Translation Solutions for Many Languages: Histories of a flawed dream (Bloomsbury Advances in Translation) 1474261108, 9781474261104

Citation preview

Translation Solutions for Many Languages

Bloomsbury Advances in Translation Series Series Editor: Jeremy Munday, Centre for Translation Studies, University of Leeds, UK Bloomsbury Advances in Translation Studies publishes cutting-edge research in the fields of translation studies. This field has grown in importance in the modern, globalized world, with international translation between languages a daily occurrence. Research into the practices, processes and theory of translation is essential and this series aims to showcase the best in international academic and professional output. Other titles in the series: Corpus-Based Translation Studies Edited by Alet Kruger, Kim Wallmach and Jeremy Munday Community Translation Mustapha Taibi and Uldis Ozolins Global Trends in Translator and Interpreter Training Edited by Séverine Hubscher-Davidson and Michał Borodo Music, Text and Translation Edited by Helen Julia Minors Quality in Professional Translation Joanna Drugan Retranslation Sharon Deane-Cox The Pragmatic Translator Massimiliano Morini Translation, Adaptation and Transformation Edited by Laurence Raw Translation and Translation Studies in the Japanese Context Edited by Nana Sato-Rossberg and Judy Wakabayashi Translation as Cognitive Activity Fabio Alves and Amparo Hurtado Albir Translating For Singing Mark Herman and Ronnie Apter Translation, Humour and Literature Edited by Delia Chiaro Translation, Humour and the Media Edited by Delia Chiaro Translating the Poetry of the Holocaust Jean Boase-Beier

Translation Solutions for Many Languages Histories of a Flawed Dream Anthony Pym

Bloomsbury Advances in Translation

Bloomsbury Academic An imprint of Bloomsbury Publishing Plc

Bloomsbury Academic An imprint of Bloomsbury Publishing Plc 50 Bedford Square London WC1B 3DP UK

1385 Broadway New York NY 10018 USA

www.bloomsbury.com BLOOMSBURY and the Diana logo are trademarks of Bloomsbury Publishing Plc First published 2016 © Anthony Pym, 2016 Anthony Pym has asserted his right under the Copyright, Designs and Patents Act, 1988, to be identified as the Author of this work. All rights reserved. No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage or retrieval system, without prior permission in writing from the publishers. No responsibility for loss caused to any individual or organization acting on or refraining from action as a result of the material in this publication can be accepted by Bloomsbury or the author. British Library Cataloguing-in-Publication Data A catalogue record for this book is available from the British Library. ISBN: HB: 978-1-4742-6110-4 ePDF: 978-1-4742-6113-5 ePub: 978-1-4742-6111-1 Library of Congress Cataloging-in-Publication Data Names: Pym, Anthony, 1956- author. Title: Translation Solutions for many languages : histories of a flawed dream / Anthony Pym. Description: London ; New York : Bloomsbury Academic, [2016] | Series: Bloomsbury Advances in Translation Identifiers: LCCN 2015040022| ISBN 9781474261104 (hardback) | ISBN 9781474261135 (epdf) | ISBN 9781474261111 (epub) Subjects: LCSH: Translating services. | Translating and interpreting--Study and teaching. | BISAC: LANGUAGE ARTS & DISCIPLINES / Linguistics / General. | LANGUAGE ARTS & DISCIPLINES / Translating & Interpreting. Classification: LCC P306.94 .P96 2016 | DDC 418/.02--dc23 LC record available at http://lccn.loc. gov/2015040022 Series: Bloomsbury Advances in Translation Typeset by Fakenham Prepress Solutions, Fakenham, Norfolk NR21 8NN

For my students.

Contents List of Figures List of Tables Preface Acknowledgements Notes on Transcriptions, Back-Translation and Capitalization 1 2 3 4 5 6 7 8 9 10 11 12

Charles Bally and the Missing Equivalents Vinay and Darbelnet Hit the Road A Tradition in Russian and Environs A Loh Road to China Spontaneous Combustion in Central Europe? Cold War Dalliance with Transformational Grammar Forays into Romance Meanwhile Back in German Disciplinary Corrections Going Japanese The Proof of the Pudding is in the Classroom A Typology of Translation Solutions for Many Languages Postscript: The Flaw in the Dream

Notes References Index

viii ix x xvi xviii 1 15 37 73 101 121 137 151 163 185 205 219 243 247 263 279

List of Figures 9.1. ‘Degrees of cultural transposition’ (from Hervey and Higgins 1992: 33) 168 9.2. Translation strategy flow chart for ‘culturally marked segments’ (realia), from Mayoral Asensio and Muñoz Martín (1997: 166; Muñoz Martín 1998) 181 12.1. Tentative positioning of main solution types in terms of accessible information on start culture (horizontal axis) and perceived location of item with respect to start culture (vertical axis) 235

List of Tables 2.1. Vinay and Darbelnet’s general table of translation solutions

24

3.1. Possible development of solution types in Russian translation theory

48

3.2. Torop’s model of translation solutions

69

4.1. Possible alignment of categories from Vinay and Darbelnet (1958) and Loh (1958)

83

4.2. Possible correlations between Russian typologies and Loh (1958)

83

8.1. Schreiber’s model of translation procedures

159

9.1. Translation solutions according to Newmark

165

9.2. Translation solution types in several French and Spanish translation manuals

171

12.1. A typology of translation solution types for many languages

220

Preface The past fifty years or so have seen many lists of the ways translators (which here includes interpreters) can solve problems. Although they are mostly presented as ‘procedures’, ‘techniques’ or ‘strategies’, the items in the lists are all based on the solutions identified in translations, not on the way translators think, so here I will call them ‘solution types’. The lists are also based on the study of different language pairs, which might explain why they are so different. With that idea in mind, I started collecting the lists, evaluating and comparing them, to see if it might be possible to produce a list that could work for many language pairs, if not for all. What I found was surprising, first because of the way the historical relations between the lists effectively traced the development of an international Translation Studies – a long extended conversation between scholars, over and above the concerns of specific languages. The second unexpected aspect was the extent to which the lists, apparently anodyne and technical, were in fact highly politicized, seeking to advance the agendas not just of various linguistic nationalisms but also of a few governmental regimes – the research became a history in which Hitler, Stalin and Mao all play roles. I had set out to produce a simple typology of solution types for many languages – and I do propose one in the last chapter – but the more exciting discoveries were made along the roads that led there. This project actually had its origins in something slightly different. In 2012–13 Kirsten Malmkjær and I were working on the use of translation in the teaching of foreign languages (Pym et al. 2013). Part of that research involved asking language teachers why they did not use translation in class; one of the not infrequent replies was along the lines of, ‘We prefer communicative methods.’ In the minds of many language instructors – and indeed in the imagination of the Department of Education in the United Kingdom (Pym et al. 2013: 88) – translation is simply not a communicative activity. And when we then asked the respondents whether translators might actually communicate something to someone, the reply, if not a puzzled look or gesture of dismay, sometimes went along the lines of, ‘That’s not the kind of translation we use in class.’ What was at stake in those encounters, and indeed in that entire research project, was a constant misunderstanding or slippage in the meaning of the

Preface

xi

word ‘translation’. This was later confirmed when Kurt Kohn, an expert in language education in Germany, responded to our study by claiming that, in the popular preconceptions of students, instructors and indeed the general populace, ‘translation means literal translation’. So the great challenge is to make people aware that translation involves more than literalism, that there are the creative processes of what we might call ‘communicative translation’, not just in translator training but in language learning as well. So what is the best way to get people to think of translation as a communicative activity? In German, the trend has been to see translation and interpreting as parts of ‘linguistic mediation’ (Sprachmittlung), or even to have the term ‘linguistic mediation’ replace the concept of translation altogether as a pedagogical activity. This ‘mediation’ would include talking about a foreign text, giving the gist, engaging in bilingual conversations, plus a range of associated exercises that include traditional translation and interpreting, albeit perhaps privileging the spoken word more than the written. In English, a rather less successful but still innovative trend is to use ‘translanguaging’ as the general term, covering the spoken and written use of L1 in the L2 language class. In both these cases, translation becomes part of something much wider; it loses its status as a specific skill-set that many language learners still want to learn and use; in effect, it continues to be locked into the common preconception that anything called ‘translation’ is basically literal translation, as close to the foreign text as possible. Given the extent of these misunderstandings and cross-purposes, there is a very good case to be made for abandoning the term ‘translation’ altogether. Perhaps we should simply replace it with ‘mediation’ or ‘translanguaging’ (or ‘localization’, in technological contexts)? Such a shift might work, particularly in primary and secondary education. Of course, it would risk losing a lot of the knowledge we have gained about the more traditional kinds of translation; it would involve the implicit capitulation of the dynamic and still-evolving academic discipline called Translation Studies. But those costs would not be irreparable. I think there might be an alternative way. All we need is some strategy whereby the term ‘translation’ resists reduction to ‘literal translation’ and at the same time remains a specific skill-set, able to be taught in language classes and to be usefully implemented in the training of volunteer and professional translators. So where might this miraculous solution be found? Perhaps it has been staring us in the face for more than fifty years, in one of the most ideologically conservative nooks of Translation Studies.

xii Preface

When Vinay and Darbelnet drew up their list of ‘translation procedures’ in 1958, their main categories were as follows, just so we know what we are talking about: Borrowing Calque Literal translation Transposition Modulation Correspondence or Reformulation (équivalence) Adaptation To produce that list, the French linguists made numerous assumptions about language and translation, and we nowadays find some of those assumptions questionable, if not entirely distasteful. The linguists, like most of those around them, unthinkingly accepted models that were essentialist: they never doubted that each phrase had a fixed meaning to be rendered by the translator. And their essentialism was implicitly nationalist: the linguists were sure the fixed meanings and acceptable turns of phrase had the same value for all citizens speaking the one language of a given nation. These days we are more inclined to accept that texts have multiple interpretations and that societies are linguistically segmented. This makes the old linguistics hard to read, let alone accept. And yet what Vinay and Darbelnet came up with, those seven major types of translation solutions, still have something to say in the context of current debates, most notably with respect to doubts about the meaning of the word ‘translation’ and the perennial question of whether humans translate in ways that machines do not. What their typology says is basically that translators can do a very wide range of things. Yes, the linguists believed in essentialist equivalence, but there was nothing simple about it: above the line in the list of solution types you have degrees of what another categorization might call ‘directional equivalence’, of the kind that works when you go from language A to language B but not from B to A; and below the line we find degrees of ‘natural equivalence’, all based on the hope that good translations are already in the target language, apparently waiting to be found (on these categories, see Pym 2010/14). In practical terms, the range of options above the line says something like the following: Yes, translation could be basically literal translation, when it can be, and that is the starting point (indicated by the line, although we will see this radically modified with respect to Asian languages), but look at all the other things translators do! Look here, all these frontline problems concerning new terminology and

Preface

xiii

phraseology are also faced by translators (who must decide between various flavours of Transcription and Calque). And then, in the second part of the list, below the line of literalist solutions, look at all the more creative solutions; see how much can be done to respond to differences in semantic structures, discursive habits, cultural differences and the like: Transposition, Modulation, Correspondence, Adaptation and so on, down into a list of lesser ‘prosodic effects’ (added to the above list) that include Explicitation and Compensation. In sum – one might say to those who still think all translation is literal translation – just look at all the decision-making skills that can come into play in the act of translation. So how can you tell us translation is just ‘literal’? Think of all the cognitive work of translators who have to decide between such alternatives – surely they are performing more than a simple mapping operation? In order to choose between most of the options, translators must necessarily be thinking in terms of a communicative activity: this solution for this text, for this text user, for this client, to achieve this purpose. Hence the question at the origin of this research: can we take the old typologies of translation solutions, clean away their often questionable ideologies, cover over their various lacunae and contradictions and use them as a way of addressing our more contemporary desire to show that translating is communicative? Of course, that question has led to many others, taking me down some rather strange paths. Where did the early typologies come from anyway? Can they be made more logical or at least easier to understand? How do they relate to the different meanings of the word ‘equivalence’? And especially, how can they now be adapted to work with Asian languages? As this project developed, it connected with several other ideas that have been of interest in recent years. One of them is the notion that translation is not just communication but is more importantly part of an event, requiring the active involvement of numerous participants. To achieve involvement, translators have to work with the full array of rhetorical devices their languages offer, taking communicative risks rather than simply staying close to convention. One could pick up some of this by re-reading Nida in terms of Tannen’s conversation analysis, for example, just as some have gone further into the sense of ‘event’ in Badiou (1988), where the only truth is the one enacted in the event itself (a concept that still sits poorly with the representative function of translation). We are increasingly aware that cross-cultural communication concerns more than giving the right information – it is successful when it involves the rhythms and emotions of the people it engages. As I went reading around Vinay

xiv Preface

and Darbelnet, going through the comparative stylistics of the early twentieth century (especially Charles Bally), so much of what I found was precisely about affectivity in language, the liveliness of discourse, the variety of expression offered in everyday phrases, and indeed an underlying love of language in all its richness, as indicated in the countless examples in which the linguists of translation once used to lose themselves, often at the expense of clear thought. True, the comparative stylistics that fed into translation theory suffered from the nationalist fantasy that all social classes participate equally in the plays of language; true, their general view of affectivity in language curiously stopped at sentence level. Still, I suspect a link has been lost somewhere: early structuralism, perhaps badly understood, cut us off from that sense of style as the liveliness of language. Written within that distanced tradition of stylistics, the initial typologies of translation solutions still convey a sense of playful diversity, of translating as a constant choice between alternatives and of languages differing not just in terms of syntax but more profoundly in the various ranges of what ‘feels right’ and how far one can tease the limits of that feeling. This is a point where my interests connect with those of Douglas Robinson (2013a, 2013b), whose concerns with performativity have nevertheless drawn him to a very different history of twentieth-century linguistics: the dynamic collective origins of norms and laws. Here I am more concerned with parts of language where there are still viable alternatives, rather than developed laws. If they were ever meant to be prescriptive rules, the lists of solutions were mostly too badly constructed to have been successful in those terms. Their virtues have always been elsewhere, mainly pedagogical, and those virtues are worth recuperating. I suspect that the typologies of solution types are not taught much these days. At the Middlebury Institute of International Studies at Monterey (where most of this research was carried out), only the Russian programme has regularly presented anything like them in recent years. Instead, what seems to resonate most among my colleagues are the large binary oppositions like ‘domesticating’ versus ‘foreignizing’ translation, ‘dynamic’ versus ‘formal’ equivalence, and so on. The great binarisms tell our students they can basically look one way or the other, toward the start text1 or toward the target receiver, and that there are many ideological and political factors informing those choices. The binarisms are perfect for filling our programmes with grandiloquent theories about translation, rather than teaching students how to do actual translations (these days marginalized in some one-year multilingual Masters, constituting one of the great shams of our times). The binarisms certainly have their place, but that place is not close to the many practical sessions where students should

Preface

xv

be learning how to solve translation problems, quickly and with professional confidence. In order to talk about those problems, in order to have a basic metalanguage to start dealing with the multiple factors involved, we need those short lists of solution types, and we need them in a form that is not swiftly reduced to binary oppositions. My ongoing suspicion is that the solution types are currently ignored not particularly because they are old, language-specific, bereft of psychological underpinning, limited to sentence level and derived from outdated linguistic approaches – although those are all serious shortcomings. I suspect the problem is more simply that many of the categories were badly constructed in the first place, and no one has yet been able to bring together the various insights that have been made in different parts of the world, in different languages. Perhaps, just perhaps, a more robust, coherent list will help us train translators a little better. In many languages?

Acknowledgements This research has benefited from the expertise, advice and hard work of many people. I have learned much from the careful editing of Professor Jeremy Munday, whose work is much appreciated. My sincere thanks also go to Dr Esther Torres-Simón, who has suggested numerous corrections, has worked on data from my classes in Monterey, has helped with Japanese and has more generally kept my research life in some kind of order; Nune Ayvazyan, who has done most of the detective work in Russian; Dr Hui Ting Ting Maggie, whose doctoral research was at the origins of this project and who found Loh and Fedorov in Hong Kong, then helped me understand them; Dr Liu Fung Ming Christy, for further help with Loh and corrections of Chinese; Professor Debbie Folaron, who found and sent several introuvables from Montreal; Professor Brian Baer, for timely information and advice on things Russian; Brian Mossop, for general enthusiasm for the same Russian things; Dr Susanna Witt, for advance notice of her work in the Soviet archives and some timely advice; Oleksandr Kal’nichenko, for detailed information on the Ukrainian theorists of the 1920s and a highly pertinent correction of the chapter on Soviet theories; Professor Zhang Ling and Professor Ye Zinan, for help with locating and understanding biographical information on Loh Dian-yang; Professor Ye Zinan, again for access to his own work on translation solutions; Dr Jaroslav Špirk, for his translations from Czech and Slovak, along with his helpful guidance through theories in those languages; Igor Tyšš, for further guidance through the history of Slovak thought on translation; Russell Scheinberg, for a year or so of discussions about English–Japanese translation and for allowing me to draw liberally from his examples; Dr Masaru Yamada, for help with rubi characters; Dr Akiko Sakamoto, for her advice and corrections of the Japanese chapter; Professor Kayoko Takeda, for a further revision of the Japanese chapter; Professor Andrew Chesterman, for correcting my occasional misunderstandings of his work and for making numerous insightful comments on the entire text (it is always a great pleasure to argue with Andrew); Dr Alberto Fuertes Puerta, who has been an invaluable sounding board for many of these ideas; Dr Carlos Teixeira, for some very careful reading and intelligent comments; Professor Ilse Feinauer and Masters students at Stellenbosch, who were subjected to an interim set

Acknowledgements

xvii

of solution types; and most especially the students in my Practicum classes in Monterey over the years, from whom I have learned much about translation and languages, east and west. My special thanks to Michael Schreiber for permission to translate his table of solutions (Table 8.1) and for agreeing with the translation; to Taylor and Francis for permission to reproduce the ‘cultural transposition’ diagram from Hervey and Higgins (Figure 9.1); to Ricardo Muñoz for his kind permission to reproduce the ‘Translation solution types in several French and Spanish translation manuals’ (Table 9.2); and to Ricardo Muñoz and Roberto Mayoral for permission to reproduce the ‘Translation strategy flow chart for “culturally marked segments”’ (Figure 9.2). A previous version of Chapter 1 has been published as ‘Charles Bally and the missing equivalents’ in Forum (2014): 45–63. Some of the material in Chapter 3 is drawn from ‘The case of the missing Russian translation theories’, written with Nune Ayvazyan and published in Translation Studies (2014), while its methodology is defended in ‘A response to the response to “The case of the missing Russian translation theories”’ in Translation Studies (2015). A more complete description of the data used in Chapter 11 has been published as ‘The pedagogical value of translation solution types’, with Esther Torres-Simón, in Perspectives: Studies in Translatology (2014).

Notes on Transcriptions, Back-Translation and Capitalization Russian names have been transcribed in accordance with the Translit system, used by the Russian government in its passports. Chinese has been transcribed into Simplified throughout, for the sake of uniformity. My sincere thanks go to Nune Ayvazyan and Ye Zinan respectively. The translation examples incorporate various styles of back-translation, sometimes morphological, sometimes at phrase level. In each case the level has been selected in accordance with the linguistic features to be highlighted. The names of the various translation solution types have been capitalized in cases where they are naming the concepts formulated by the various theorists. When not, then not.

1

Charles Bally and the Missing Equivalents

Chapter summary: In the early twentieth century, the Swiss linguist Charles Bally developed a comparative stylistics based on translation and functional equivalence. The difficulties encountered by his students when translating between French and German had alerted him to the peculiarities of the two languages, which set in motion his analyses. Bally claimed that his stylistics could constitute a ‘method’ for translation at a more advanced level, where the approach ideally enabled students to ‘work through the idea’ rather than mechanically move from form to form. Bally thus implicitly recognized two kinds of translation – one mechanical, the other communicative – and a loose dynamic relation between linguistic description, language teaching and the learning of translation skills. He stopped short, however, of proposing any specific translation solutions. Others would later do precisely that.

Vinay and Darbelnet (1958/72) do not call their work a ‘typology of translation solutions’. Far from it: their approach is named ‘stylistique comparée’, comparative stylistics. So what does that term mean? Where did it come from? A history of German Translation Studies (Siever 2010: 32) identifies the founding of comparative stylistics with the publication of Alfred Malblanc’s Pour une stylistique comparée du français et de l’allemand in 1944. Vinay and Darbelnet do indeed recognize a debt to Monsieur Malblanc, who was in charge of the series where they published their work, chez Didier in Paris. However, Vinay and Darbelnet themselves attribute the origin of comparative stylistics not to Malblanc but to the Swiss linguist Charles Bally, who, they say (1958/72: 32), formulated the corresponding concept of ‘external stylistics’ in 1932, which was then merely applied by Malblanc under the name of ‘comparative stylistics’. Vinay and Darbelnet certainly owed much more to Bally than they did to Malblanc. Indeed, they cite Bally some twenty-three times in the first edition of their Stylistique comparée, whereas Malblanc gets little more than a few nods of respect.

2

Translation Solutions for Many Languages

So if you are going to find out where Vinay and Darbelnet were shooting from, Bally could be the man to look at.

Charles who? Charles Bally is widely recognized as the founder of linguistic stylistics and, within that frame, modern phraseology. It would be nice to say he is a forgotten figure in English-language linguistics,1 but I am not sure he was ever really present enough to have been forgotten. He is usually referred to as no more than the co-editor, alongside Albert Sechehaye, of Saussure’s Cours de linguistique générale (1916), compiled from what are called ‘students’ notes’. In discussions of that coediting, Bally is sometimes mentioned as sharing the blame for concealing many of the less systemic questionings that have since been discovered in Saussure’s Écrits (2002) – Bally would be the over-eager apprentice who excessively simplified and formalized his master. And yet, as Bally’s independent interest in stylistics might suggest, that characterization is not entirely fair. Charles Bally (1865–1947) was only eight years younger than Saussure. Prior to attending Saussure’s lectures, he had studied classics in Geneva, completed his doctorate on Euripides in Berlin from 1886 to 1889, been a tutor to the Greek royal family for four years and had taught as a privat-docent at the University of Geneva from 1893, in addition to teaching at a business school and an introductory grammar school (Progymnasium). That is, he was not just a student and not just a linguist: he earned his keep by teaching French, particularly as a foreign language. This teaching experience, coupled with his knowledge of German linguistics, brought him to stylistics. Bally published his Précis de stylistique in 1905 and his Traité de stylistique française in 1909. Since Saussure did not begin teaching his course in general linguistics until 1907, there is little reason to see Bally as simply applying Saussure. Some of the systemic notions were actually formulated in nuce by Bally prior to Saussure, albeit without the revolutionary clarity. From 1913 Bally held the chair of General Linguistics and Comparison of Indo-European Languages at the University of Geneva, the chair that had previously been held by Saussure. So what does ‘stylistics’ mean in this context? Bally set out to study the ‘expressive resources of a whole language’, not just of some expressions or of some writers (1905: 11). In this, his project differed from the kind of literary stylistics that would study the language used by one author, for example, and



Charles Bally and the Missing Equivalents

3

it refrained from assuming, as had some of his German predecessors, that the works of great writers could help shape a whole language. Bally wanted to look at the entire French language, at all its levels of usage, and give a systematic description of its peculiarities at the level of parole, linguistic expression, rather than the underlying obligatory grammar. This is where his association with Saussurean tradition becomes difficult: Bally set out to study a whole language, but he did not want to do so at the level of obligatory grammatical rules. Instead of cold oppositions and pure differences of the kind one might find in the later Prague structuralism, for example, he emphasized the affective uses of language, the musical and rhythmic qualities, the varieties of possible expression and the constant imbrication of utterance with ‘life’, in a sense that, given the dates, would owe much to Bergson. The important point here is that stylistics, although assuming the same extension as the entire language system (langue), is not looking for abstract binding laws. Instead, it is an affair of tendencies, possibilities, options, relative frequencies and what ‘sounds right’ in a particular language. Bally’s various formulations of this point are often wavering. In his Traité de stylistique française, in almost confessional mode, he puts it as follows: With respect to method, one must distinguish between what is normal [la règle] and what is a law. Stylistics has to study expressions without adopting a systemic frame of mind; it has to identify the general tendencies without excessive concern for rigor. (1909/51: 1; my translation here and throughout)

You might imagine how difficult it is to translate the word ‘règle’ in the above passage. A ‘règle’ is a rule, but here it is not a ‘law’. So what is the difference between a rule and a law? In Bally’s mind, it is a question of attitude and approach: he works close to the examples, of which there are many, and he mostly works bottom-up, from the examples to the abstraction, rather than top-down, from the hypothetical law that then seeks examples. So much for Bally as the ‘student’ who over-systematized Saussure! The one reproach one might make of Bally’s own work is precisely his lack of clarity with respect to theory and methodology (as we shall see below). But that fault might also be called an excessive love of language. Interestingly, Vinay and Darbelnet echo Bally’s vague distinction as far as they can: ‘Grammar is the domain of servitudes, whereas options belong to the domain of stylistics’ (1958/72: 32). Vinay and Darbelnet immediately qualify this with respect to their own research project, messing up the apparently linear boundary:

4

Translation Solutions for Many Languages Even though options dominate internal stylistics, which is focused on the facts of expression [as in Bally], external stylistics [what Vinay and Darbelnet call ‘comparative stylistics’] concerns both servitude and option. (1958/72: 33)

That is, when it is a question of speaking or writing in French only, then stylistics concerns the choices you can make. But when you are moving from English or German into French, then it seems there are grammatical laws and frequent patterns of expression that you simply have to accept. This accounts for why Vinay and Darbelnet, like many of those who have followed them, make distinctions between the ‘obligatory’ and ‘optional’ modalities of the various solution types they present, whereas this kind of distinction is of little interest to Bally. One might nevertheless ask why such a distinction between two kinds of stylistics should be made. Surely there are rules (or ‘laws’) and options for the speaker of French, just as there are rules and options for the translator going into French? If a stylistics can be formulated just for French, why should it be fundamentally different from a stylistics that works for movements between French and other languages? I can see no strong reason for any difference. Indeed, one might remain faithful to Bally, to the origins of linguistic stylistics, if everything in stylistics were regarded as optional and grammar were left to the laws of grammarians. Which is something we might want to do.

A methodical linguist The real interest of Bally is that his starting point was translation, his methodology was based on equivalence, he recognized that his whole procedure was in effect a way of training translators and yet he never said anything systematic about translation. You will not find his name on any list of translation theorists. When I attempt to grasp his methodology for translation, I am thus looking for something that is fundamentally not there, but could be there. Bally’s major text, for methodology and much else, is his Linguistique générale et linguistique française, first published in 1932. As the title indicates, the first part elaborates concepts and procedures that are presumed to work for all languages, while the second part applies those concepts and procedures to the stylistics of French. Beneath the cover, though, what is happening in this book is a constant process of translation between French and German. In fact,



Charles Bally and the Missing Equivalents

5

the method is explicitly marked by Bally’s experience as a teacher of French to German-speaking students: As I went through French texts with foreign students, as I translated German texts into French with them, I was naturally led to reflect on the difficulties they encountered and the differences they found between the two languages. These observations gradually became impressions of a more general order, which in turn allowed me to glimpse the deep and divergent tendencies of the French and German tongues. Text analysis and translation thus ceased to be the purposes of my work and instead became practical ways of defining these general overviews. (‘Préface à la première édition’, 1932/65: 8)

The bottom-up approach is very clear. So, too, is the functional role of translation. Further on, Bally gives special weight to the need to study the mistakes that Germans make when they speak and write French, and the difficulties encountered by translators. Both kinds of data are seen as revealing the key points of different linguistic systems (1932/65: 28). And so, despite a title that talks about universal linguistics and French, this is a work that starts from translations between French and German (especially if we can envisage an explication du texte as a mode of translation) and is actually full of detailed comparisons of the way French and German differ in expressive modes and capacities. Much of the work functions as a direct precedent of Malblanc’s comparison of French and German (1944) and thus of the linguistic and institutional space necessary for Vinay and Darbelnet (1958/72). Bally, however, is not directly concerned with helping anyone translate. As can be seen from the citation above, his working method takes him from the particular to the general, from translational activity to an awareness of the differences between languages. And then, in the presentation of his findings, this inductive method is reversed: he goes from the principles of ‘general linguistics’ to the specificities of French as compared with German (with the title blatantly omitting not just the German but also any hint of how translation set the whole exercise in motion). Bally takes pains to point out that his first concern in all of this is ‘methodological’, since adherence to strict method will protect him from the errors of other kinds of stylistics: Our mother tongue is constantly mixed with our own life, with the life of our society and of the nation; how can this not give rise to erroneous views, displacing disinterested observation with a priori engagements and conventional ideas, introduced from the outside without control? (1932/65: 13)

6

Translation Solutions for Many Languages

The enemy of scientific objectivity is the very liveliness of an embodied language and its imbrication in society. Yet there is no suggestion of social determinism here: Bally is keenly aware that there is no one-to-one parallelism between the way a language evolves and the societies that speak the language, not only because languages change much more slowly than other institutions do, but also because they are imposed on speakers of other languages. So there is no consistent correspondence between language and culture (1932/65: 14): Bally’s cannot be a linguistics of national psychologies. That would be the first historical specificity of his approach. A second idea that he challenges, as mentioned, is that since great speakers apparently shape a language, stylistics should be of individual expression. Bally retorts, in Humboldtian mode, that ‘thought works on a language, but language also shapes thought’ (1932/65: 15). A third traditional tenet to be questioned is that one language is more logical or clearer than another, where Bally has no trouble indicating that German is clearer to Germans and French is clearer to the French. These notes are important because Bally was working against previous studies that were overtly shaped by precisely those presuppositions, notably Fritz Strohmeyer’s historical-psychological portrait of French in Französische Grammatik auf sprachhistorisch-psychologischer Grundlage (1921) and Der Stil der französische Sprache (1924), as well as Karl Vossler’s Frankreichs Kultur und Sprache:  Geschichte der französischen Schriftsprache (1929), which had generally seen French as shaping and being shaped by works of literature. In order to break with all of that, Bally proposes that the linguistic method – the method that is linguistics rather than anything coming ‘from outside’ – is based on pure comparison. The foundational assumption of that comparison is that several elements, within the one language, can share ‘functional equivalences’ (équivalences fonctionnelles), defined as follows: Pieces of a grammatical system can replace each other [peuvent s’échanger] because of their shared function, even when their semantic and stylistic values are not identical. Compare the following ‘functional equivalents’: the house my father owns, of which my father is the owner, belonging to my father, owned by my father and finally, my father’s house. (1932/65: 40)

The functional equivalence here allows identification of … what, exactly? Bally does not really say. It is not just an underlying grammatical category like the genitive or the possessive, which would be simple and constant enough: OBJECT + DE + OWNER in French and OBJECT + OF + OWNER or OWNER + ’S + OBJECT in English. That would lie in the land of grammatical laws,



Charles Bally and the Missing Equivalents

7

whereas Bally’s example includes structures that do not correspond to those laws (‘belonging to’, etc.). The examples concern more than grammar; they are marked by semantic values: these equivalent expressions can scarcely be aligned with cases like ‘My father’s smile’ or ‘My father’s death’. The equivalence here stops short of producing functions for each possible concept in a language. The equivalence of interest to stylistics is somewhere between the grammar and the dictionary. However, Bally’s project, despite its firm insistence on method, seems to have no clear grid for organizing the ‘functions’ that this process of comparison can isolate. What he presents is a tool for investigation, not a map of the area to study. It is rather like giving a young geologist a pick-hammer and saying, ‘Off you go, find some geology!’ On one level, Bally’s use of equivalence could be among the most obvious; any speaker can say the same thing (or almost) in different ways: reformulation exists. Bally nevertheless insists that such functional equivalences are ‘the basis of all linguistic systems’ (1932/65: 35), which is a statement that requires some thought. If it is true, then systems are not based on pure differences, as in a simplistic reading of Saussure. If it is not true, then language could not talk about language and there would be no linguistics. Bally, as a lover of language, does not want to go down either of those paths. He asks us to accept that, with some give and take, reformulation does exist and different languages do restrict it in different ways. Bally perhaps wanted the rigor of a methodological tool to save him from the presuppositions of the German tradition, but he was not a blind Saussurean: he never accepted the rigor of a system that had no life – a language was something dynamic waiting to be explored. Does reformulation have any theoretical right to exist? That is not a question we can answer. Yet its possible existence should be a key element in the training of translators and interpreters. I remember a lecture given by Daniel Gile in Monterey whose sole point was that reformulation is always possible within any one language, so there was no reason why any translator or interpreter should feel obliged to stick to the start-text form. If the author or speaker can reformulate, then so can the translator or interpreter. A linguistics that actively recognizes reformulation should be of some interest to translators.2 Bally’s use of functional equivalence has certain productive virtues. For instance, he recognizes a mode of ‘grammatical equivalence’, actually based on word classes, that allows for the concept of ‘transposition’ to be formulated: A linguistic sign can change its grammatical value and yet retain its semantic value by adopting the function of a lexical category (noun, verb, adjective,

8

Translation Solutions for Many Languages adverb) to which it did not previously belong. Thus the nouns planet and country, without changing their signification, become (functionally) adjectives in planetary (system) and country (house) … (1932/65: 116)

The resulting notion of ‘transposition’ would be of some importance for Vinay and Darbelnet. A third example of equivalence as a methodological tool is rather more exciting and might resonate with any scholar struggling to find time and silence in which to grapple with old books. Bally claims that the following series of utterances and actions could be functionally equivalent to each other (1932/65: 41): 1. I want you to leave. 2. I order you to leave. 3. You have to leave. 4. You must leave. 5. Leave the room! 6. Get out! 7. Out! 8. [Gesture toward the door and facial expression indicating irritation.] 9. [Throwing the person out.] Bally observes that the series moves from the linguistically explicit to the linguistically implicit. That is, the assumed functional equivalence allows the linguistic category of ‘implicitation’ to be identified, along with some of its degrees – this, too, would appear in Vinay and Darbelnet. Bally actually claims that each movement down the list sees a logical category disappear from the expression, while the mind of the receiver (the person who is supposed to leave) can still make up for the missing expression. He thus manages to build a psychological observation and a linguistic category on the basis of a methodological banality: any act of comparison requires the assumption of an invariant or common core that allows the differences to be identified; all linguistics must be doing something like this; the name for the operator in Bally is ‘equivalence’. I pause a moment to consider how much more Bally could have done with this same example. For instance, observe the way the first person disappears as you move down the list, then the second person, the third, and finally we see the linguistic, corporeal and interpersonal actions flow into each other. Why not posit that, instead of the linguistic expression becoming increasingly implicit



Charles Bally and the Missing Equivalents

9

as we move down, one can see the potential action, the possible grimace and the linguistically marked anger all becoming implicit as we move up: ‘I want you to leave’, although formally explicit, implies that relative failure to comply may activate any or all of the other utterances and actions. Put another way, ‘I order you to leave’ is a well-formed performative that can only actually perform to the extent that its force and relation with action are implicit (and explicit in all the other expressions). Bally’s example seems on the cusp of a performative linguistics: he is relating language with action in a novel and intriguing way. And yet he has little interest in any of that. He makes no comment on the system of pronouns – Benveniste (1966) would later claim that the pronominal system, not equivalence, was the basis of human languages; Bally did not want to go there (and this in a chapter called ‘Théorie de l’énonciation’). The kind of performative French linguistics that Robinson (2003) explores in relation to translation was developed completely to the side of Bally’s stylistics. Let me take the example just a little further. One might posit that the physical action at the bottom of the list might be possible within all languages and cultures (the restricting condition concerns power, not different language systems); the grimace and gesture at the penultimate step could be more culture specific (an Italian would be cutting one hand with the other), and so on up the list, such that the utterance most embedded in its language would be the one at the top, marked by the explicit first person, possibly with marked degrees of politeness (as in Japanese?). This seems counter-intuitive, since the most explicit utterances should be the least embedded and thus the easiest to translate (cf. Pym 1992/2010: 126). I leave those paradoxes for another day. The interesting point, here, is that it would have cost nothing for Bally to recognize that innumerable languages could articulate the action and the grimace; many could capture the monosyllabic interjection; and why not imagine, with increasing frustration, an enraged Heinrich reverting to his mother-tongue ‘Raus!’, while an infuriated Inés screams ‘¡Fuera!’, as one tends to do, furtively, when driving in heavy traffic and the like. Why does Bally not mention this? Just as Bally does not want to see performativity, he does not want to entertain the possibility of interlingual equivalence. For him, equivalence is something that operates within one language system only. That is strange. When Bally comes to compare French and German, there seems to be no mention of any kind of equivalence between their items.3 The two languages somehow line up as if by magic: ‘Wenn du kommst: Quand tu viendras’ (1932/65: 355). It would seem logical to apply the same principle to such interlingual alignment, but that just does not happen (as far as I can see).

10

Translation Solutions for Many Languages

In his two-volume Traité de stylistique française (1909/1951), Bally thus pushes the notion of equivalence into territory that is extremely idealist and unnecessarily monolingual. He talks about pieces of language having a ‘logical equivalent’, which is held to be a ‘simple notion’ that can then be used in order to classify those pieces of language (1909/51: 1.30, 96). However, there is also ‘equivalence in context’ (since words function in contexts) and from there we find the combination of the two kinds of equivalence giving something like a pure language: When the logical equivalence of an identifying term is accompanied by an equivalence in context, one has constituted a mode of intellectual expression or language of the pure idea. (1909/51: 1.105)

What seemed like no more than a methodological assumption in the earlier works here becomes a view of language as expressing a series of simple ideal forms. Then again, at one point realizing that in some cases there are numerous potential equivalents and their differences are not intangible (or, to speak with Derrida, there is no ‘transcendental signified’), Bally reverts to the argument of methodological necessity: ‘these equivalences are only justifiable to the extent that we are only looking for points of comparison so that we can identify the stylistic value of the linguistic facts’ (1909/51: 1.109).

For or against translation? That reservation notwithstanding, Bally claims, right at the end of the explanatory part of the Traité, that his method of reducing pieces of language to ideas as ‘logical equivalents’ in ‘equivalents of context’ can and should constitute – wait for it – ‘a veritable method of translation’ (toute une méthode de traduction) … which he will leave for another day (1909/51: 1.138).4 The second volume of the Traité then comprises numerous exercises for the learning of French stylistics, and many of the activities do indeed involve translation. But it was not Bally who wrote the ‘méthode de traduction’. As I have noted, Bally was thinking from a translational situation; his method was based on a concept of equivalence that could have been made translingual; and he himself saw that his stylistics offered a way of training translators. So why did he not have anything systematic to say about translation? At the very beginning of the Traité we find a few words that are rather less enthusiastic about the issue:



Charles Bally and the Missing Equivalents

11

The formal learning of a language system [langue] tends to favor an automatic and mechanical view of language use [langage]; too often it makes one think in a passive way, which is not the best preparation for disinterested scientific observation. […] The pedagogical use of translation [la méthode de traduction] promotes this mechanical work precisely because it accustoms the learner to exchange linguistic symbols – words – for each other, without obliging the mind to work through the idea. (1909/51: 1.2; italics in the text)

This is intriguing. Here at the beginning of his treatise, Bally associates translation with a ‘mechanical’ process that jumps from form to form without analysing the ‘idea’ (the idealized use of a piece of language in a situation).5 And then, at the end of the same work, he claims his analytical method, in isolating ‘ideas’, can provide exactly the same thing, a ‘method of translation’. One can only conclude that there are two kinds of translation being referred to here. The one at the beginning of the book is bad, since it is a mapping operation that goes from form to form, as in the ‘grammar translation’ teaching method where translation was basically a backward-looking check on acquisition. The second, however, is positive, since it is made to analyse precisely the situated ‘idea’ that the first concept avoids. This seems to be confirmed in the earlier Précis, where Bally similarly claims that ‘if translation is a mechanical exercise, a juggling with words, the result can only be bad’ (1905: 165). However in that work we actually find a whole ‘Appendice’ on translation, added despite the fact that ‘in theory, translation should not concern us here’ (1905: 163). So why the marginal status of translation, relegated to the appendix? On the one hand, Bally admits that the passage from one language to another, especially as manifested by the difficulties of learners, remains an efficient way of approaching what is particular to a language. On the other, however, he consistently mistrusts all translation activities that are based on word-for-word principles, or that work on words or phrases in isolation, or indeed that ask learners to translate complex passages before they are adequately prepared. Such uses of translation avoid passing through the ‘idea’, thus cultivating blindness to all the intricacies of the way subjective desires and emotions (‘l’affectivité’) operate in language. In fact, translation, thus understood, effectively shortcuts all the work of stylistics and is thus almost necessarily marginalized, as an inferior and misleading rival. There is, however, an alternative use of translation, envisaged as an advanced learning activity: When translation is introduced at the appropriate stage, not at the beginning of language studies but at a time when the learner has access to many materials

12

Translation Solutions for Many Languages that have been taught in a controlled way, it is not without value. Approached in a certain way, translation can become a touchstone for the lessons learned and the knowledge acquired. (1905: 163)

So we have a bad kind of translation, the one commonly used in the grammartranslation method of language acquisition at the lower levels, and a good kind of translation, which is a high-level application of grammar and stylistics, albeit also at the level of a check on acquisition.6 It is in this pedagogical sense that stylistic analysis can become a ‘method for translation’, while translation itself cannot be a method for learning a language. In a slightly later lecture, Bally once again allies translation with ‘mechanical correspondence from language to language’ (1910: 4) but this time he sees pedagogical composition exercises and translation as both being based on ‘synthetic combinations of expressions’ (1910: 5; italics in the text). These combinations are apparently too complex for the learner to appreciate what is actually happening in a language, in this instance in the acquisition of abstract nouns. Bally’s stylistics should once again provide a ‘preparatory study’ for such compositions and translations, basically by ‘describing the fundamental ideas expressed in the abstract vocabulary and the means of expression that language use gives us for rendering them’ (1910: 5; italics in the text). Stylistics should thus precede a better kind of translation. The bad kind of translation might be as literal as possible, applying the rules of grammar but actually doing too many things at once. The second would seem an even more complicated affair, about which Bally had very little to say apart from the fact that it should follow his stylistic analyses.

A delayed legacy Bally’s comments on translation are dated about fifty years prior to Vinay and Darbelnet: they were there, but they were by no means about to cause a revolution. And on the more methodological level, Bally’s observations on equivalence seem almost banal but could have become something else. In his Linguistique générale, equivalence is simply attesting the fact of reformulation and the necessary assumptions of comparison. In his Traité de stylistique it momentarily opens onto a very naïve idealization of semantics. And all those things probably would be mere wishful thinking were it not for the subsequent history of thought on translation.



Charles Bally and the Missing Equivalents

13

Vinay and Darbelnet, as one might imagine, eventually found in Bally an open door to walk through: he had announced a complex ‘method of translation’ based on assumed equivalences in particular situations; they just had to write it up. As I have noted, Vinay and Darbelnet also found several of their technical terms and concepts in Bally: the analyses of ‘transposition’ and ‘explicitation’ were waiting to be used, as indeed were basic concepts of ‘loan’, ‘syntactic calque’ and many of the grammatical terms and observations about the peculiarities of French. More important, there was a whole methodology there: Bally had gone from practical translation problems to the idealism of logical and contextual equivalence; Vinay and Darbelnet just had to go the other way, taking the assumed equivalences and using them to give guidelines for translation.

2

Vinay and Darbelnet Hit the Road

Chapter summary: In 1958 the French linguists Vinay and Darbelnet, working in Canada, published a highly influential list of translation solutions. They drew on Bally and borrowed some organizational concepts from Malblanc. Although their list covers a wide range of options, it is ideologically weighted in favour of solutions that protect the génie of each language: corresponding idiomatic expressions, different sports fulfilling the same cultural function, and so on. This can be seen as politically defending ‘correct French’ from the degrading influences of American English and Canadian bilinguals. An ideology of a ‘superior civilization’ then runs through the less-than-innocent humour of their text, privileging authors over translators and heaping invective on the use of English as a lingua franca. This evasion of reasoned argument enabled the solutions to be presented as self-evident within Francophone cultures, without serious debate. And then they were copied, almost everywhere.

One day in the mid 1950s, Jean-Paul Vinay and Jean Darbelnet were driving from New York to Montreal. Their account, in the introduction to their landmark Stylistique comparée du français et de l’anglais of 1958, is one of the most entertaining passages in Translation Studies: The story begins on the highway from New York to Montréal. After the crowds and crossroads of Manhattan, suddenly all is calm: the sober order of a long double carriageway […]. Yet traffic signs punctuate the route. We first read them without paying much attention, to check we are heading the right way, but then with more interest, since we are linguists going to Montreal and we cannot help but discuss language: Linguistics will out! As the signs multiply, our initial impressions are confirmed: we are indeed in the United States, in an Englishspeaking country, not because of the landscape but because of the stylistics of the traffic signs. They are very clear, of course, but no one would have written like that in French. […] (1958/72: 17; my translation, here and throughout)

16

Translation Solutions for Many Languages

The country is not the landscape; it is not even the English language; it is in the way the signs are phrased. This is clearly not an affair of grammar alone. It might have something to do with the ‘stylistics’ that the linguists put in the title of their book: The driver reads out aloud, the passenger uses the back of an envelope to note down the main texts that this ever-watchful administration lavishes upon its highway travellers: KEEP TO THE RIGHT. NO PASSING. SLOW MEN AT WORK. STOP WHEN SCHOOL BUS STOPS. THICKLY SETTLED. STAY IN SINGLE FILE. SLIPPERY WHEN WET. TRUCKS ENTERING ON THE LEFT. CATTLE CROSSING. DUAL HIGHWAY ENDS. Is one not immediately struck by the almost paternal and softly authoritarian nature of these meta-traffic injunctions? We are advised to stay in the same lane as the other cars, to stop if the school bus stops, to go more slowly because some of our contemporaries are working, to be aware that the two lanes, separated by a small strip of green, are about to join. For the French, none of this sounds at all official. It is as if we were having a polite if silent conversation with the State of New York Department of Main Roads, on small notes secretly handed out from each new thicket of red maple or spruce. A truly charming administration, so kind as to inform us, on the threshold of a promise-filled escapade: THIS SIGN LEGALLY CLOSES THIS ROAD. […]

Note here the assumption of a unitary voice behind the signs. Locating the voice is not easy, though: the conversation with the Department of Main Roads is humorous because it is as unlikely as the trees handing out the signs. Part of that voice is surely associated with the English language and culture, at least as a system that makes certain dispositions and language behaviours easier than others. However, the way the voice somehow comes from a whole language is one of the many things that stylistics has never made clear, and it is not very clear here. This is surely a voice from the authority behind the language, from an official government institution, of the kind that sets limits and guidelines as to which uses of language are officially correct. That might be the first kind of political voice that stylistics is attuned to hear: institutional limits and guidelines, even when soft and friendly. The journey continues: We soon reach the Canadian border, where the language of our forefathers is music to our ears. The Canadian highway is built on the same principles as the American one except that its signs are bilingual. After SLOW, written on the road-surface in enormous letters, comes LENTEMENT, which takes up the entire width of the highway. What an unwieldy adverb! What a pity French never made the adjective LENT into an adverb. … But, come to think of it,



Vinay and Darbelnet Hit the Road

17

is LENTEMENT really the equivalent of SLOW? We begin to have doubts, as one always does when shifting between languages, when our SLIPPERY WHEN WET reappears around a bend, followed by the French GLISSANT SI HUMIDE. Whoa!, as the Lone Ranger would say, let’s pause a while on this SOFT SHOULDER, thankfully caressed by no translation and meditate this SI, this ‘if ’, more slippery than an acre of ice. No monolingual French speaker would ever come straight out with that, nor would they spray paint over the road for the sake of a long adverb ending in -MENT. Here we reach a key point, a sort of turnpike between two languages. But of course—parbleu!—instead of LENTEMENT it should have been RALENTIR! And as for our road slippery ‘if ’ wet, they should have said, to respect the spirit of the French language …

But does it really matter? This is indeed a key example. The English adverbial ‘Slow’ corresponds to the French infinitive ‘Ralentir’, and a translation mistake rears its ugly head when the adverbial is rendered as an adverbial. Fair enough. So what are the linguists actually noting here in Canada? Are they translations, perhaps bad translations, to be compared with the pure non-translations of France? Not really: the signs are just signs, not particularly presented as translations. We don’t know who wrote them; there is no evidence of one language being the start and the other being the end of a translation process. And who, for that matter, are these linguists? Both, as it happens, were from France, not Canada: they are resolutely not from the government authority that controls these particular roads. So when they hear ‘French’, they are actually hearing voices from their ancestors across the ocean. Here are the biographies: Jean Darbelnet (1904–90) was born in Paris and studied at the Sorbonne before working as a reader at the universities of Wales, Edinburgh and Manchester. In 1938–9 he taught French at Harvard before moving to Canada in 1940, where he taught at McGill until 1946, setting up a three-year programme of night classes in translation. He later taught at Bowdoin College and Laval University. Jean-Paul Vinay (1910–99) was born in Paris and studied English at the Sorbonne before obtaining an MA in Phonetics and Philology from the University of London in 1937. He served with the French army in 1939–40 as a liaison officer with the British Expeditionary Forces. In 1946 he moved to Canada, where he taught at the University of Montreal, where he was head of the Department of Linguistics and Translation. In addition to his work on translation he directed the publication of a bilingual Canadian Dictionary in 1962 and became well known through a television course called Speaking French.

18

Translation Solutions for Many Languages

In 1967 he moved to the University of Victoria in British Columbia. He was awarded the French Legion of Honour and the Order of Canada. So the ‘language of [their] forefathers’ is not Canadian French but very much the French of the colonial centre, the metropolis, the source of ultimate authority. And their mission, driving into Canada in order to teach French, is to defend the purity of the centralized language, or something like that. So there are different voices, from different institutions, all conveying different kinds of authority. Vinay and Darbelnet are actually arguing that one kind of authority, one mode of postcolonial belonging, should win out over the others. Then again, if the purpose of the signs is to save lives, surely it would have been better to run some empirical tests, to see whether the infinitive or the adverbial is better at getting drivers to slow down? But that would not be stylistics, would it? And for that matter, are we sure that the English is always ‘Slow’? There are actually many signs out there that say ‘Drive slowly’, with the infinitive plus adverbial, and ‘Slow down’, which is grammatically the same as the French. Then in Sacramento it seems they have tried ‘Please Slow Drively’,1 which should be quite good at getting puzzled drivers to decelerate. Yes, reformulation exists; there is perhaps not just one voice available in English. We left the linguists grappling with a problem not in English, but in French: Well, what exactly? The phrase [for ‘road slippery when …’] did not come straight to mind. We are clearly dealing with a second kind of stylistics here, based not just on one language but on two at the same time. The solution to our problem would require a comparative study, a comparative stylistics. The translator had done no more than translate. And we, impregnated as we were with adjectival and conditional viscosity, hesitated to correct the translation. This is because in order to make a comparison, one needs two objects to compare and we only had one, the English text, whose manifest character dominated us entirely. In order to compare, we would have needed a French text that was equivalent (we will have to define this term), that was not influenced by a semiological process, a text spontaneously emitted by a monolingual brain, in response to a situation that is in all ways comparable.

Vinay and Darbelnet’s stylistics thereby stays clear of translations; it does not trust them: translations always risk contamination. Authority and purity are magically located in ‘a monolingual brain’, which is somewhere else, presumably back in Paris. Who has a monolingual brain? Certainly not the linguists, who have no problem assuming knowledge of at least two languages and their past



Vinay and Darbelnet Hit the Road

19

is marked by movements between at least two cultures. In fact, no one cited by Vinay and Darbelnet could be safely identified as a monolingual – nor perhaps can anyone at all, if we accept that we learn different varieties of one or several languages as we progress through life. This ‘monolingual brain’ is more like an active construction of an ideal land, virtually stored in the imaginary of the linguists, a land that has only one voice, only one authority, and that no longer exists and probably never did. It is an ideal, to which one needs an … equivalent? Did you notice ‘equivalent’ in the above passage? Vinay and Darbelnet have actually been using the word all along (as in, ‘Is Lentement really the equivalent of Slow?’). Here, though, they recognize the need for an explanation: ‘a French text that was equivalent (we will have to define this term)’. The strange thing is that they never really offer a formal definition of ‘equivalent’. In the Glossary to their 1958 book they certainly include équivalence, defined as follows: Translation procedure that accounts for the same situation as in the original, but with an entirely different expression, for example ‘The story so far’ rendered as ‘résumé des chapitres précédents’ [‘summary of the previous chapters’].

But that is the solution type called équivalence in French, translated by Chesterman (Vinay and Darbelnet 1958/89: 67) as ‘total syntagmatic change’, tentatively named ‘correspondence’ in my own work, but also understandable enough as ‘reformulation’. One might quibble about an example or two (the French rendition of ‘the story so far’ looks like a change of perspective, which could just as easily be classified as ‘modulation’), but the general territory of the beast looks plain enough. That usage is clearly quite different from the more general term ‘equivalent’, as in ‘there is no equivalent for this word’. It is this second, looser, more general sense that Vinay and Darbelnet say, correctly, that they have to define. And then they do not define it, as far as I can see. Or do they? We read on: Our hesitation was well justified. We were moving along new routes, half way between the two languages whose specific principles we presume to know. And yet we hesitate to carry out the passage of a text from one language to the other. Our doubts concern two points: a) the choice of a French text that owes nothing to the English text but covers the same reality, and b) the reasons that make us choose one translation rather than another.

Strange that they should need two things: the product of a monolingual brain, and only then the ‘reasons’ that could make a translation anything like that product. If we focus for a moment just on the first part, ‘the choice of a text’, isn’t

20

Translation Solutions for Many Languages

that where we might find the missing definition? Is an ‘equivalent’ a text in the target language that ‘owes nothing to the start text but covers the same reality’? If so, the game is played not just in the essentialist assumption of a ‘monolingual brain’ but also in the similarly precarious belief of ‘the same reality’. Who could say the realities were the same? Certainly not the ‘monolingual brain’. Only the linguists themselves could presume to locate the sameness of reality. So they come out as the ultimate authority, able to rival, in the field of language, the authority of government institutions, and able magically to muster the legacies of ancestors. Linguistics will out! All things considered, there are at least three subjectivities involved here: the infallible monolingual brain situated somewhere around Paris; the all-seeing but potentially contaminated linguists who travel between the languages; and the eminently fallible translator who produces the signs in need of correction (even though the signs are not marked as translations). To these we could add two further subjectivities: the voice fancifully attributed to the State of New York Department of Main Roads, and then a voice that presumably speaks from something like the Quebec Transportation Department, which fails to be heard because the (non-)translations are not trusted. The trouble is, if we add these two further voices as part of the ‘reality’, how could the realities ever be the same? We thus reach a third major and equally precarious assumption (alongside ‘the monolingual brain’ and ‘the same reality’): a non-linguistic reality – which is a reality about which we obviously cannot speak. Put all those assumptions together and Vinay and Darbelnet do not seem to be doing very well. They have brought across some of the terms used by Bally (‘equivalent’ is certainly one of them), but there is strangely no reference to anything like Bally’s concern with function, nor really with his attention to the purely methodological work of the linguist. Instead of visible method, here we have dismissive irony. Vinay and Darbelnet proceed undaunted, pragmatists making sense of their journey: Hence our conclusion, already quite obvious: the passage from language A to language B, to express the same reality X – the passage is usually called ‘translation’ –, is the object of a specific discipline, comparative in nature, whose goal is to explain this mechanism and to assist in the carrying out of the passage by isolating laws valid for the two languages concerned. Translation thus becomes a special case, a practical application, of comparative stylistics. (1958/72: 20)

The task of comparative stylistics would be to produce a complete set of such ‘procedures’ (procédés). Vinay and Darbelnet then propose that translators



Vinay and Darbelnet Hit the Road

21

who want to produce natural-sounding texts should employ those procedures. The secondary title of their book is thus ‘Méthode de traduction’, Method of Translation – comparative rhetoric is presented as a way of translating, even though none of the examples need to have ever been produced by translators (or so it seems from the lack of method). This translation method is glossed as ‘a set of rules that govern the miracle of a perfect translation’, which is indeed miraculous. Given the multiplicity of voices hanging around, who or what could ever pretend to recognize the perfect translation? Perhaps only the French forefathers, who speak through the travelling French linguists, protected from contamination by methodless linguistics.

Malblanc and the German connection Vinay and Darbelnet’s Stylistique comparée of 1958 was published in Montreal by Beauchemin and simultaneously in Paris by Didier, in a collection directed by the linguist Alfred Malblanc, who was quite possibly pulling a few ideological strings. So who was Alfred Malblanc? Not a lot is known. According to the records, he gained a degree in Protestant Theology in Montauban, France, and his thesis, on the anti-religion of the French proto-sociologist Guyau, was published in Geneva in 1904. So there may be a Swiss connection. Malblanc then appears from 1934 to 1941 as the author, co-author, editor, co-editor, annotator or commentator of a series of readers for German learners of French, all of them published by Teubner in Leipzig and Berlin. The series was originally called ‘Teubners neusprachliche Lektüre’ (Teubner’s Modern Language Readers); from 1937 the series title changed to ‘Teubners neusprachliche Lesestoffe im Dienste nationalpolitischer Erziehung’ (the same thing but now ‘…in the Service of National Political Education’), with titles that include Richelieu, Jeanne d’Arc and Napoleon. When the Nazis occupied Paris in 1940, Malblanc possibly felt quite at home. The publication of his Pour une stylistique comparée du français et de l’allemand in Paris in 1944 would have been just prior to the end of the Nazi occupation. Vinay and Darbelnet’s 1958 title (‘Stylistique comparée du français et de …’) looks like a public wink to this precedent in Malblanc’s 1944 work. The common source for the general approach in both texts, however, was Charles Bally. As we have seen, Bally’s Linguistique générale et linguistique française (1932) was

22

Translation Solutions for Many Languages

a comparison of French and German in everything except the title. Malblanc’s first ideological contribution was to give Bally’s approach the title it deserved, naming two languages instead of one. His second intervention was to take what Bally termed ‘stylistique externe’ and call it ‘stylistique comparée’. That said, the weight of Germanic tradition is more apparent in Malblanc than in Bally. Malblanc notes that he is taking the linguistic facts and adding ‘some psychological observations’ (1944: 113), eloquently manipulating the metaphor of the language as a person (psychology is normally for people) in a way that Bally had not. Indeed, there is a collusive crossover from the notion that a language system (langue) has a psychological profile to the actually quite different proposition that the creative use of language (parole) expresses a collective personality; the project of writing a stylistics for a whole language necessarily confounds Saussure’s two major categories. Yet there was no time to dwell on the philosophies of the enterprise: much of Malblanc’s text actually reads rather like a manual for a French student of German – in 1944 the French did not know how long the Germans were going to stay. In 1963 Malblanc published a second book, Stylistique comparée du français et de l’allemand (now without the Pour …), in the same collection of which he was director. In this text we are told that the inspiration comes more deeply from Wilhelm von Humboldt (curiously Gallicized as ‘G. de Humboldt’), notably Humboldt’s idea that languages express worldviews – and for Malblanc, the worldview makes the language a person, with a particular personality.2 This second book differs from his 1944 text in several important ways: it actually takes over from Vinay and Darbelnet the table of translation solutions, reproduces their glossary of technical terms and adds the subtitle ‘essai de représentation linguistique et étude de traduction’ (emphasis mine). Malblanc might initially have influenced and encouraged Vinay and Darbelnet, but their influence on him, and indeed their main innovation, was a shortlist of solutions for translators.

Vinay and Darbelnet’s main solution types Vinay and Darbelnet were not great theoretical linguists – they were probably having too much fun. They certainly drew on Bally, citing him profusely with respect to ‘loan’, ‘syntactic calque’, ‘transposition’, ‘explicitation’, with occasional portrayal of the specific characteristics (‘genie’) of the French language. They also refer to a thesis by Georges Panneton, defended in Montreal in 1945, on



Vinay and Darbelnet Hit the Road

23

Transposition as a ‘principle of translation’. And they explicitly say they have taken the term ‘modulation’ from Panneton (1958/72: 51), whom we shall soon meet. For the rest, they are linguists who are looking closely at languages. Perhaps because of their proximity to practical application, Vinay and Darbelnet’s work had a marked impact on the training of translators and indeed on the development of Translation Studies. Their book was revised in 1960, reprinted in 1968, 1972, 1977 and 1984, and translated into English by Juan Sager and M. J. Hamel in 1995. Their typology of translation solutions has also inspired many similar lists, as we shall see later in this book. Vinay and Darbelnet’s basic schema is represented in Table 2.1, which I have translated from the 1972 edition (Vinay and Darbelnet 1958/72: 55). Here I omit aspects like the distinctions between the ‘optional’ and ‘obligatory’ modes of Modulation and Transposition, and I replace the French term ‘équivalence’ with ‘correspondence’, since our contemporary sense of ‘equivalence’ would probably cover all the examples enlisted in the entire table. Vinay and Darbelnet also offer a list of supplementary ‘prosodic effects’, to which we will come in due course. The visual form of the table is meaningful. At the centre stands Literal Translation: if translators cannot find a solution on that level, then they should move either upwards (incorporating elements from the start text) or downwards (drawing on elements in the target language and/or culture). The upward movement is into the territory of ‘direct translation’ and apparently involves less difficulty (I would prefer to talk about less risk); the downward movement involves ‘oblique translation’ (a visual metaphor whose virtues I have never really appreciated3) and is said to be more difficult (or more risk-laden). The safety zone is undoubtedly in the middle. The seven solution types each come with examples on three levels of usage: lexis, collocation (agencement) and message. Malblanc (1963) reproduces exactly the same table but gives examples comparing French and German. The description of these ‘procedures’ takes just nine pages (1958/72: 46–54), which I summarize: Loan (emprunt): Foreign words can be used in the translation in order to fill in a lexical gap or to give local colour. All languages are full of old loans from other languages; the ones that are of interest here are the new loans, which still look and sound foreign. Calque: Some loans comprise literal translations of each of the elements in the foreign phrase. Again, the calques of interest are the ones that are still fresh. The

24

Table 2.1  Vinay and Darbelnet’s general table of translation solutions Collocation

Message

1. Loan

Fr. Bulldozer Eng. Fuselage

Fr. science-fiction Eng. à la mode

Fr. five o’Clock Tea Eng. Bon voyage

2. Calque

Fr. économiquement faible Eng. Normal School

Fr. Lutétia Palace Eng. Governor General

Fr. Compliments de la Saison Eng. Take it or leave it.

3. Literal Translation

Fr. Encre Eng. Ink

Fr. L’encre est sur la table Eng. The ink is on the table

Fr. Quelle heure est-il? Eng. What time is it?

4. Transposition

Fr. Expéditeur: Eng. From:

Fr. Depuis la revalorisation du bois Eng. As timber becomes more valuable

Fr. Défense de fumer Eng. No smoking

5. Modulation

Fr. peu profond Eng. Shallow

Fr. Donnez un peu de votre sang Eng. Give a pint of your blood

Fr. Complet Eng. No Vacancies

6. Correspondence (équivalence)

Fr. (milit.) la soupe Eng. (milit.) tea

Fr. Comme un chien dans un jeu de quilles Eng. Like a bull in a china shop

Fr. château de cartes Eng. hollow triumph

7. Adaptation

Fr. Cyclisme Br.Eng. cricket Am.Eng. baseball

Fr. en un clin d’œil Fr. Bon appetit! Eng. before you could say Jack Robinson Am.Eng. Hi!

Source: Vinay and Darbelnet 1958/72: 55; my translation.

Translation Solutions for Many Languages

Lexis



Vinay and Darbelnet Hit the Road

25

example ‘économiquement faible’ (economically weak) is apparently a calque from German (1958/72: 48) – as a financial term it translates as ‘distressed’, apparently. Vinay and Darbelnet give some examples of calques they consider atrocious (‘pénibles’) and even ‘abominations of desolations’ (to risk a millennial Biblical calque that we will soon pick up on): ‘thérapie occupationnelle’, ‘les quatre Grands’, ‘Banque pour le Commerce et le Développement’, etc. Of course, almost all of these calques are now well and truly part of the French language – the linguists were fighting a losing battle even as they spoke. And the one notable exception in this list is ‘les quatre Grands’ (the four great powers: the United States, the USSR, the United Kingdom, France), which might have presumed to rule the world in 1958.4 The linguists’ hopes and assumptions were a product of the times. Literal Translation: The translator works word-for-word and only has to respect the obligations of grammar (‘les servitudes linguistiques’). This solution is only viable with any frequency between languages that are linguistically and culturally cognate. An interesting point here, and one that will become of more interest later, is that Vinay and Darbelnet have no category for changes that only concern the order of words and phrases. For example: ST: The black ink is on the table. TT: L’encre noire est sur la table. Here the noun and adjective have changed order, so is this still literal translation? One would hope so, since the switch can be classified as a fairly trivial act of obeisance to grammar. Word order is certainly dealt with elsewhere in Vinay and Darbelnet (1958/72: 201ff.), but its absence as a category is nevertheless intriguing. These first three solution types are apparently the ones handled by machine translation. If they result in something unacceptable, then further stylistic procedures are necessary. And those further procedures assume, not similarity of linguistic form, but an ‘identity of situation’ (1958/72: 49). Transposition: ‘One part of speech is replaced by another, without changing the meaning of the message’ (1958/72: 50). This can happen both within a language and between languages: ST: Après qu’il sera revenu ST: Après son retour ST: After he comes back ST: After his return

26

Translation Solutions for Many Languages

In these examples we have transpositions between the verb (sera revenu, comes back) and the noun phrase (retour, return). It is easy to see Vinay and Darbelnet extending the kinds of lists compiled by Bally, turning the intralingual procedure into an interlingual one. There is a strange note in the text here: ‘The original and the transposed expressions are not necessarily equivalent from the point of view of stylistics’ (1958/72: 50), basically because the transposed one is usually more literary. This implies that the ‘meaning of the message’ is the same, but the ‘stylistic value’ is not. Many questions could and should be asked about the extent to which these terms ‘message’, ‘meaning’ and ‘equivalents’ are separable. But Vinay and Darbelnet do not ask those questions. A particular kind of Transposition is called French chassé-croisé (‘interchange’ in English, which conjures up images of an intersection of four-lane highways, with fancy flyovers). This is where two semantic elements change grammatical category and change places, all in the one go (1958/72: 105–6): ST: Blown away ST: Emporté par le vent [taken away by the wind] So the value of ‘away’ is expressed as a verb in French and thus adopts the first position in the French phrase. It is under chassé-croisé that we find one of Vinay and Darbelnet’s more famous examples: ST: He swam across the river. ST: Il traversa la rivière à la nage. [He crossed the river by swimming.] (1958/72: 58) This sparks off a characterization of the two languages in play: ‘the English sentence is organized around the word as image, whereas the French uses the word as abstract sign’ (1958/72: 58). And the abstractions climb even higher: ‘In general, French words are mostly situated at a higher level of abstraction than are the corresponding English words’ (1958/72: 59). A lot of philosophy from just a few examples! Modulation: ‘Variation in the message, obtained by a change in the perspective, in the lighting’ (1958/72: 51), considered necessary when the above procedures give a translation that is correct but ‘goes against the génie of target language’. Some modulations are obligatory, as in:



Vinay and Darbelnet Hit the Road

27

ST: The time when ST: Le moment où [The time where] (1958/72: 51) Others are more optional: ST: It is not difficult to show … ST: Il est facile de démontrer … [It is easy to show …] (1958/72: 51) Correspondence (équivalence): ‘Two texts can account for the same situation using completely different stylistic and structural means’ (1958/1972: 52). When you are hammering a nail and you hit your fingers, in English you should scream ‘Ouch!’, and in French ‘Aïe!’, apparently (personally I shout a lot of rather more unprintable things, in both languages). Most of the examples are fixed expressions, idioms, proverbs or whatever you want to call them, of the kind: ST: Too many cooks spoil the broth. ST: Deux patrons font chavirer la barque. [Two bosses make the boat capsize.] The alignment of these sayings is intimately pleasing, creating the harmonious image of two completely different, unrelated and yet equal ways of living a life. Vinay and Darbelnet make little mention of the many idiomatic expressions that find no passively awaiting ‘equivalents’ of this kind, or the many that move from language to language through literal renditions. And yet, when you go looking for an idiom and you fail to find one, that sense of a linguistic gap, that intimate frustration, says something about the psychological validity of this category. Adaptation: When the situation itself does not exist in the target culture, another situation can be used as an equivalent (so we actually have an ‘équivalence de situations’). The one example here is strange. When a father returns from a long trip away, in French, he apparently kisses his daughter on the mouth (not any more, I sense, but okay, on the cheek, perhaps both cheeks?), whereas a British or American father might simply give his daughter a warm hug. And when you’re a father who speaks English to his French daughter, as is my case, Adaptation is never going to be an easy way out. (I have no idea what I do.) In addition to these general solutions, Vinay and Darbelnet outline a series of further but apparently minor procedures, mysteriously called ‘prosodic effects’. Most of these operate at sentence level, largely to comply with the constraints of

28

Translation Solutions for Many Languages

the target language. The main effects can be summarized in terms of three pairs (I take this summary from Pym 2010/14): Amplification/Reduction: Amplification is when the translation uses more words than the start text to express the same idea (1958/72: 5): ST: the charge against him (four words) ST: l’accusation portée contre lui (five words) [the charge brought against him] (1958/72: 5, 183) One variant of this is étoffement, where a word grammatically requires collocation of another word. In fact, the above example is cited in two places: once as an example of amplification, then again to illustrate étoffement (the categories and sub-categories are not easy to follow). Reduction (économie) is the opposite (just read the above example in the opposite direction). The counting of words here seems rather trivial: as in the case of word-order change in literal translation, it might just be a matter of obeying grammatical rules. Worse, the example here, the explicit mention in French of the verb ‘brought’ that is implicit in English, looks a lot like simple Explicitation (see below). Note that Vinay and Darbelnet avoid the terms ‘addition’ or ‘deletion’, since there was doubt as to whether such things still counted as translation. So in principle, everything that is amplified could be explicitation, and everything that is reduced could be implicitation. Explicitation/Implicitation: Explicitation is when the translation gives specifications that are only implicit in the start text (1958/72: 9): ST: students of St. Mary’s TT: étudiantes de l’école St. Mary [female students at St. Mary’s school] (1958/72: 117) Here the French specifies that the students are women and St. Mary’s is a school. Implicitation is the opposite (the directionality of the above example could be reversed, if it is common knowledge in the target culture that St. Mary’s is a school for girls). This category will become particularly important when we move to Asian languages, which are generally more implicit than English, to degrees that go well beyond what Vinay and Darbelnet were contemplating. Generalization/Particularization: Generalization is when a specific (or concrete) term is translated as a more general (or abstract) one (1958/72: 9). Example: ‘mutton’ (sheep meat) becomes ‘mouton’ in French (both the animal and the



Vinay and Darbelnet Hit the Road

29

meat), or the American ‘alien’ becomes ‘étranger’ (which includes the concepts of both ‘foreigner’ and ‘alien’). A ‘window’ in English can be translated in French as ‘guichet’ (when you buy a train ticket, for example), ‘fenêtre’ (when you want some fresh air) or ‘devanture’ (in the case of a shop window) (1958/72: 64). Particularization is then a shift in the opposite direction. Compensation: A value created on one level or textual feature can be re-created or referred to on a different level of textual feature: ST: On se tutoie … [We can use the intimate second-person pronoun …] TT: My friends call me Bill. (1958/72: 190) The French intimate second person these days has no corresponding feature in English (since ‘thou’ has become archaic, with religious overtones), so its value is compensated for by use of the person’s shortened given name. Fair enough.

Is there rhyme or reason in the list? Vinay and Darbelnet divide their book into three large sections for their three levels of discourse: lexis, collocation and ‘message’. This allows the above solution types to be explained and exemplified in different ways for each level. Unfortunately the typology of ‘prosodic effects’ thereby becomes particularly difficult to keep track of. The list has no theoretical reason to be closed (there could be more categories, there could be fewer) and there are no strong reasons why some solutions are relegated to the B team (‘prosodic effects’). In all, the inundation of examples and quick leaps to abstraction cover over a lack of careful planning. When research is conducted bottom-up, from texts to categories, numerous other solution types appear. Indeed, quite a few others are named in the course of Vinay and Darbelnet’s various discussions. The question is then whether we need a ‘miscellaneous’ category, or whether the more occasional solutions can be construed as sub-types of the more general types.

Translation preferences and culture as distinction Instead of reasons, Vinay and Darbelnet distribute values: good words for everything that is ‘oblique’ translation, bad words for everything that runs the risk of

30

Translation Solutions for Many Languages

interference from English, more or less. Since this binary ideology runs counter to the very reasons why the solution types are of interest to me, it is as well to dispense with it here. I start from the values expressed in the nine pages where Vinay and Darbelnet first describe their solution types. In their initial editions, Vinay and Darbelnet posit that innocuous calques like ‘les quatre Grands’ (‘the four great powers’) are ‘in the minds of some translators, the most concrete expression of the abomination of desolation’ (1958/72: 48). What the hell is this? If you know your Bible (Daniel 9.27, Matthew 24.15, Mark 13.14 and the apocryphal 1 Maccabees 1.54), the allusion should probably work as a very weak joke, since it is itself a calque from Biblical Hebrew and Greek, I suppose. This kind of discursive sophistication is scarcely innocent. It separates an inner readership of intimes from the rest, creating smug satisfaction if you are on the right side and perplexity if not. This discursive division between civilized and less-civilized then becomes the strategy underlying the distribution of all further values. When writing in defence of Correspondence, Vinay and Darbelnet heap gratuitous invective on mediators who might inadvertently transfer a structure from one language to another. We are told that certain ‘so-called bilinguals suffer from permanent contact with the two languages and finish up knowing neither of them’ (1958/72: 52). This is no doubt a light-hearted quip, like the Biblical allusion, yet it is lapidary in more than one sense. There are indeed people who are left stranded between languages, with no formal education in either and who are seriously disempowered as a result. In my native Australia there are cases where the transmission of indigenous languages has been broken by force and the schooling in English is inadequate – people are effectively stranded between two languages, left without linguistic agency. In Catalonia, where I live, many adult speakers of Catalan had no formal education in that language (thanks to the centralist language policy under Franco) and are thus sometimes in serious doubt as to what is official Catalan and what is official Spanish, which is another mode of being stranded and disempowered on a long-term basis. The problem in such cases, however, is not the fact of bilingualism but the forceful exclusion of a language from school systems. If you look at things quantitatively, most of the world’s language users live in constant contact with more than one language, and many of them enjoy and play with the wealth of language resources thus made available: code-switching and code-mixing can be relished without any loss of communicative value. Monolingual speakers, especially of English, are actually the great exception here. To suggest that constant contact with two languages necessarily leads to degradation is a very cheap shot, possibly



Vinay and Darbelnet Hit the Road

31

aimed at the many Quebecois who live bilingual lives. And the implication, of course, is that linguistic decisions should be made by those who maintain a significant degree of monolingualism, or who are magically able to keep the languages separate in their brains, in a way that others somehow cannot. Vinay and Darbelnet’s readers are invited to eschew one condition and to aspire to the other: the Quebeckers should really try to speak like the French, and implicitly like the expert linguists who have come from France. This is how the play of opposed voices becomes a distribution of cultural values. Further clues about this postcolonial distribution of values are given in the very next sentence, where we learn that a ‘perfectly organized language’ should not accept interference by translators: […] only authors can allow themselves such escapades, since the success or failure of the calques will then reflect on them alone. In a translation, you have to stick to the classic forms. (1958/72: 52)

This implies, of course, that the translator is not an author, even though Vinay and Darbelnet are elsewhere demanding that translators engage in degrees of Transposition and Modulation that require significant amounts of creative linguistic work. Since the argument is about responsibility, it could mean that the translator is not a ‘principal’, in something like Goffman’s sense of the person who takes responsibility for the utterance. Goffman, however, was talking about responsibility for what is said;5 here, in translation, it is a question of responsibility for the way things are said. Vinay and Darbelnet give no actual reason why translators should not bear that responsibility, why they should not be named as such and why they should not be treated as authors in precisely that respect. In place of argument, they present no more than authority – no doubt the authority that they themselves assume as authors, from France.
 When Vinay and Darbelnet justify Adaptation, they indulge in further invective about the dire consequences for the world if these ‘oblique’ translation strategies are not used properly: ‘one fears that four-fifths of the globe will live exclusively on bad translations and intellectually starve on slop for cats’ (1958/72: 54).6 What? Yes, that’s what they say, more or less: ‘bouillie pour les chats’, perhaps ‘boiled left-overs’, but certainly for cats. Sager and Hamel render this as ‘a diet of linguistic pap’ (1958/95: 40), which is not bad – ‘pap’ is a lovely word to use, sparingly, although it does seem to soften the blow and rather loses the sense of sub-human degradation coming from mixes. Chesterman (translating Vinay and Darbelnet 1958/1989: 69) goes for ‘this diet of pulped catfood’, which does retain the mix and the lower form of life, but I thought

32

Translation Solutions for Many Languages

pulping was what happened to printed pap (finding this kind of translation solution is indeed difficult and creative!). The message is nevertheless clear: If you, poor translator, risk allowing English to interfere with good French, the result will be a global stew and your readers will not be proper humans. The degraded literalism apparently tolerated by international organizations is thereby summarily dismissed by Vinay and Darbelnet as ‘gibberish that has no name in any language but which René Étiemble rightly calls a “trans-Atlantic pidgin”’ (1958/72: 53).7 Literalist translations, it seems, will reduce us to a world of cats, gibberish and non-language. This use of insult instead of argument was once widespread in Francophone discussions of translation, and Vinay and Darbelnet were by no means the most extreme in this regard. As noted, Georges Panneton, whom Vinay and Darbelnet lean on for their terms ‘transposition’ and ‘modulation’, presented his thesis in the heady days of October 1945 – the end of the Second World War, with France and Canada on the winning side, might have justified a certain euphoria about the cultural tasks to come. Panneton opposes Transposition (of which he names Modulation as a variant) to ‘literal or timid translation’ in such a way that Transposition alone can give the text ‘freedom to develop an unlimited power of interpretation’ (1945: i). Panneton is steadfastly opposed to ‘infiltration by Anglicisms’ (ibid.) and assures us that ‘no other language has shone on a par with French’ (1945: ii). The overarching ideology is one of progress, development and attainment of a cultural pinnacle from which French and English can presume to speak for the world: Our sketch of a general history of translation traces the gradual rise of the art of transposition, running parallel to literary progress, with the global movement of translation reaching its highest point in the ethnic development and evolution of two prototypical adult neo-European languages, English and French, whose double drive is undoubtedly primordial on the American continent. (1945: ii)

I hesitate to cite much more – but there is plenty in the same vein.8 World linguistic evolution, it seems, has been led by Indo-European, the two adult strands of which meet at its highest of high points: yes, of course, a bilingual Canada: ‘The prestigious sign of bilingualism, evocative of the highest civilizations […] bestows on us a unique distinction in the hierarchy of peoples. Need it be proved?’ (1945: 7). Well, um …Vinay and Darbelnet, we have seen, were far more circumspect on the topic of bilingualism. Although they certainly shared Panneton’s translation preferences, they toned down the evolutionary glory, at least. But neither they nor Panneton believed that such assertions required any evidence.



Vinay and Darbelnet Hit the Road

33

Things similar can also be traced in the wake of Vinay and Darbelnet. Here I return to one of their best-known examples (if you google the French, more than half the hits refer to the sentence as a translation example): ST: He swam across the river. TT: Il traversa la rivière à la nage. [He crossed the river by swimming.] (1958/72: 58) Dominique Aury (pen name of Anne Cécile Desclos), in her preface to Mounin (1963: ix), ironically rejoices in the way Vinay and Darbelnet intervene to help the French from rendering the English word-for-word as ‘Il nagea à travers la rivière’. That literalism would mean translating the English into ‘Patagonian’ (‘traduire en patagon’), she says. Into what? One supposes the reference is to an imaginary language that is a long way from civilization and is spoken by indigenous non-Indo-Europeans, or cats, in any case by people who are not like ‘us’. And this is amusing, unless you are from Patagonia, where you might speak Spanish, or perhaps Welsh, while pre-colonial languages like Tehuelche have died. But why should that matter to the pinnacles of civilization? In a fine but largely forgotten polemical paper, Jean-Louis Laugier (1973) picks up the same example, notes the implicit racism with regard to Patagonians and asks why, after all, the French translation should not be the word-for-word ‘Il nagea à travers la rivière’: ‘We should at least recognize this marvelous possibility of reading English in French’ (1973: 31). Further, if the purpose of literary translation is to introduce the reader to new and exciting forms, tantalizingly alien at times, then why not the word-for-word version? Laugier agrees that the reader might not learn much English from the exercise, ‘but we at least show something and we let in a little fresh air [nous nous aérons un peu]’ (ibid.).9 So what actual arguments could be used against Laugier’s fresh air? Barbara Folkart (1991), perhaps alone, has taken Laugier seriously, perhaps too seriously. She gratuitously theorizes his proposed river-swimming rendition as ‘material translation’, ‘ethnolinguistic translation’ and ‘exoticism’, understood as ‘this grain that constitutes the presence of the medium in the message’ (1991: 300) – you’re doing very well when you get retro allusions to Barthes and McLuhan all in the one sentence! Folkart criticizes Laugier’s proposal because it is only one sentence, bereft of context. Yes, there is a serious argument here: if you put Laugier’s literalism in an actual context, using close literalism for a whole literary work, the effect might become unreadable, or overwhelming, or philological, and it ultimately offers the reader the quite obvious piece of information that the start text was in English: ‘instead of true access to the Other,

34

Translation Solutions for Many Languages

Laugier offers the reader no more than junk tourism [tourisme de pacotille]: America in a fortnight, English on cassettes, to listen to while you sleep’ (Folkart 1991: 303). In a later study, Folkart further ironizes Laugier’s literalist rivercrossing as a ‘lofty ideal’ (2007: 6). Once again, an attempted argument dissolves into little more than facile derision.10 Why use argument or evidence if you can just pooh-pooh? Henri Meschonnic (1999: 110) also briefly mentions the literalist river, dismissing the banality of grammatical correctness: ‘[…] a problem of the language system [la langue]. For “comparative stylistics”. Of the kind he swam across the river, “il traversa la rivière à la nage”’. Meschonnic uses the reference to make a good point about the intelligent use of orality in translations for the theatre, yet he has nothing but disdain for the low-level correspondences that operate on the level of grammatical obligation. That whole argument was a long way beneath him. As Bourdieu (1979/84) remarked of (French) cultural practices in general, the chief import of all these non-arguments is to set up and manipulate a distinction between the different ‘tastes’ of social classes, mostly between a superior ‘us’ and an inferior ‘them’. Those cultural distinctions become selfjustifying, functioning so as to legitimize social distinctions, over and above the distributions of wealth or political authority. Something like that is happening here, in the various appeals to self-evidence. Could there ever be a definitive argument in favour of one kind of solution or another? Probably not, or at least not in terms of the values discussed here. Elsewhere (Pym 2000) I have described the history of translation principles as a constant negotiation process in which there are moments of relatively stable hierarchies (‘regimes’) but no lasting victories: a principle that is almost forgotten in one age, or culture might become key in another. The reluctance to engage in debate, however, the simple assumption of self-evident superiority, risks having precisely the opposite effect: once others begin to doubt your axiomatic superiority, they will have no substantial reason to mimick your translation preferences. As the prestige and protagonism of French culture has declined on the international stage, many of those who have cited and adapted Vinay and Darbelnet have blissfully overlooked the strong ideologies that framed the initial formulation of their solution types. The solution types proved useful, but the ideology had no strong reason to travel. The comparative study of solution types should initially be interested in the possibility that each typology is determined not just by the languages it draws examples from, but more profoundly by the culturally specific political ideologies



Vinay and Darbelnet Hit the Road

35

about those languages, especially about their relative prestige and relations of power. In this particular case, the ideological component would seem to be particularly marked; it raises the question of whether the apparently technical categories really can be separated, not just from specific languages, but also from historical prejudices. Those are questions that can only be answered by tracing the adventures of the Vinay and Darbelnet typology as it travelled out into the world.

Arrival at the point of departure Did Vinay and Darbelnet know where they were going? Were they aware of the course they were steering, between the gratuitous insulting of bilinguals and the unflinching appeal to a colonial past? Their prose does not seem too worried; it is urbane and playful. Their drive of discovery, indeed their psychic drive to discovery, moved by the desire to disrobe true translation, began in New York and headed toward Montreal but then magically crossed a big river à la nage and figuratively finished elsewhere, in the language and the colonial centre that were of most apparent value: The car rumbles along at its regular rhythm. And now our eyes are ravished by a parade of desired translations: DOUBLER À GAUCHE, PRIORITÉ À DROITE […]. We are already through the Saint Cloud underpass and there is the Seine, the Bois de Boulogne, Paris. (1957/74: 22)

And it was indeed from publication in Paris that their influence would mostly begin.

3

A Tradition in Russian and Environs

Chapter summary: Twentieth-century Russian translation theory produced a set of ideas that mark it as a separate approach to translation solutions. First, translation was seen as conforming to just one criterion (‘adequacy’) rather than operating between two poles (such as ‘foreignizing’ vs. ‘domesticating’). Second, translators were seen as using a wide range of solutions in order to achieve adequacy. And third, translation solutions depended on specific text types and purposes. The development of these ideas in the 1950s was entangled in Stalin’s pronouncement on linguistics in 1950 (encouraging Fedorov and others to adopt a linguistic discourse) and then by Stalin’s death in 1953 (after which the linguistic line was openly challenged). In the ensuing struggle between linguistic and literary approaches we find a 1958 proposal for a unified Translation Studies. Because of Cold War politics, however, very little of this development reached Western languages, and some of the Russian metalanguage is so different from that of Vinay and Darbelnet that real misunderstandings can result.

So Bally leads to Vinay and Darbelnet (1958) and thereby to Malblanc (1963). Where do the influences go from there? Any account of the next steps must actually back up quite a way, to works that predated Vinay and Darbelnet and owed apparently little to Bally. A semi-cryptic clue about this can be found in Peter Fawcett’s Translation and Language (1997). Fawcett only presents the Vinay and Darbelnet typology after first summarizing a rival typology, formulated by the Russian theorist Yakov Retsker (1974/2007), ostensibly as ‘reorganized’ by the also Russian Aleksandr Shveytser (1973/1987: 25–8). Fawcett, unfortunately, does not spell out how someone writing in 1973 could have ‘reorganized’ categories published in 1974, and he does not really explain why the work of those scholars, in the 1970s, should be positioned prior to Vinay and Darbelnet (1958). As Fawcett’s editor at the time, I should have delved into those mysteries and I sincerely

38

Translation Solutions for Many Languages

regret not having done so – things Russian were simply too hard in those days, and too difficult to read. Had I scraped a little, I might have encouraged Peter to explain why most references to this period are actually not to Retsker or Shveytser but to Andrey Fedorov’s Введение в теорию перевода (Introduction to the Theory of Translation), published in 1953. Now, that was indeed prior to Vinay and Darbelnet (1958)! Fedorov’s work has more recently been hailed as perhaps the first systematically linguistic approach to translation (in Mossop 2013) and was influential in Russian and beyond, as we shall see when we get to China. Fedorov, however, did not present any full list of solution types, which only compounds the mysteries. History thus imposes a brief dive into the Soviet Union, into a tradition of theory that all Russian departments love telling me, undoubtedly correctly, that I fail to understand. I enter undaunted but not unaccompanied.1

Three ideas to situate translation solutions Although there is no neat typology of translation solutions in Fedorov’s work in the 1950s, there is at least a sketch of one in his very early work, notably in his 1927 article ‘Проблема стихотворного перевода’, translated into English as ‘The Problem of Verse Translation’ (Fedorov 1927/74). In that piece, published when he was 21 and actually dated two years prior, Fedorov starts from the proposition that translations should not be of forms but of the historical functions of forms, where the sense of ‘function’, drawn explicitly from Tynyanov, is the position an element has in a dynamic system. Fedorov’s approach at this stage also includes a surprisingly radical hermeneutic component in that the value of the form is considered to be unknowable independently of the translation: Translation is not the reproduction of a work but the creation of something new – according to a model which gives rise to varying interpretation, a model which is not uniform but multifaceted. It is impossible to equate the two systems – translation and original – and not only because they are essentially unequal but also because we do not know the most equatable quantities. The original is not a stable, previously defined quantity; a translation is not one of the aspects of the original but, as it were, an aspect of an aspect of that multifaceted system; it is a particular one-time interpretation of the system of the original, and we have perhaps no criterion for assessing its correctness or incorrectness. (1927/74: 24)



A Tradition in Russian and Environs

39

One wonders if the young Russian Germanist had read Walter Benjamin’s 1923 essay ‘Die Aufgabe des Übersetzers’ (The task of the translator), at least with respect to the fleeting, tangential status of translations, extending part of the start text. Whatever the case, this radical positioning leads Fedorov to exclude the possibility of ‘equality’ (равенство) and to instead focus on what we might translate as the ‘deviations’, ‘infringements’ or ‘violations’ (нарушения) that translators enact in order to attain ‘accuracy’ (точность) in the rendition of verse. That position raises the question of what is being violated or deviated from (surely some level of equality operates as a conceptual horizon?), but that rather obvious point somehow goes unmade. The ‘deviations’ or ‘violations’ look like what I have been calling solution types, of which six are offered, mostly without convenient labels. I risk a rough reduction (the labels are mine): Omission: ‘Every translation offers many examples of the nonrendering of a number of words’ (1927/74: 15). Addition: ‘introduction by the translator of an element (or group of elements) not given in the original, which does not accompany the loss of some element’ (1927/74: 15). This sounds a little like explicitation, which was a concept perhaps not available at the time. Substitution (замена): the dropping of an element and its replacement by another element not present in the start text. Constructive-Semantic Violation (нарушение): meanings are used differently in the start and the target texts, as when poetic form requires semantic changes. Correspondence: translation on the level of the relationship between lexical levels in different languages (1927/74: 17), related to degrees of required archaization or foreignization. This comes with a nice example illustrating the values of whole languages: when Heine in German uses the single French word in ‘Madame, ich liebe Sie’, this becomes ‘Madame, je vous adore’ in Russian, since the foreignizing effect in Russian – where French was more at home – required more French words (1927/74: 19). Changes in Element Order: Changes in the order of elements at the sentence or text levels (all described as ‘syntactics’) in order to retain rhetorical effects. The young Fedorov explicitly takes the idea from Vinogradov’s use of the term ‘composition’ in stylistics, but he prefers ‘syntactics’ because it is more specific to features of element order (1927/74: 22n.). There you have it: a typology of solution types, more or less, in 1927. True, the descriptions are best suited to problems in the translation of verse, where

40

Translation Solutions for Many Languages

the priorities of form may require radical alteration of content and element order. What is more interesting, perhaps, is that the nature of Fedorov’s overall ‘accuracy’ depends in part on the nature of the text being translated: The accuracy of a scholarly prose translation is accuracy from one particular point of view – the point of view of meaning; the verbal form is immaterial, only it must not contradict the norms of the language into which it is being translated, i.e., not interfere with the perception of the sense; its function is that of an auxiliary, incidental, concomitant element. [On the other hand,] absolute accuracy in a literary translation – particularly a verse translation – would be accuracy in all aspects, the recreation of the interaction of all the elements of the original … (Fedorov 1927/74: 24)

So we have three basic ideas here: the notion that translation has just one goal (here called ‘accuracy’), a typology of the changes made to meet that goal and a certain dependence on text types (admittedly restricted to scholarly vs. literary texts). These ideas can certainly be found in texts prior to 1927 and indeed in the general debates of the late 1920s (as we shall soon see), so there is no pressing need to attribute originality to Fedorov. I nevertheless submit that these ideas, when presented together, were not just passably novel in the 1920s, they were also remarkable for their eventual fecundity in the Soviet tradition, for their future use by others beyond that tradition and for their absence from most bibliographies. What particularly interests me here is the way those ideas were contextualized in the 1920s and early 1930s, and then again in the 1950s.

First idea: Translation is just one thing Fedorov’s 1927 article was responded to in the same year by the Ukrainian translation critic Volodimir Derzhavin, whose text generally argues for a more foreignizing approach than one might gather from Fedorov. Derzhavin uses ‘adequacy’ (адекватність), rather than ‘accuracy’, as the aim of all translations, and this notion of ‘adequacy’ would seem to have been in circulation at least since Batyushkov (1920).2 Whatever the case, it is ‘adequacy’ that was to predominate from then on. I will return to Derzhavin soon. If we move forward just slightly to 1934, we find the same term ‘adequate’ (адекватный) being used in an encyclopaedia article signed by A. A. Smirnov:3 An adequate translation conveys the author’s entire intention (both thoughtout and unconscious) as [realized by] a certain ideological and emotional



A Tradition in Russian and Environs

41

literary impact on the reader, to match as far as possible – through exact equivalents [точные эквиваленты] or satisfactory substitution [субституты (подстановки)] – all the resources used by the author with respect to imagery, color, rhythm, etc., these resources being considered not as ends in themselves but only as a means to achieve the overall effect. (Smirnov 1934; italics mine)4

The intriguing idea here is that an ‘adequate’ translation combines both exact matches (this is the sense of ‘equivalents’ in the above passage) and ‘substitutions’ (which would be most of the things in our lists of translation solutions, including Fedorov’s early sketch). Fedorov (1927/74) had effectively denied equality; Derzhavin (1927) had not mentioned it and is in any case not cited here; so I am not really sure where Smirnov’s model came from. I take the author of the encyclopaedia article to be Aleksandr Aleksandrovich Smirnov (1883–1962), a medievalist and professor at Leningrad State University and a translator and editor at the Academia publishing house from 1924 (Bedson and Schulz 2015: 122–30), where numerous classics of world literature were published in Russian translation. The entry includes a healthy bibliography, with references to Fedorov (1927), Chukovskiy and Fedorov (1930), some other Russian (not Ukrainian) works from the twentieth century5 and nineteenth-century theories in German. The important point, at least for the thought about solution types, is the combination of ‘equivalence’ and ‘substitution’ as two complementary items that together make up ‘adequacy’.

Second idea: Translators use a range of solutions The second idea would be that the kind of translation solutions listed in Fedorov (1927/74) can be generated in some controlled way and made applicable to more than the translation of verse. This effectively means working from the kind of divisions used in Smirnov’s encyclopaedia article. A logical development along these lines is formulated in an article by Yakov Retsker in the 1950 collective volume Вопросы теории и методики учебного перевода (Theoretical and methodological questions of translation teaching). Retsker proposes that there are just three basic kinds of translational relationship, mainly on the level of terms and phrases: (1) equivalence, (2) ‘analogue’ correspondence, as in the case of synonyms, and (3) ‘adequate substitution’ (адекватная замена). Smirnov’s two branches of ‘adequacy’ (‘equivalence’ and ‘substitution’) thus become three in Retsker. In this small system, ‘equivalence’ clearly means what Retsker would later explain as a one-to-one ‘permanent equivalent correspondence, which for

42

Translation Solutions for Many Languages

a certain time and place no longer depends on the context’ (Retsker 1974/2007: 157), whereas ‘analogue translation’ is where the translator decides between similar options on the basis of each particular situation. The third category, ‘adequate substitution’, is then where Retsker brings in a rather heterogeneous list of options: 1. Concretization of undifferentiated and abstract concepts 2. Logical development of concepts 3. Translation through antonym 4. Compensation Strangely enough, these four items, formulated in 1950, are precisely the ones presented in Fawcett (1997) – which might explain how they could be found in Shveytser (1973). One small mystery solved! Retsker (1974/2007) did nevertheless develop these four into a list of seven ‘transformations’, a nice number for our purposes: Differentiation, Specification, Generalization of Values, Semantic Development, Antonymic Transfer, Holistic Transformation and Compensation, all of which can be mixed and matched, as in Vinay and Darbelnet. Those seven types at least looked like they were on the same level. They became influential on Russian translation theory, although the typologies that are mainly cited today are those of Shveytser (1973/87) and Barkhudarov (also transcribed as ‘Barhudarov’) (1975), whom we shall meet toward the end of this chapter. In 1950, the innovative idea was not so much the embryonic list of solution types but the three-term framework (in French one would say the dispositif, the ‘mechanism’) that neatly related them to each other. In practical terms, Retsker was saying, okay, you use the established equivalents, you pump the dictionary for what it is worth, and if all that fails, you start on the really interesting translation solutions, of which here are four, or seven. This posits that there are three basic ways to translate (although it gives no strong reason why there should not be more or less than three), plus some tricks of the trade. In theoretical terms, the novelty lies in the way these three terms offer a linguistic description of the relations between the two sides of a translation. The focus is not on comparing two languages in the blind hope that the differences will somehow be of interest and use to translators. As Retsker put it in the title of his 1950 article, ‘О закономерных соответствиях при переводе на родной язык’ (On regular correspondences in translation into the vernacular), the object of study is the ‘correspondences’ actually created in translation. Of course, there remains considerable doubt about how those correspondences are to be discovered:



A Tradition in Russian and Environs

43

Retsker’s title uses ‘закономерныe’ to describe the correspondences, a term that can be rendered as ‘regular’, perhaps in the sense of rule-governed. Then again, Retsker begins this same 1950 article talking about the need to study a group of ‘good-quality translations’, so he could also be referring to statistical regularity.6 Whatever the case, the three terms of what we might call a ‘typology of correspondences’ formed a space in which it became possible to do a linguistics of translation, rather than just apply linguistics to translation. We will see this notion of ‘correspondences’ appear later in German.

Third idea: Translation strategies depend on text types and purposes The third idea, found in nuce in Fedorov (1927/74) but with antecedents as far back as Jerome (395/1962) and then Schleiermacher (1813/1963), starts from the notion that different ways of translating correspond to different text types. The formalization of this now unsurprising point is sometimes attributed to the German theorist Katharina Reiss (1971/2000), who initially distinguished between three main text types (expressive, conative and referential), based on the three linguistic persons in a manner formulated by the German psychologist Karl Bühler (1934/82). It would be convenient to imagine that Bühler’s triad had an influence on Soviet thought and then on Reiss, but I suspect not: his functionalism was more influential on the development of Prague School linguistics (see, for example, Waugh et al. 2013: 618) and there were other, similar ideas in the air. A more likely precedent, and indeed an earlier one, can be found once again in the Ukrainian Volodimir Derzhavin, in this quite remarkable passage from 1927: A human language performs simultaneously (but in every particular case to various extents) three functions: communicative, cognitive and artistic, which are not predisposed to translation in the same degree … So there exist three types of translation: translation-account, translation-transcription (not used separately) and translation-stylization, only the last one being artistic to one degree or another. (Derzhavin 1927: 44, translated in Kal’nychenko 2011: 262)

So where did Derzhavin derive his three functions from? From his text I gather he was in part responding to the young Fedorov (1927/74), but that does not explain why he turned the two functions into three. Derzhavin gives part of the game away when he notes that the ‘communicative function’ had been worked

44

Translation Solutions for Many Languages

on by ‘the Geneva School (de Saussure, Bally)’ (1927: n.3), which sounds quite strange for anyone who has read systemic Saussure but not stylistic Bally. A more explicit connection is found in the Ukrainian translation theorist Oleksandr Finkel who, writing in the same years (1929) and citing Derzhavin, draws his main descriptions of text types explicitly from French linguistics: scientific prose, administrative prose and journalism, the first two of which are found in our old friend Bally (1909/51, sections 132, 238), as is, incidentally, Finkel’s description of calques (1929/2007: 66). Finkel nevertheless sticks to a two-term stylistic description, rather like the young Fedorov’s, declaring that ‘the degree of difficulty and accuracy in the translation of various prose genres varies and depends on the ratio of the structural and aesthetic elements in the genres, on the one hand, and the specific phraseology, on the other’ (1929/2007: 66).7 So it is not a question of types as such, but of a sliding scale of opposed values. This could all be unexciting (so what if there are two or three text types?) were it not for a later scholar who similarly made use of Bally. In 1950 we find one L. N. Sobolev (1950: 143) saying much the same thing as Derzhavin, albeit from the more nuanced and complex claim that the degree to which a translation is ‘precise’, ‘exact’ or ‘accurate’ (точный) – also glossed as ‘truthful’ (правдивый) (1950: 142) – varies in accordance with ‘the purposes of the translation [цели перевода], the nature of the start text, and the reader for whom the translation is intended’ (1950: 143, italics mine).8 This is a key sentence on several fronts. All these pronouncements could be precursors for the ‘principle of the necessary degree of precision’ that Hönig and Kussmaul (1982/96) formulated in German more than thirty years later, similarly proposing that the appropriate amount of information is determined by the required function of the translation. Yet we should also note that Sobolev’s reference to ‘exactitude’ brings in a continuous variable that, in principle, need not be reduced to any simple binarism of translation methods, just as text types can also be described in terms of sliding scales. This was at a time when others, including Retsker, seemed more inclined to look for logical categories in translation. Further, Sobolev’s sentence is secretly exciting for those of us who went through debates about Vermeer’s Skopos theory in the 1980s and 1990s. Vermeer posited that the way you translate depends on the Skopos (‘purpose’) of the translation (in Reiss and Vermeer 1984: 100), and here we find almost the same thing right back in 1950: Sobolev’s term цель (‘purpose’) is indeed the standard Russian translation for the Greek philosophical term σκοπός (purpose, goal, target), so Sobolev was intimating Skopos theory avant la lettre, perhaps. Then again, in his very next sentence Sobolev steers into safer ports: ‘Therefore, it is



A Tradition in Russian and Environs

45

convenient to consider specific criteria of accuracy separately, depending on the nature of texts: artistic, journalistic and business’ (1950: 143). And then we are into a more categorical mode of thought, with the three text types, as in Reiss. Sobolev offers no real reason for this reduction of purposes and readers to text types, but it is not hard to imagine a justification: as most translators realize as soon as they start work on a text, and as Halliday’s theory of text as interaction envisages (Halliday 1977), the nature of the text ensues from its situational purpose and engaged receivers; it is not an initial category in itself. Whatever the toss, in 1950 Sobolev mentioned Skopos but then immediately moved on to texts. The idea that translation types depend on text types was picked up by Fedorov (1953: 198), as indeed was the mention of purpose: ‘individual speech styles that are general to all languages … may be described according to the purpose, the functions and the comparison with the target language’ (ibid.; italics mine). Fedorov similarly appeals to text types, presenting the following threesome (which he says is drawn from previous theories, unnamed): News reports, documentaries and scientific texts, where the translator must pay careful attention to terms; Publicity and texts that have a ‘purely propagandistic intention’ (сугубо пропагандистская установка), where the effect on the reader is what counts (1953: 198); and Artistic (literary) works, where ‘it is important to reproduce the individual particularities of the literary text’. (1953: 256)

Here the three linguistic persons are more visible than in Sobolev (the first type is based on the third person, or things in the world; the second type is designed to have an effect on the second person, the receiver; the third type expresses the first person, the artist). In Fedorov, this idea about text types is very loosely in tune with the second big idea, the one about the different solution types, although I can see no simple mapping between the two sets of three terms. Together, these three ideas effectively situate something like Vinay and Darbelnet’s solution types, and they do so thanks to a mode of thought that was clearly operating at levels above the sentence: there was an awareness of texts, genres and thus of general communicative purposes. The Russian tradition incorporated a level of reflection that drew on Bally for some points but cast its net much wider than comparative stylistics. A further consequence of these ideas is the way they position literary translation (called ‘artistic’ translation in Russian and considered a suitable place

46

Translation Solutions for Many Languages

for ‘transformations’) within a more general theory of all translation. As long as there were ‘regular correspondences’ of some kind, there could be scientific study, hence some kind of linguistics in the sense of a systemic description of linguistic processes and products. Literary translation, which many claimed was an art, was thus implicitly made subject to science. This appears to have been relatively new, although the application of science to literature could be dated from French Symbolism and then Russian Formalism. All these ideas were expressed prior to Fedorov’s main book in 1953.

A development of translation solutions Fedorov’s 1953 book builds on much of the previous work. With respect to the typology of translation solutions, he makes two formal contributions, both of which could seem simplistic. This first is a typology of how a word can be translated: (1) there is no correspondence, so something has to be imported or invented, (2) there is a partial correspondence, so we are left with some kind of similarity, (3) the word has several meanings and can be translated by several different words, (4) in rare cases there is absolute correspondence, as can happen with technical terms, and (5) there is ‘inability to find any correspondence to the word’, which is similarly rare (1953: 122–3). This general approach would be picked up later in German (notably in Kade 1968). The second contribution is a categorization of ways to render foreign terms and names, known in this tradition as ‘realia’ (1953: 140–4): 1. Transliteration: words like esquire and miss can simply be transliterated in Russian: эсквайр and мисс. 2. Creation of a new word by building from particles: skyscraper is небоскреб (sky + scraper). Here we will call this solution type ‘componential’, for want of a better term (in Russian this is sometimes called калькирование, ‘calquing’, which invites confusion with Vinay and Darbelnet’s ‘Calque’). 3. The use of a target-language word with a close, approximate, similar function: instead of the transliteration консьерж (concierge), one might use indigenous Russian terms like швейцар (porter) or привратник (gatekeeper). This can be seen as accepting a level of equivalence for the first two solutions, then incorporating the notion of approximate translation or similarity found in Smirnov (1934) and Retsker (1950). Compare this with Vinay and Darbelnet,



A Tradition in Russian and Environs

47

who only offer Borrowing (emprunt) and Calque for the treatment of foreign terms; they seem to have no major category corresponding to the Russians’ appeal to ‘similar function’. For the rest, Fedorov frustrates my project by not presenting any compact list of seven to ten solution types. He certainly talks about such things: literal translation (дословный перевод) (1953: 155), permutation (перестановка) (change in word order) (1953: 185), grammatical restructuring (грамматическая перестройка) (1953: 156), modification (видоизменение) (1953: 150) and adaptation (приспособление) (1953: 151). If you really wanted to find something like Vinay and Darbelnet, you could, but you would really have to search for it. There is no synoptic table and no particular linguistic formalism that could be used to standardize, interrelate or generate these terms. The typologies that could really compete with Vinay and Darbelnet would be developed by others slightly later, in the 1970s: Retsker, Shveytser, Komissarov and Barkhudarov. Fedorov’s work on translation solutions is nevertheless important because of the way he justifies the central concept of ‘adequate’ translation, on the one hand, and his attempt to balance this against the criteria of different text types, on the other. When describing the translational relation between start and target texts, Fedorov (1953) uses the terms ‘correspondence’ (соответствие) and then ‘adequacy’ (адекватность), interestingly glossed with the Russian compound полноценность, meaning ‘full value’. So if there is a notion of equivalence in Fedorov, it is not a question of measuring values that are ‘equal’ on some isolated level, but of giving something that is the ‘full value’ of the start text, in keeping with the principle of dialectical unity of form and content. The notion of ‘full value’ can be seen as similar to Smirnov’s 1934 proposal that ‘adequacy’ is a summation of ‘equivalence’ plus ‘substitution’. Fedorov then describes ‘adequate translation’ as ‘exhaustive accuracy in the transfer of the semantic content of the original and full functional-stylistic correspondence to it’ (1953: 114).9 More completely: Adequacy of translation means transferring a specific relation between the content and form in the original by reproducing the features of the form or creating functional correspondence to those features. Adequacy of translation presupposes a certain balance between the whole and the parts and especially between the general character of the work and the degree of closeness to the original in the transmission of each particular segment of it. (1953: 114)

48 Table 3.1  Possible development of solution types in Russian translation theory (items in square brackets are not presented in a systematic way) Smirnov (1934)

Retsker (1950)

Fedorov (1953)

Adequacy

Equivalence

Equivalence

Transliteration Componential

Substitution

Analogue correspondence

Similar function

Synonyms

Approximate

Substitution

[Permutation Grammatical Restructuring Modification Adaptation]

Differentiation Specification Generalization Logical Development Antonyms Holistic Transformation Compensation

Permutation Substitution Addition Omission

Concretization Development Antonyms Compensation

Retsker (1974/2007)

Barkhudarov (1975) Transliteration Componential Explanation

Translation Solutions for Many Languages

Batyushkov (1920)



A Tradition in Russian and Environs

49

Note that binary opposition creeps back in here: ‘reproducing form’ or ‘functional correspondence’. This is nevertheless at a secondary level, allowing for ‘adequacy’ to rise above the opposition. In the end, the binarism may be no more than the two directions in which translation solutions can be sought at phrase or sentence level, as in Vinay and Darbelnet. Indeed, the above description locates ‘adequacy’ at the level of the text as a whole (in relation to its parts), suggesting that it is not in itself a term to be applied to each phrase-level solution when considered in isolation. As for the psychological tendency to binarism, when Fedorov comments on phrase-level solutions he usually formulates more than two ways of solving problems. If anything, he prefers divisions into three. If we focus on no more than the technical problem of naming solution types, it is possible to see a system developing as we move from Smirnov (1934) through to the more developed typologies of the mid 1970s (see Table 3.1). Here I am picking out no more than a few visible buoys, most of which are anchored to developments that are deeper and more complex. I am also hesitant about putting in Fedorov’s terms for the types of ‘substitution’, since they are not presented in any succinct list in his 1953 book. Despite all of that, Table 3.1 does show something growing. It is not hard to compare these solution types to Vinay and Darbelnet’s. Read in the Soviet context, however, the differences were far more important than the similitudes. You see, there was a spanner in the works: the relation between translation solutions and text types was not quite as easy as the theories made it seem. This was because literary scholars did not exactly leap to the idea of having their translations dictated to by the abstractions of science. That particular conflict requires a little history.

Here are the politics Although Fedorov has his name on none of the really original ideas in Soviet translation theory, I suspect he was the key historical figure nevertheless. This is not just because he was able to package together some key concepts in his 1953 book; it is also because of the collective development of Soviet thought prior to 1953. To grasp this, it is helpful to trace Fedorov’s biography through the turbulent politics of the period. Andrey Venediktovich Fedorov (1906–97) studied at the State Institute for the History of the Arts in Leningrad (Saint Petersburg), where the Formalists of the early twentieth century had created a programme. The indications are that he was a

50

Translation Solutions for Many Languages

student of Yuriy Tynyanov, who had worked on the theory of cultural systems and wrote on Russian translations of Heine,10 as well as the phonologist Lev Shcherba, the comparativist Viktor Zhirmunskiy and the linguist Viktor Vinogradov, who studied linguistic styles in Russian literary texts, with some influence from Bally.11 Fedorov would have thus started out as something like a literary historian grounded in study of systems and abstractions; he had no particular need to look to Geneva for theory; he was not trained as a ‘linguist’ in any restrictive sense (although this is what some critics would later say about him). His intellectual background included the Russian Formalists who worked in Saint Petersburg and Moscow at the beginning of the twentieth century and who, like Bally and the Geneva school, were out of favour in the Soviet Union in the 1930s and 1940s. Fedorov’s interest in translation might be dated from the 1927 paper we have seen (presented in 1925), which was followed in 1929 by a paper that also dealt with Russian translations of Heine. In 1930, at the age of twenty-four, Fedorov has a version of his essay ‘Приемы и задачи художественного перевода’ (Methods and problems of literary translation) printed as an accompaniment to Korney Chukovskiy’s reflections on literary translation in Искусство перевода (The Art of Translation) (see Leighton 1984: xxxii).12 Retsker (1974/2007: 6) reports that Fedorov gave lectures on translation theory at the Gorki Literature Institute in Moscow in the 1930s.13 Fedorov’s early work is indeed among the literature cited by Smirnov in 1934, so he was very much part of that mix. One might imagine that Fedorov’s intellectual development would then be continuous and accumulative, and well it might have been, at least in terms of private conversation and thought. In the world of public theorizing of translation, however, there was decided discontinuity. Through to the late 1920s, it seems, the public discussion of translation could fairly easily embrace debate about the existence or otherwise of rules; it could involve exchanges between literary translators and the systems-based work of the Formalists. It also appears to have been able to extend across language boundaries. As Ukrainian asserted its national status – the Academy of Sciences of Ukraine was founded in 1918; Ukrainian was taught in schools; it was the language of operas performed in Kiev from 1926 (Chernetsky 2011: 46) – literary translation into that language assumed a social mission, building up a national literature and stimulating serious public thought about translation. That was a very particular period. If you look at the dates, you find Tynyanov, Fedorov, Derzhavin and Finkel were all publishing on translation in 1927–9, and their exchanges were as much about evaluating literary translations and discussing actual practice as they were about abstract theory.



A Tradition in Russian and Environs

51

That space of debate disappeared abruptly in 1932–3, when Stalinist ideology turned to centralized control and imposed Russian language and culture (Chernetsky 2011; Kal’nychenko 2011). The shared space disappeared because many of the people in it disappeared: the proponents of Ukrainian cultural independence were sent to gulags or otherwise silenced (Ukrainians call this the period of their ‘Executed Renaissance’, rozstriliane vidrodzhennia); their contributions are significantly not cited in Smirnov’s encyclopaedia article of 1934. At the same time, there was a reaction against what we might term ‘non-social’ linguistics. Charles Bally, for example, could clearly be cited and discussed in the 1920s (as we have seen), but not in the 1930s,14 and Bally-like stylistic linguistics, including Vinogradov’s (cited in Fedorov 1927/74), was more or less banished. The heritage of the Russian Formalists was also severely questioned. So how would someone like Fedorov ride out such storms? Here are a few biographical notes that I have been able to glean. In 1934 Fedorov is named as the author of a twenty-six-page Теория и практика перевода немецкой технической литературы на русский язык (Theory and Practice of German Scientific and Technical Texts into Russian), which sounds like a practical textbook, a long way from literary history. Shadrin (2011) reports that in 1940 Fedorov defended his doctoral thesis on ‘The Linguistic Foundations of Translation Theory’ (Лингвистические основы теории перевода), which might also suggest he was moving away from literature and into more neutral technicalities. He nevertheless has a work called О художественном переводе (On Literary Translation) listed as being published in 1941. Fedorov was then a translator at the Leningrad front during the Second World War (Shadrin 2011), after which there seems to be something of a gap. In 1952 Fedorov’s name appears on a translation of Goethe’s Wahlverwandtschaften, so his concern with literary translation had not disappeared. And then in 1953 we have his book on translation theory, the second edition of which, in 1958, added the word ‘linguistic’ to its title. After that, the biography reads like fairly clear sailing: Fedorov went on to author more than two hundred publications on the theory, history and criticism of translation, on general and comparative stylistics, the history of Russian poetry and the international relations of Russian literature. He was head of the Department of German Philology at Leningrad State University from 1963 to 1979. As I have mentioned, one of the questions raised by the biography is why Fedorov should have been associated with a narrowly ‘linguistic’ approach to translation in the 1950s. One possible explanation here is that ‘linguistics’ has a wide application in the Russian tradition, including any systematic approach

52

Translation Solutions for Many Languages

to the study of both literary and non-literary texts, and thus embracing the Formalist legacy. That wider sense would certainly fit in with the biography. Yet Fedorov himself suggests a possible motivation for his ‘linguistic’ approach in the first pages of the 1953 book: J. V. Stalin’s brilliant work Marxism and the Problems of Linguistics took Soviet linguistics out of the deadlock it had reached because of N. Y. Marr and his disciples, who had created a despotic regime in linguistics. Translation theory now has the chance to be treated on the grounds of an authentically scientific linguistic approach and the correct treatment of the central problems of linguistics … Before the appearance of Stalin’s work on linguistics, these questions, which have priority not only in the sphere of philological science as a whole but also with regard to translation theory, either were not put forward or were not given the right solutions. It is evident that only in the light of Stalin’s work on language can the problems of translation be solved. (1953: 97)

Not surprisingly, these passages were deleted in later editions of Fedorov. They might be taken as a form of necessary lip service to the state ideology of the time and thus dispensable in a later era. There nevertheless remains a sense in which the praise of Stalin works as more than a plea for imprimatur. The reference is to an article published in Pravda in 1950, where Stalin declared that a language was not the institutional product of a particular class or economic base (it was not part of the ‘superstructure’): Language is not a product of one or another base, old or new, within the given society, but of the whole course of the history of the society and of the history of the bases for many centuries … Hence the functional role of language, as a means of intercourse between people, consists not in serving one class to the detriment of other classes, but in equally serving the entire society, all the classes of society. (Stalin 1950/54: 5)

In the Soviet context, Stalin’s intervention ostensibly put paid to long-standing arguments against the formalist study of language, in which the likes of Bally were included. Thanks to Stalin, in principle, language could now be studied more or less on its own terms, including in terms of synchronic systems. Mossop (2013) argues that Stalin’s pronouncement opened the way for what is perhaps the first formally linguistic approach to translation, namely Fedorov’s 1953 book. The actual causal relations are nevertheless difficult to gauge. A simplistic reading (offered for example in Brang 1955/63) is that, for as long as language was considered part of the superstructure, it could be handled



A Tradition in Russian and Environs

53

as pure ideology and could thus be altered for ideological purposes – translators could freely make every text sound like Marxist-Leninist ideology. Following Stalin’s decision, ‘such changes are in theory no longer allowed’ (Brang 1955/63: 398). That is far too easy, especially since Fedorov’s theorizing in 1953, after Stalin’s intervention, actually allows for the correction of ‘facts’ (as we shall see below). It seems more plausible that Stalin’s articles actively politicized linguistics, bringing the discipline to the centre of intellectual attention and thereby opening new opportunities. Viktor Vinogradov, the Russian linguist who had been influenced by Bally in his early work on phraseology, had been banished into internal exile in 1934 but was brought back in 1950 and made Director of the Institute of Linguistics in Moscow. According to one report (Cary 1957), which I would like to see corroborated, a chair in Translation Theory was established at the Institute of Linguistics in the same year, 1950. Thanks to all this attention, so the story goes, Stalin created a moment when linguistics was well positioned institutionally and was able to explain translation in ways that need not be indebted to the dominant position previously held by literary studies. So where might Fedorov have found the intellectual tools needed for a linguistic theory of translation? His background, we have seen, was in literary history of a systemic and formalized kind, which might certainly be termed ‘linguistic’ but was not borrowed directly from any systematic account of language alone. If you follow the citations, there is a strange lack of anything like a ‘master linguist’ in his works. Fedorov’s 1953 book actually makes no reference to Tynyanov or anything close to the Formalists, cites little from his own earlier papers and makes no more than passing references to Chukovskiy.15 Fedorov does approvingly cite Viktor Vinogradov, the linguist who had been brought back from exile and made Director of the Linguistics Institute; this could be no more than a politically correct name to cite were it not for reports that Vinogradov had been one of Fedorov’s professors,16 so it remains a name of significance – yet there is little to suggest that the restoration of Vinogradov brought back any underground cult of Bally (much as I would love to discover such a connection). And for pointers on translation, we have seen that Fedorov refers to the collective volume Вопросы теории и методики учебного перевода (Theoretical and methodological questions of translation teaching), published in the same year as Stalin’s pronouncements (1950) and thus also presumably on safe ground. More importantly, albeit strategically, Fedorov loudly complains that in previous Soviet approaches ‘little attention was paid to the grammatical questions of translations and comparative study of two languages can map the differences between them, which helps to find corresponding solutions.

54

Translation Solutions for Many Languages

However, there has not been contact between translation theorists and comparative grammarians’ (1953: 101, italics mine). The implication is that Fedorov’s project will be, like Bally’s and Vinay and Darbelnet’s avant la lettre, some kind of comparative study that can help locate translation solutions. And the ‘lack of contact’ here might refer to a separation of literary translation from the rest – a separation that had actually been bridged, as we have seen, by the trajectory of Fedorov’s interests since the 1920s and indeed in his training. Why should this relation between linguistics and literary translation have been so important? Part of the answer certainly lies in the World Literature series initiated by Gorkiy, in which some 120 titles had been translated into Russian by the mid 1950s (or so reports Cary 1959), creating a prestigious ‘Soviet school of translation’ exclusively in the literary field (Witt 2016). Strangely, in the Soviet context of the 1950s, this seems to have affected the very possibility of studying translation in a unified way. Indeed, it bore on one of the origins of what we these days call Translation Studies. Here I read Fedorov’s 1953 text: That is why one is so surprised at Prof. A. A. Reformatskiy when he asks, in his article ‘Linguistic problems of translation’ [1952], ‘Is a science of translation possible?’ and he answers, ‘No, such a science is impossible; translation practice can use the knowledge of many sciences, but cannot have a science of its own. This is the consequence of the diversity of translation types and genres.’ This argument is completely groundless. Indeed, it can be very difficult and complex to systematize and generalize the different forms (формы) adapted by the correlation of regularities (закономерности) between two languages when working with different genres (жанры) and different types of material, but that does not mean it is impossible to carry out that task. (1953: 15)

If the way you translate really does depend on the text type (we have seen the idea announced in 1927, then again in 1950 and now picked up by Fedorov in 1953), then each text type could have its own way of studying translation and there would be no unifying Translation Studies. On the other hand, if the concept of a common ‘adequacy’ is ranked higher than the principle of the different text types, then it is possible to study translation from a shared perspective, with a unified approach that includes a typology of translation solutions. I will return to this discussion later. For the moment, let it suffice to note that, in envisaging a common scientific approach, Fedorov’s starting point in 1953 is not so much the aspiration to a general ‘adequacy’ in translation as it is the observation that all kinds of translation deal with language – which is indeed their ‘common



A Tradition in Russian and Environs

55

denominator’. Hence, it seems, the usefulness of ‘linguistics’ and the need to talk with comparative grammarians. At the same time, as Fedorov makes equally clear, the intellectual challenge of identifying the ‘forms of regularities’ is by no means as banal as this ‘common denominator’ might appear. And there was no suggestion that such forms would somehow be laws legislating the way all translators must act, let alone promulgating primitive literalism. When negotiating the relations between linguistics and literary studies, Fedorov is actually quite cunning in that he places literary translation in advance of linguistics, so that linguistics has to catch up. This might be no more than a trope of false modesty, perhaps a wink at Stalin’s criticism of Soviet linguists, or facile flattery of the literary translators who were to oppose him. Whatever the case, that simple argument implicitly positions Fedorov outside any traditional linguistics (how else could he see that it needed to catch up?), at the same time as it posits that literary translators are people that linguists can learn from (before the relationship might work the other way around). So, given this scenario, how could linguistic science ‘catch up’ to the advances made in literary translation? Fedorov proposes the following principles (1953: 98–100): 1. The translator has ideological responsibility for the quality of the translation, hence the requirement that the translation be truthful, in order to give the Soviet reader the full picture of the translated materials. 2. The translation must use ‘complete language’ rather than literalism, with no kind of violation against the mother tongue in favour of the start language. It should be noted that in practice one does find cases of literalism (or so-called ‘translationese’), which present both the start and target languages. 3. Since the literary work has dialectical unity of content and form, its translation must concord with the start-text function [функция] (of both separate elements and the work as a whole). 4. Since in the case of literature the start text is a single (meaningful, artistic) whole, where each separate element has its own meaningful (and also artistic) role and which, in its turn, has a definite cultural background … the translation must transfer the uniqueness of the original (the speech features of the given genre, the individual style of the author, the historical context), which would correspond to the meaningful and artistic roles fulfilled both by the original as a whole and by its individual features. This requirement is only fully feasible on the basis of analyses of the author’s

56

Translation Solutions for Many Languages

style in terms of linguistics, which is only starting to be worked on by Soviet scholars. 5. Translation theory and practice should be approached through constant use of scientific data on the history of a nation, history of language, history of culture, history of literature, etc., i.e. the principle of studying phenomena in their interrelationships. This is also associated with the requirement to take into account the actual conditions and the purposes [цели] for which the translation is carried out, whether it is a literary translation designed for publication, a documentary-business translation, or an interpreting assignment. 6. The principle of translatability must be accepted, i.e. the possibility of a complete translation, which is only attainable because we do not proceed from individual elements … but from the complex whole in which they are joined together and acquire their specific meaning. (Translation by Nune Ayvazyan.) For my purposes here, the most important principle is perhaps the last one, since acceptance of translatability is the necessary premise for any study of translation solutions. Principles 2, 3 and 4 then work together: if the start text has no primal separation of form from content, of signifier from signified, then it is not possible to separate translation strategies on the basis of any similar binary opposition: the translation must work on all levels at once, using whatever resources are necessary. The use of comparative linguistics, of science, would thus correct the excesses of literalism, leading to ‘adequate’ translation: translation will progress thanks to science. Pleasantly surprising, in view of later translation theories, is the way principle 5 extends scientific inquiry into the purpose and genre of the translation, which are here combined with an awareness of the way wider social and cultural systems function. But then, progress also had to pay the piper: the first principle, read in context, requires the translator to remain truthful to historical reality, presumably as interpreted in terms of Soviet ideology, rather than as fidelity to the start text. This ‘ideological responsibility’ resonates with the above principle that texts with a ‘purely propagandistic intention’ should be altered so as to achieve the desired effect (1953: 198) and potentially extends it to all texts. And this is also what could be meant by ‘the purposes [цели] for which the translation is carried out’ in principle 5. In the Stalinist context, the appeal to what we now call Skopos could also be seen as a justification of Party propaganda.17 Summarized as such, Fedorov’s approach actually owes a great deal to literary studies – he had hardly sold his soul to any non-literary linguistics. The ‘dialectic



A Tradition in Russian and Environs

57

unity’ of form and content (principle 3) is seen as a quality of literary works (we are not told if it extends to other kinds of texts). And the text as a ‘meaningful whole’ (principle 4) makes specific reference to ‘artistic roles’, which may or may not exist beyond the realm of literature. These items constitute a legacy of the Formalist tradition – Fedorov was trained in system-based literary history and he could be speaking from that position in most of these principles. Yet the 1953 work is presented as being linguistic in its overall conception, purporting to apply to all kinds of texts, as befits a general scientific theory and responding to Stalin’s call for a new orientation in linguistics. Its techniques were thus potentially applicable to all translations. And that was precisely the point that most rankled linguists and translators after its first publication.

Literature vs. science and a possible birth of Translation Studies Fedorov’s 1953 book was indeed criticized by theorists of literary translation (as noted in the fifth edition).18 Retsker recalls that I. A. Kashkin, ‘one of the founders of the Soviet school of literary translation and educator of a whole galaxy of talented translators, accused Fedorov of “formalism”’ (1974/2007: 6). It seems Kashkin was not against theory as such; he was more against having art constrained by science. Cary, commenting on the 1954 Congress of the Union of Soviet Writers, notes: In a collection published the following year, literary translation is presented as belonging to literature. Whereas Fedorov had denounced the ‘literary’ deviation, now his approach is qualified as ‘linguistic’ deviation. His theory comes in for a heavy beating … (1957: 187)

If the basic argument is that literary translation belongs to literature and should therefore not be subject to abstractions coming from other disciplines, there is room for respect and discussion: medical translation belongs to medicine, film subtitles should be dealt with in cinema studies, court interpreters are bound by the rules of each legal system (which they are) and should thus be studied in faculties of law. That is a legitimate debate. And yet the solution types we are exploring here are somehow so basic and obvious (at first sight) that they are surely offering a common conceptual framework for all text types? That debate, however, is a long way from the mutual accusation of ‘deviations’. In the postStalinist climate (Stalin died in March 1953), the debate was heavy, dangerous and by all accounts spirited.

58

Translation Solutions for Many Languages

So why were the literary translators and theorists apparently so set against Fedorov? What did it mean to be accused of ‘formalism’ or ‘deviation’? And what kind of ‘science’ was being invoked anyway? One sense of formalism may involve a particular way of translating. This is the sense in which we find Kashkin using the term in 1951, accusing ‘formalism’ of separating form from content, apparently with very negative consequences: In their deliberately arcane versions, formalist translators mutilated the Russian language, imitating the foreign language as a matter of principle even when there was no stylistic justification for it, such as a need to give a sense of local color or to highlight characteristics of direct speech. (Kashkin 1951: 2; cit. Levý 1963/2011: 16)

The worrying part is that this position was cited by Levý, in Czech and German, making it one of the few fragments that actually escaped from Russian, back in the day. Highly literalist translators have certainly existed; some may have exerted revolutionary zeal in this way (as did Lu Xun in China, for example); they might have been language enthusiasts of the kind that Stalin was writing against. Or it could be, more probably, that Kashkin here is simply twisting Stalin’s basic argument in order to further his own agenda – Witt (2016) identifies his position as an attempt to formulate a correlative of Socialist Realism in the field of translation theory and to promote it as the ‘Soviet school of translation’, corresponding to anti-cosmopolitan thought. Whatever the aesthetic or political motivations, no names appear to be given for these ‘formalist’ translators, and Fedorov’s linguistic approach (let alone Sobolev’s and Retsker’s) can scarcely be construed as justification of such literalism.19 Reading Fedorov’s text, even the hardline 1953 version, it is difficult to imagine his approach justifying, or indeed practising, translations that ignore content, that fail to see the work as a whole or that produce mindless mimicking of the foreign. I suspect that the theoretical debate was not just about different ways of translating (or ways of relating to the foreign), nor just about anti-literalism, as defending the linguistic purity of the target language. I think it also had to do with the social status of the translator and the concept of science as progress. This may require explanation. When I first started studying translation, in the late 1970s, there were innumerable papers addressing the question of whether translation was ‘art’ or ‘science’. I never understood why the issue might be of concern to anyone – what difference would it make? Now I am beginning to understand who could have been caring. Consider the following, lifted from Chukovskiy’s diary for 1918 (1991/2005: 40):



A Tradition in Russian and Environs

59

I had a run-in with Gumilyov at the meeting. A gifted craftsman, he came up with the idea of creating a ‘Rules for Translators’. To my mind, no rules exist. How can you have rules in literature, when one translator ad-libs and the result is top-notch and another conveys the rhythm and everything and it doesn’t go anywhere? Where are the rules? Well, he lost his temper and started shouting. Still, he’s amusing and I like him. (12 November 1918)

Chukovskiy would go on to write a book that was very clearly in favour of art and against the existence of rules associated with that art. So why should Fedorov now be placed on the other side of that debate? In the above encounter – in 1918 linguistic arguments could be vigorous but amicable (later they could have you banished from state institutions, or worse)20 – the question is whether there should be set rules that translators must follow. On that particular question, Chukovskiy, Fedorov and Kashkin would very probably agree that any such rules would be so basic as to be uninteresting. All three would more or less be on the same page. Quite another question, though, is whether there are regularities in what translators do when seen from the outside, when translation is seen as a collective historical activity carried out within dynamic cultural systems, and whether those regularities form clear options, points at which a decision can take the translator down one of several paths. The formalist studies of the beginning of the twentieth century were very much in terms of that second kind of question, analysing literatures as systems and marginally viewing translations as ways of relating systems. If that is what is meant by ‘formalism’, then it is quite different from having a scientist tell translators how to do their job, and it is a world away from translating ‘form’ to the exclusion of ‘content’. Part of the problem here (as mentioned in our discussion of Bally) is what is meant by a ‘rule’? If it were clear that any rules of translation are general tendencies, subject to historical social conditions – as, for example, in the sociological concept of translation ‘norms’ in Toury (1995/2012) – there might not have been any debate at all. In context, though, the underlying problem was that during the 1930s the image of intellectual ‘formalism’ had been disassociated from sociology and had thus been opposed to properly ‘Marxist’ linguistics. Formalist linguists had been sent into exile; the debate was no longer just a lively discussion among intellectuals. And the specific problem here, in Fedorov (1953), is that the placing of literary translation within a general theory of ‘regular correspondences’ does indeed bring in stronger senses of ‘rules’: there are the obligations of grammar, and technical texts do have fixed equivalents, after all. It is easy, in such a context, to stretch the pejorative sense of ‘formalist’

60

Translation Solutions for Many Languages

so that it includes grammar and terminology and thus opposes all literary art. But is that what Fedorov was actually doing in his book? Not if you read the book. Was that debate important? Perhaps not, in the long run, since the linguists somehow won out, at least for a while (but now they are out of favour again). I nevertheless note one further report on the affair. Edmond Cary (1959) at one point remarks that this Russian argument between linguists and littéraires momentarily led to a proposed solution, during a conference in Moscow in 1958: instead of opting for one approach and opposing the other, the question was asked: ‘why not have a separate science for all forms of translation?’: At first blush, one is tempted to smile and shrug one’s shoulders: this would merely avoid the difficulty rather than resolve it. Upon reflection, though, one wonders if it might not be, at least given the current state of affairs, the most equitable solution, the one that best accounts for the positions on the ground. (1959: 19n.)

That might be seen as a conceptual beginning for Translation Studies, born of a conflict between literature and linguistics, well prior to the papers and meetings that claim to have set up Translation Studies in the 1970s. Unfortunately the news seems not to have travelled to all corners of the globe.

So how was this reported in the West? The work of the Soviet theorists became known in Western Europe through several early reports, notably by Brang (1955/63) and Cary (1957, 1959), plus reviews in 1956 and 195821 (which I have not been able to find). There are also echoes in the 1960s, notably in Kade, whom we will meet later. The ideas of the Soviet theorists nevertheless failed to leave a lasting mark beyond Russian, much to the detriment of Translation Studies. Some absences seem curiously marked by the Cold War presuppositions. The ground-breaking anthology of theoretical texts edited by Störig in 1963, for example, omits Fedorov, ostensibly because the Russian examples meant nothing to the German reader (or so says Störig the editor) but also because ‘Fedorov’s theoretical approach is fundamentally based on Stalin’s pronouncements on linguistics and should these days hardly be of any significance in the Soviet Union itself ’ (1963: 384). To this day, Fedorov’s main text has not been translated into any of the major Western European languages.



A Tradition in Russian and Environs

61

Störig did nevertheless reprint Peter Brang’s article on the Soviet theorists. Brang (1955/63: 388) makes the point that Fedorov was in fact reacting against a narrowly Marxist linguistics, following Stalin’s pronouncements of 1950, which would seem to contradict Störig’s dismissal of the theory just four pages previous. He also explains the idea of ‘practice’ being ahead of ‘theory’, referring to the All Union Conference of Translators held in December 1951. Brang then offers a ten-page summary of Fedorov’s book, including the three ways of dealing with foreign realia and the three kinds of texts, which Brang insists are derived from Stalin’s theory of language as a treasury of key words – personally I can’t find the connection. In Brang’s summary, Fedorov is making general points about the different kinds of language used in the three text types, but it is not abundantly clear from his description that the way you translate depends on the text type – the novelty of the idea remains obscure. In 1959, one year after publication of Vinay and Darbelnet, Edmond Cary published an article on Soviet translation theories. At the time, Cary was many things: co-founder of the very young journal Babel, co-founder of the Société Française des Traducteurs, co-founder and Secretary General of the Fédération Internationale des Traducteurs and an interpreter working for UNESCO. A review by Cary had every chance of being read seriously. So how was it that an influential French interpreter and co-founder of associations became an intermediary for Russian theory? Edmond Cary, it turns out, was born in Saint Petersburg in 1912 with the name ‘Kirill Znosko-Borovskiy’ – which might explain his knowledge of Russian (see Ballard 1985: 9–10 for further biographical details). His family moved to Paris after 1917 – that might explain why he is not wholly in agreement with Soviet ideologies, and perhaps why he picks a petty fight with Fedorov. Cary stresses the importance of translation for Russian culture, not just because of its relations with foreign cultures but also for the many texts that move between the official languages within the Soviet Union – an aspect that is strangely absent from many of the theories we have been looking at. At the time, the Soviet Union was the country with the world’s highest number of translations per year. Cary sees Soviet translation theory as having begun from the translator-training courses organized for the World Literature series managed by Gorkiy – the courses would have started from those given by Chukovskiy in 1919 and then led to Chukovskiy and Fedorov’s Искусство перевода (The art of translation) of 1930. Cary notes the extreme difference between the two authors of that work: Chukovskiy writes as an artist giving advice, Fedorov as a scholar mainly concerned with questions of verse form. Cary then remarks in passing

62

Translation Solutions for Many Languages

that many Soviet theorists assume the translator is a reader who has learned the foreign language and culture through print: ‘the [nineteenth-century] group of travelling intellectuals has disappeared, replaced by translators who are less complete [plus frustes] and only know the foreign from the printed page’ (1957: 182). Cary also takes note of Smirnov’s 1934 article, translating Smirnov’s ‘адекватность’ (‘adequacy’) as ‘pleine équivalence’ (‘full equivalence’), which seems not to be what Smirnov would have done, and then he uses the same translation in his summary of Fedorov. True, Smirnov and Fedorov did gloss ‘adequacy’ as ‘full value’, but neither of them was using the term ‘equivalence’ in this way. Cary appreciates the way Fedorov develops a translation concept that goes beyond purely literary concerns, and he accepts the important proposition that ‘each of these different types imposes different requirements on the work of the translator’ (1957: 186). Yet he is unhappy with the idea that the ‘common denominator’ of all these types of translation should be ‘linguistic’. He seems to think that many other kinds of translation (dubbing, conference interpreting, children’s literature) somehow break the linguistic mould (it is not clear why) to the point where only truisms remain. He is fundamentally upset that the translator’s art, or even thought process, should be reduced to decisions between formal categories, and this is indeed the way he conceptualizes ‘linguistics’, as nothing more than formalism (on the debate between Cary and Mounin over this point, see Ballard 1985). In this way, the possibly revolutionary Russian ideas about text types remain hidden under a very restrictive misconception of linguistics. Something similar happens in Cary’s presentation of Sobolev’s ideas from 1950. Cary does translate the key sentence that I have cited above: ‘The degree of exactitude in translation varies in accordance with the purpose [la destination] of the translation, the nature of the text translated and the readers for whom the translation is intended’ (trans. Cary 1957: 186) and he actually repeats this principle in his own course on translation (1985: 85). The French term ‘destination’ is not exactly wrong as a rendition of ‘цель’ (‘purpose’), but it could also be read in its unhelpful physical sense. To make matters worse, Cary’s commentary then describes this ‘destination’ as concerning different historical conceptions of translation (which is a good thing to promise, but is not what Sobolev was about), rather than a position with radically practical consequences concerning the social use of language. This then degenerates into an argument in which Cary sincerely regrets Fedorov’s dogmatic defence of translatability and the consequent abstractions in which general principles are applied to all



A Tradition in Russian and Environs

63

kinds of translation. Read with hindsight, there is something rather sad in the way the truly novel ideas of the Soviet theorists were obscured by the sterility of main debates of the day (art vs. science, literature vs. linguistics, untranslatability vs. translatability). In 1959 Cary returns with a short review of the second edition of Fedorov’s book, which had appeared in 1958. Cary notes the lively debates that had ensued between Fedorov’s ‘linguistic opinions’ and the Soviet ‘littéraires’, observing that Fedorov’s second edition had toned down some of the ‘rigorous but simplistic systematicity’ (1959: 19). All of this wins Cary’s approval, since Fedorov’s work has become ‘more nuanced’, ‘more complex’, and also because Cary himself does not believe that it is possible to ‘reduce the different text genres to abstract (linguistic) “common denominators”’ (1959: 19n.). In sum, one senses that Cary wanted the Russians to be known but was himself more attached to the literary translators than to the linguists. He might have appreciated that Fedorov was more than a linguist; he might even have seen that part of Fedorov’s appeal to linguistics was a convenient political shield following Stalin’s pronouncements of 1950; but he was not about to entertain any intrusion by progressive science. When all is said and done, he belonged to the ‘heroic age’ of Western translation practice – he shared the mystique of those prodigious polyglots who performed near-magical feats in the service of humanity. Let linguistics be applied to the machines; the mystique of the master was not to be touched. For many readers who had no Russian (Mounin, for example, and the likes of Vinay and Darbelnet), Cary was the main source for knowledge on Soviet translation theory. He was an authoritative and dynamic intermediary and he would undoubtedly have done more had he not been tragically killed when an Air India flight crashed into Mont Blanc on 24 January 1966 (Ballard 1985: 9). The mediating voice was cut short. Cary’s ultimately negative reductions are also found in some later theorists who perhaps should have been reading in Russian. The Czech scholar Jiří Levý (1963/2011: 5), for example, records only the ‘pointless and fruitless’ polemics between Fedorov and Chukovskiy, between science and literature; the German theorist Rudolf Jumpelt (1961: 6), who worked with Cary in the Fédération Internationale des Traducteurs, picks up from Fedorov only the idea that linguistics might be the ‘common denominator’ (gemeinsamer Nenner) beneath all text types; Otto Kade (1968: 24), who cites Fedorov many times (always in Russian, without translation), openly regrets that the Russian had not made the relation between content and form more dialectic22 (Cary had also remarked

64

Translation Solutions for Many Languages

that the concept of translatability should have been more dialectic); and Wilss, writing in 1977, refers to an unpublished German translation of Fedorov’s main work23 to remark little more than that Fedorov had seen translation as a ‘purely linguistic operation’, as belonging to the field of linguistics, whereas it must also be seen as a ‘psycholinguistic and sociolinguistic process’ (Wilss 1977: 76, 77). If they had looked a little harder through the smokescreen of Stalinist linguistics, those readers might have gained some inkling of the three ideas that started this chapter. It seems, however, that many of the Western and Westernized translation theorists – this does not include Kade! – saw the Soviets as being held back by Stalin, being reduced to banalities by linguistics and remaining locked in a sterile debate between literature and linguistics (a debate that is hardly evident in the theories themselves). And nowhere in any of these presentations would the Western or Westernized reader have any notion that the Russian thinkers were steadily developing a typology of solution types. Hence, perhaps, a huge and lasting historical disconnect.

Postscript 1: A durable classification Since the 1970s the linguistic classifications of solution types seem to have become standard fare in Russian-language theory, presumably because of their usefulness for training in a country with extensive internal and external translation demands. The main classifications are nevertheless not from Fedorov, who was more important for the theoretical and ideological debates we have seen. The references tend to be to a series of textbooks produced in the 1970s. One of them is Retsker (1974/2007), who describes the ‘formal relations between concepts’ as giving the seven types of ‘lexical transformation’ mentioned above (1974/2007: 43). Most of those categories also appear in one form or another in Barkhudarov (1975), which is among the most lasting of the typologies. Here I look briefly at Barkhudarov only, for the purposes of later comparisons with similarly contemporary traditions. Barkhudarov (1975) recognizes five main types of translation solutions: Transliteration/Transcription, Componential Translation (called Калькирование, ‘tracing’, ‘calque’), Added Explanation, Approximate (analogue) Translation and Transformation (1975: 96–101). The first three are ways of retaining the foreignness of the foreign expression; they can be seen as Fedorov’s categories plus an explanation of the expression. ‘Approximate’ Translation had also been within the Russian tradition at least since Retsker (1950) (see Table



A Tradition in Russian and Environs

65

3.1). An important shift, however, has occurred in the sense of the term ‘equivalence’, which was used by Smirnov (1934) and Retsker (1950) to designate the kind of one-to-one relation established between words and phrases on the level of form and which Retsker (1974/2007: 13) described as a match at word or phrase level that is ‘permanent’ in that it ‘usually does not depend on context’.24 Now in Barkhudarov the same term ‘equivalence’ (эквивалентность) refers to ‘maintaining the content, i.e. the values’ (1975: 11) by performing whatever transformations are necessary. In these same pages, not by chance, Barkhudarov is citing Catford (1965), who used ‘equivalence’ in more or less this way. So that term changed, but the rest remained virtually the same. The more interesting part of Barkhudarov concerns his ‘translational transformations’, which are then further classified into four main types: Permutations, Substitutions, Additions and Omissions, all of which usually work in combination (1975: 229). In the previous parlance, all of these would presumably have been classified as ‘substitutions’; now the linguistic catch-term ‘transformations’ has won the day. The transformations are described as follows: Permutations (перестановки): Change in the order of words and phrases. For the sentence ‘A suburban train was derailed near London last night’, the order of phrases in Russian can be exactly the opposite of that in English (1975: 190). Barkhudarov claims that this is because English obeys fixed syntactic preferences, whereas phrase order in Russian expresses new vs. old information. More generally, a highly inflected language has more expressive element order than does a weakly inflected language. The interesting point is that this is a category that is missing from Vinay and Darbelnet. Substitutions (замены): One kind of element replaces another. This can happen at several levels: Word forms, requiring obligatory adjustments in accordance with grammar and lexicon. Parts of speech: noun for pronoun, noun for verb, adjective for noun and so on. Sentence structure: when a sentence is переструктурирование, ‘restructuring’), as in: ST: Не was met by his sister. TT: Его встретила сестра. [To him met sister.] (1975: 198)

entirely

recast

(called

66

Translation Solutions for Many Languages

Most of the examples here seem to involve active-to-passive transformation, although there are also transformations between main and subordinate clauses: ST: I like watching her dance. TT: Я люблю смотреть, как она танцует. [I love to watch, as she dances.] (1975: 203) The other types of ‘substitution’ listed are Concretization (the translation uses items with a narrower meaning, for instance when ‘I said’ is rendered as ‘I asked’), Generalization (the opposite of Concretization), Antonymic Translation’ (transformations between positives and negated negatives) and Compensation (Компенсация), which I will discuss below. Additions (добавления): The presentation of items that are not expressed in the surface of the start text but are implied. Barkhudarov cites the transformational linguistics of Zellig Harris at this point and the notion of ‘relevant words’ (1975: 220) – I will return to this in a later chapter on transformational grammar. ST: The new American Secretary of State has proposed a world conference on food supplies. TT: Новый государственный секретарь США предложил созвать всемирную конференцию по вопросам продовольственных ресурсов. [The new American Secretary of State proposed to convene a world conference on food resources.] (1975: 221) Here, Barkhudarov argues, the item ‘world conference’ calls up the relevant verb ‘to convene’ in English but does not do so in the same way in Russian. The elided verb must thus be inserted in the Russian translation. In Vinay and Darbelnet, this would count as Explicitation. Omissions (опущения): The leaving out of words that are semantically redundant (1975: 225–6), apparently because Russian tends to use few synonyms and does not like repetition: ST: just and equitable treatment TT: справедливое отношение [fair treatment] (1975: 226) This four-term classification is easy to remember and seems quite robust. Then again, many things can be thrown into the middle ‘substitutions’ pot, and it is not made clear how they relate to each other. All the categories come with numerous examples, mostly drawn from textbook English novels, with the



A Tradition in Russian and Environs

67

general interest being in minor transformations that remain quite close to the text. The relation between what is obligatory and what is optional is not always spelled out, but excursions beyond the domain of ‘servitude’, while certainly creative, do not go far. In all, Barkhudarov sounds like a teacher anxious to train translators to stay within a fairly narrow band of solutions, perhaps in the interests of collective reliability. This restrictive view informs what Barkhudarov has to say about Compensation, listed as one type of ‘substitution’. The term here means more or less what it means in Vinay and Darbelnet and most of the examples concern how to express regional qualities or colloquial registers in different parts of the text, as indeed is the case in Vinay and Darbelnet. Barkhudarov nevertheless does not insist on Compensation being on a different level or location, as some other usages do. He finds a nice example in Somerset Maugham: ST: ‘Why don’t you write a good thrilling detective story?’ – ‘Me?’, exclaimed Mrs Albert Forrester, for the first time in her life regardless of grammar. TT: – А почему бы вам не написать детективный роман, такой, чтобы дух захватывало? – Чего? – воскликнула миссис Форрестер, впервые в жизни забыв о грамматике. [‘Why do you not write a thrilling detective novel?’ – ‘What?’, exclaimed Mrs Forrester, for the first time in her life forgetting about grammar.] (1975: 219–20, italics mine) Here Mrs Forrester’s use of ‘Me?’ (instead of the hypercorrect ‘I’) is a very minor grammatical deviance, making the narrator’s commentary mildly amusing. In the Russian this is captured by the colloquial form ‘Чего?’ (‘What?’), instead of the more formal ‘что’. In Vinay and Darbelnet, the solution would perhaps not count as Compensation because it occurs in the same place and on the same linguistic level as in the start text; it would more easily fall into the category of Équivalence, where two different expressions (like idioms) are formally different but fulfil the same function, or it might be seen as Adaptation. The surprising thing is that Barkhudarov appears to have no separate categories operating on those two levels, so they can all be poured into Compensation. In sum, the kind of radically creative solution that Vinay and Darbelnet encouraged, in the name of preserving the spirit of each language, seems to be of no particular interest to Barkhudarov. The structures of good Russian must be maintained, and the science of their maintenance was kept close to linguistics.

68

Translation Solutions for Many Languages

Postscript 2: Total system in Estonia In 1995 the Estonian translation theorist Peeter Torop produced what appears to be the most systemic and elegant of typologies, combining some basic terms from the Russian tradition of translation theory with the appeal to system developed within Tartu-school semiotics (basically after Lotman). His typology claims to provide a metalanguage for all kinds of translation, between all kinds of semiotic systems (music, image, dance, etc.) and for all kinds of texts (1995/2000: 95), which means that his terms and concepts are rather different from the traditional grammar-based categories. Torop’s work has been championed by the translator and theorist Bruno Osimo, whose Italian translation of Total’nyi perevod (Total Translation) (1995/2000) I am following here, along with later commentaries by the same mediator (2008). Torop’s model (Table 3.2) is actually of ‘types of translation’, which in theory can concern a general approach to the whole start text, developed for the sake of categorizing texts rather than training anyone how to translate. Here I am going somewhat against the grain, seeing it as a top-down approach to the problem of organizing translation solutions, basically because I am interested in the way the upper-level categories are used. Torop is recognizably within the tradition of Fedorov when he allows that there is just one kind of translation, called ‘adequate’, and that this one type aspires to the full value of the text. His division of adequacy into ‘recoding’ (mapping onto a new code) and ‘transposition’ (content recast as new content) then follows more or less the conceptual geometry that we have traced from Smirnov (1934) (see Table 3.1) and it might indeed be retroactively mapped back onto the Vinay and Darbelnet distinction between ‘direct’ and ‘oblique’ translation. So far, so good: the top two rows of Table 3.2 are within the Russian tradition and remain compatible with a fairly traditional model.25 And then, if you look at the bottom line of Table 3.2, it is possible to see a progression from narrow constraints to relative freedom, in fact giving a series of answers to Sobolev’s 1950 question about what degree of exactitude is required in translation (not that Torop cites Sobolev at all). So how does Torop actually get from the traditional categories to the detailed descriptors of degrees of exactitude? He uses two distinctions, both of which are rather more complicated than they seem: 1. There are two phases of the translation process: ‘analysis’ refers to the translator processing the start text; ‘synthesis’ is the phase when



A Tradition in Russian and Environs

69

the target text is assembled for its particular reader and purpose. Each of these phases can be dominant within each of the two types of translation. 2. There are two ways of approaching what is to be translated: ‘dominantcentred’ focuses on the language plane that is dominant in the start text, while ‘autonomous’ can change that focus, rendering just one plane or another of the start text – for example, by rendering verse as prose (Osimo 2008: chapter 21). Applying these two binary divisions in conjunction with the others, we get a neat set of eight translation types (Table 3.2). Table 3.2  Torop’s model of translation solutions Adequate translation Recoding

Transposition

Analysis

Synthesis

Analysis

Synthesis

Dominant-centred Autonomous Macro-stylistic Precise

Dominant-centred Autonomous Micro-stylistic Citational

Dominant-centred Autonomous Thematic Descriptive

Dominant-centred Autonomous Expressive Free

Source: Torop 1995/2000: 104; Osimo 2008: chapter 21.

Each of the eight types comes with examples from the various modes of translating poetry (Torop 1995/2000: 102–3): Macrostylistic: ‘Pure analytical recoding’; the translation is characterized by ‘exact observation of the rhyme and verse structure’. Precise: ‘In the start text the plane of expression is dominant and the translation is limited to this’, as in interlinear versions. Microstylistic: ‘the main goal is to recreate the author’s expressive means’, as in ‘exoticizing and localizing translations’. Citational: ‘the aim is to obtain lexical precision within formal limits’, as in ‘literal translation’. Thematic: the plane of expression is subordinated to the plane of content, allowing such things as the translation of strict verse as vers libre. Descriptive: ‘a conscious rejection of the work’s integrity’, such as a prose translation of verse. Expressive (or ‘receptive’): ‘the reconstruction of the content in order to maintain the expressivity’, as in Nida’s ‘dynamic equivalence’.

70

Translation Solutions for Many Languages

Free: ‘free interpretation of the plane of content’, in the style of ‘liberally adapted from …’. Most of these descriptions are actually quite hard to separate from each other, in both name and content, and no textual examples are given to guide the perplexed. In any case, the whole house seems to have been built on moving sands. For example, the distinction between ‘analysis’ and ‘synthesis’ is drawn from ideas about the translation process, but there are no data and no analyses of any actual process (only a discussion of theories), so one is left with the strange idea that, in some cases, the translator only got half-way through (and is thus stuck on ‘analysis’). In effect, this is an idealistic binary division between startoriented and target-oriented, which is undone by most close analyses of actual translation processes. The second problematic distinction relies on Jakobson’s concept of ‘the dominant’, defined as a ‘focusing component of a work of art: it rules, determines and transforms the remaining components’ (Jakobson 1935/81: 751). The dominant can be different in each text, so it cannot be identified with any particular linguistic category. This means that it is left precariously open to interpretation: one translator might produce an exact ‘macrostylistic’ translation for a text whose dominant seems to be the form, while a second translator might work into vers libre or prose, arguing that the dominant is in the content of the start text. And who is to say that either translator is wrong? In fact, the closer you look at the descriptions, the less clear they become. The hugely idealistic reliance on binary distinctions, which promise to make things so neat, produces little but a set of difficult terms that seem not to speak to any divisions that are operative in practice, neither socially nor cognitively.26 Not surprisingly, this metalanguage for ‘types of translation’ has found few adepts, perhaps in part due to the historical failure of semiotics to gain ascendency as a discipline on the international stage, but perhaps also because of an inherent lack of binary divisions in the actual gamut of translation solutions, which are simply not happy when boxed in like this. As it stands, Torop remains a brave attempt at top-down systematization. His work nevertheless runs into problems that also plagued other adepts of systemic organization, as we shall see when we meet a few Germans.



A Tradition in Russian and Environs

71

Postscript 3: A postcard from Ukraine I close with a brief counter-example from Ukrainian. Gorbachov (2013), writing in English, purports to apply the typology from Vyacheslav Karaban’s Переклад англійської наукової і технічної літератури (Translation of English Scientific and Technical Literature) (2004) to an English translation of the Ukrainian Constitution. Here I am interested in the mess that results when the categories are rendered into English. The translation solutions are seen as ‘grammatical transformations’, of which there are four basic types: Replacement, Transposition, Addition and Omission. The descriptions and examples given by Gorbachov, who also cites Miram et al. (2005), include the following. Replacement: ‘any change in the target text at the morphological, lexical, or syntactic levels’ (2013: 29): ST: за ознаками раси [according to-PREP features-N.INSTR race-N.GEN] TT: based on race Here the preposition ‘за’ is rendered by the phrasal verb ‘based on’. If we can put aside our doubts about the linguistic categories, it seems clear enough that the category of Replacement is what Vinay and Darbelnet called Transposition. At the same time, Gorbachov cites cases of resegmentation (the splitting and joining of sentences) as examples of Replacement, so it is hard to say what is going on. Addition: ‘a device intended to compensate for structural elements implicitly present in the source text or paradigm forms that are missing in the target language’ (2013: 30): ST: Green Party federal election money TT: на вибори на федеральному рівні [for-PREP election-N.ACC.PL on-PREP federal-ADJ level-N.LOC] This is clearly what Vinay and Darbelnet call Explicitation, unpacking the noun phrase so as to indicate semantic relations (‘на’, for, on) and location (‘рівні’, level). Transposition: ‘a change in the order of the syntactic elements or words in the target sentence as compared to that of the source sentence’ (2013: 30). As noted above, a pure change in the order of words and phrases seems to be missing

72

Translation Solutions for Many Languages

from Vinay and Darbelnet. In any case, this sense of ‘transposition’ is clearly not the one used by Vinay and Darbelnet, even though it has tons of etymological justification (‘trans-position’ should probably be a movement between positions). Omission: ‘the reduction of the elements of the source text that are considered redundant from the viewpoint of the target language’s structural patterns and stylistics’ (2013: 31): ST: Причини виникнення війни [Reasons-N.NOM occurrence-N.GEN war-N.GEN] TT: Causes of war The abstract noun for ‘occurrence’ has been made implicit here, and this Omission is clearly what Vinay and Darbelnet call Implicitation. In summary, if read from the perspective of Vinay and Darbelnet, Transposition is here called Replacement, word-order change is Transposition, Explicitation is Addition, Implicitation is Omission and they are all ‘transformations’. This is not simply confusion – it is the trace of a different tradition, struggling to translate itself.

4

A Loh Road to China

Chapter summary: In 1958, the same year as Vinay and Darbelnet, Loh Dianyang published a translation textbook in Beijing, with a rather similar list of translation solutions. The link with Western thought had come from the Chinese translation of the Russian Fedorov’s 1953 book, tempered with studies in English rhetoric. Loh’s solution types differ from Vinay and Darbelnet’s in that they account for different writing systems, consecrate mistranslations (such was the power of authority), and Repetition is listed as a way of achieving naturalness. Loh’s general categories (albeit without Repetition) underlie much of the subsequent Chinese work on solution types, although his general approach has been attacked by nationalist theorists who argue that Chinese is so different from Western languages that no Western-inspired categories can be applied to translation into or from it.

Any attempt to adapt Vinay and Darbelnet to English–Chinese is necessarily preceded by Chinese categorizations of very similar solution types. The earliest of these seems to have been by Loh Dian-yang1 (or Lu Dianyang, 陆殿扬), whose two-volume Translation: Its Principles and Techniques (英汉翻译理论与技巧) was published in English in 1958, coincidentally the same year as the first edition of Vinay and Darbelnet. Loh’s textbook was picked up by later generations of scholars in China, especially through the work of Zhang et al. (1980), and thereby significantly shaped the way translation solutions are organized in contemporary Chinese training programmes. It is also a convenient point of reference for me because it was published in English, indeed in excellent English, in Beijing and in 1958. Loh’s typology is strikingly similar to Vinay and Darbelnet’s in approach and aims (the two have been compared in Zhang and Pan 2009). Indeed, one would suspect an influence of some kind, or at least the trace of common forerunners. Loh’s work nevertheless makes no mention of Vinay or Darbelnet, nor does it

74

Translation Solutions for Many Languages

betray any knowledge of French – there are no references to Saussure, Bally or Malblanc, for example. Given the intellectual climate of Communist China in the 1950s, one might more readily suspect influence from Russian – and Andrey Fedorov is indeed the only contemporary translation scholar cited by Loh.

Dian-yang who? Since Loh is relatively unknown even among contemporary translation scholars in China, I reproduce parts of an official biography:2 Loh Dian-yang (1891–1972) showed his extraordinary intelligence while young: at the age of 11 he won first place in an examination organized by Mr. Zhang Yipeng, an advocate of Western learning, for all students under 15 in greater Shanghai. Loh was then recommended for admission to Shanghai Nanyang Public School, affiliated with the predecessor of National Jiaotong University. Upon graduating from the public school at the age of 19, he became a highschool teacher of English. At 29 he became Principal of Jiangsu Provincial No.1 Middle School and a professor at Southeast University. In 1928 he served in the Education Bureau in Hangzhou, Shanghai, as head, general secretary and other posts, winning the favour of Chen Bulei, who was Director of Education in Zhejiang province, and Chen Li-fu, who would later become Minister of Education of the Republic of China. In 1935 Loh entered the Zhengzhong Book Bureau, which had been founded by the Kuomintang government to strengthen their control over culture and education … Loh Dian-yang gradually rose to the position of Senior Editor at the bureau. In Nanjing he spent 20,000 yuan (at that time a big barrel of rice was only five or six yuan) on building a three-floor house with a garage, boiler room and reception room in the Western style. Loh attached great importance to the training and education of his children, and the family home had its own chemistry laboratory. Mao Yanwen, second wife of the Premier of the central government of Republic of China Xiong Xiling, was invited to teach the children English. Loh was the author of Practical English Rhetoric, Translation: its principles and techniques, Geography for Junior Middle School Students, English for Junior Middle School Students, How to speak English, A Handbook of English Public Speaking, A Handbook of Practical English Rhetoric, A Comparison of Word Order between English and Chinese, among others.

The last-mentioned book, on comparative word order, appears to have been published in 1958, the same year as the translation method. Some paragraphs further on, we read:3



A Loh Road to China

75

In 1933 Loh Dian-yang joined the Kuomintang. After the victory [of the Communist Party in 1949] he left the Kuomintang and entered what would become the PLA [People’s Liberation Army] University of Foreign Languages, where he was Professor and Chair of the English Department, with a third-class merit citation from the PLA. At this time [during the Cultural Revolution] he was coerced into declaring that he was a spy for the Kuomintang.

Loh’s teaching at the military university was significant in more than political terms, if only because his years there would seem to have overlapped with Zhang Peiji’s time as a much younger member of the same faculty.4 Zhang was to take over many of Loh’s translation categories (in Zhang et al. 1980) and thus ensure their lasting influence. So Loh was a brilliant student (his written English is truly brilliant, to judge by his books) who worked as a foreign-language teacher, school principal and then top government bureaucrat in the Kuomintang regime, accumulating considerable wealth. With the defeat of Chiang Kai-shek in 1949, he managed to move to a top university post, working for the Chinese army. Not surprisingly, he paid more than lip service to contemporary politics in his 1958 book, where we find quite a few translation examples like ‘Long live the camp of peace headed by the Soviet Union!’ (1958: 1.170; cf. Tao 2005). There were also consequences concerning the translation principles themselves, as we shall soon see. Those were turbulent times; Loh would appear to have ridden out several storms. The Cultural Revolution did not treat him kindly, however. Not only was he forced to denounce his past, as noted above, but his daughter Lu Lanxiu, a heroic opponent of the Cultural Revolution, was paraded through the streets and shot by firing squad in 1970. Loh died two years later.

Asymmetric power and totalitarian translation policies The relation between politics and translation textbooks is not just a matter of a few politically correct examples. Much also depends on national language policy and the power relations between languages. I pause to consider this. For as long as comparative stylistics in the first half of the twentieth century was dealing with French, German and English – and these were the languages for which the solution types were formalized – everyone could comfortably assume the languages were of about the same size and shared a similar degree of complexity. Each had its systemic tendencies or ‘personality’; each could expect to see those identity markers preserved. The underlying claim of Bally’s stylistics

76

Translation Solutions for Many Languages

was not just that such identities existed, but also that they should not change – the languages had little reason to borrow major expressive resources from each other. After all, those languages had expressed industrialization, science, class mobility, colonialism and a generalized modernity. They could look at each other with mutual respect or rivalry. Any comparative stylistics involving Chinese, however, or pre-Soviet Russian, could not enjoy the luxury of assuming such a balance of power. When technologies and new ideas were moving from centres to peripheries, the general assumption was not of equal or rival languages, but of dynamic asymmetries and imbalances of power, with some languages being weaker and others stronger. In fact, in world translation history, the assumption of asymmetry is more likely to be the general case, with apparent equality only being assumed, after the Renaissance, between the larger national languages of European print culture. For the rest, languages are unequally endowed and translators move not only knowledge and information but also expressive resources. In other words, in situations of asymmetry, translation is likely to make the stronger language alter the established structures of the weaker one. One of the fundamental questions facing modernist language policy is whether this should happen. Should a weaker language accept outside resources (‘influences’, ‘interference’, ‘contamination’ – none of the terms are neutral)? Or should the language be maintained and cultivated on its own terms, generating new forms from within, as with the expression of collective identity, resisting or recasting resources from the outside? The comparative stylistics developed between French, German and English are clear on this point: translations should respect the established structures of each language. That is why there are so many categories for complex transformations. The linguists wanted not just to describe language systems, but also to make sure that the differences between languages remained intact. The linguists, however, were not the only ones to opine. It is a seldomremarked fact that three of the great totalitarian leaders of the twentieth century – Hitler, Stalin and Mao – all had something to say about relations between languages, and thus indirectly about translation. We have seen Stalin’s statements of 1950, where ‘the functional role of language, as a means of intercourse between people, consists not in serving one class to the detriment of other classes, but in equally serving the entire society’ (1950/54: 5). What that means for translation is not immediately clear, but it might assume significance in the light of similar statements by other leaders. Adolf Hitler, in a Führererlass of 1940, declared himself in favour of importing foreign linguistic forms, since the German language had to be developed so as



A Loh Road to China

77

to become a world language (cf. Von Polenz 1967/79). This is logical enough: National Socialism needed nationalism, but not so much as to isolate its language from the advances of an international modernity. Hitler was arguing against linguistic nationalists who wanted German to develop all terms from within its own expressive resources. That argument seems to be working opposite to Stalin’s: German should be allowed to change from the outside, whereas the Russian language should probably stay more or less as it is; there is no need to change it radically just because the relations of production have changed. On another level, however, both Hitler and Stalin are saying much the same thing: there is no direct social determination of a language, so the radical reformers, the shapers of ideologies, should not mess with it. A descriptive linguistic approach to translation thus became possible, as we have seen in Fedorov. And Mao Zedong? In 1956 he commented on a long-standing debate about whether cultural forms should remain national or could import from abroad: National forms may incorporate some foreign elements. There’s surely no need to write novels in the old style with each chapter headed by a couplet giving the gist. But the language and presentation should be Chinese. Lu Xun was for a national style. But he also advocated very close translation. Personally, I’m for very close translation of theoretical works because it has the advantage of accuracy. We should be very clear on this fundamental principle: It is also necessary to learn basic principles from the West. To insist that the scalpel must be in the Chinese style is absurd. (1956/91: 102; italics mine)5

It is the same argument, basically: do not be excessively nationalist or classbased; be prepared to bring in cultural products from the outside; incorporate modernity. The reference to the scalpel is intriguing – when your health is in play, you want the best-shaped scalpel possible, no matter where it is from. I am reminded of the Spanish Cardinal Cisneros, who after 1492 went to Granada and destroyed between four and five thousand Islamic books in order to impose Christianity (totalitarian cultural policies existed well before the twentieth century), but he kept thirty or forty medical treatises in Arabic – even the most extremist ideologues do not take risks with their health (cf. Vallejo 1913: 35). But I digress. Hitler, Stalin and Mao would all be saying, in different ways, that a language is not, and need not be, determined by the people that use it. As totalitarian leaders, they felt they had every right to say what should be done with language, since a ‘totalitarian’ discourse is precisely one that presumes to address all

78

Translation Solutions for Many Languages

aspects of social life (cf. Faye 1972). At the same time, they wilfully created these gaps in their own imagined control: a language is not to be considered just one further element in the totally controlled social system; it can and should receive resources from the outside. To do otherwise would mean looking backward and risk missing the rewards of modernity. I hasten to recall that Charles Bally made much the same argument at the beginnings of his stylistics: a language, for him, was not to be considered the direct expression of a people; it demanded to be studied in and of itself. This may explain why comparative stylistics was applied to translation in the years directly following the great leaders’ pronouncements: Malblanc 1944 followed Hitler 1940; Fedorov 1953 followed Stalin 1950; Loh 1958 followed Mao 1956. Is this just fancy footwork with the dates? Not entirely: Fedorov does discuss Stalin’s articles at some length; Loh makes significant reference to Mao; and Malblanc, well no, he does not mention Hitler and he does tend to believe in more of a Volkspsychologie. But the dates are there anyway. Just as Stalin intervened in a long-standing disagreement among linguists, so Mao was explicitly commenting on a long debate between translators.

Chinese translation theories prior to Loh If you ask about Chinese translation theory since the nineteenth century, the first name people mention is Yan Fu and his ‘three requirements difficult to fulfill: faithfulness, comprehensibility and elegance’ (译事三难：信达雅) (1901/2004: 69).6 More important than the principles, in the present context, is the translation practice that went with them. The following is a passage from Yan Fu’s version of Huxley’s essay ‘Evolution and Ethics’, pertinent here because the three ‘difficulties’ were expressed in the preface to the translation. Here I take the example from Luo (2003: 204): first we have Huxley, then Yan Fu, then the back-translation, perhaps generous in word choice, provided by Luo: It may be safely assumed that, two thousands [sic] years ago, before Caesar set feet in Southern Britain, the whole country-side visible from the windows of the room in which I write, was in what is called ‘the state of nature’. Except, it may be, by raising a few sepulchral mounds, such as those which still, here and there, break the flowing contours of the downs, man’s hands had made no mark upon it; and the thin veil of vegetation which overspread the broad-backed heights and the shelving sides of the combs was unaffected by his industry.



A Loh Road to China

79

赫胥黎独处一室之中，在英伦之南，背山而面野。槛外诸境， 历历如 在几下。乃悬想二千年前，当罗马大将恺彻未到此时，此间有何景物。 计惟有天造草昧，人工未施，其借征人境者，不过几处荒坟，散见坡陀 起伏间。而灌木丛林，蒙茸山麓, 未经删治如今日者， 则无疑也。 [Huxley sat alone in his study, which was situated in southern England. In front of the house was a stretch of land and behind it a mountain. The whole countryside was visible in a natural state from the window. Huxley just could not help thinking about what it was like two thousand years ago when Caesar, the Roman general, had not yet set foot on the land. Unlike today, it must have been a land of burial mounds, a thin veil of vegetation spreading across the broad-backed heights and the shelving sides of the ancient combs.]

Whatever ‘elegance’ and ‘expressiveness’ meant in theory, Yan Fu’s practice involved a great deal of work: here he changes from the first person to the third (apparently a measure of ‘elegance’); he removes a good number of circumstantials; he explicitates Caesar; he omits the final reference to ‘industry’. The point is that Yan Fu’s translating was a long way from what ‘translation’ means nowadays in China. When I give my Chinese students the above passage to translate, they deploy none of the adaptations, omissions and explicitations made by Yan Fu; they follow the text in a much closer way, with as few deviations as possible. And when I show them Yan Fu, they laugh, since the Chinese is old-fashioned. That might be where modernist theorizing was starting from. Yan Fu wanted to bring in foreign knowledge, but he did not particularly want to bring in foreign expressive resources. If we move forward to the May Fourth Movement, initiated by student demonstrations in Beijing on 4 May 1919, we find a strong anti-imperialist discourse that wavers between the rejection of Confucian values, the regeneration of national culture and the rejection of outside influence. In terms of translation theory, the most cited figure is Lu Xun, who reacted against the kind of discursive freedom we find in Yan Fu, calling for (and practising) a form of close literalism that he called ‘stiff translation’ (yingyi). That literalism was even at the risk of incomprehensibility, violating normal Chinese word order (Chan 2001: 202) and attempting to produce a new kind of Chinese for educated readers. Chan (2001: 206) cites and translates a 1931 letter by Lu Xun: Neither Chinese speech nor writing is precise enough in its manner of expression. The key to good writing is to avoid clichés and empty words … To cure this ailment, I believe we have to do it the hard way and seek to render thought in wayward syntactical structures. (1931/2004: 159)

80

Translation Solutions for Many Languages

Although Lu Xun’s general aim is often described as a ‘Europeanization’ of Chinese, his actual claims on this point are interestingly mixed: change can come from the past, the regional and the foreign. Understandably, there were many arguments for and against this extreme position; there were calls for a kind of Chinese that was accessible to the new wider readerships, that would enable the development of an inclusive national culture. What interests me here is that, as Chan (2001) sees very clearly, the debate itself, both for and against literalism, was thoroughly marked by modernity on both sides. The arguments were not just about how to represent a foreign text, or how to write Chinese: they were about how to bring modernity to a whole national culture, particularly with respect to education. So when Mao Zedong refers to Lu Xun in 1956, he is recognizing a whole previous debate for and against literalism, a debate in which no one was anywhere near the extreme freedoms seen in the above example from Yan Fu. There were basically disagreements about what should be done within the structure of the sentence. It is against that background that a Chinese professor of English picked up a Russian linguistic theory of ‘adequate translation’ and noted that Fedorov’s ideal united the opposites of being both ‘faithful’ and ‘smooth’. Loh does not link this with the ‘still Chinese with the advantage of accuracy’ noted by Mao, but he does see it in terms of Mao’s theory of progress through contradiction, with the corresponding call to ‘let a hundred flowers blossom’ (Loh 1958: 1.16): If the development of translation work proceeds by way of the disclosure of internal contradictions and by way of free discussion in the translation field until the question of right and wrong is settled [Loh elsewhere cites Stalin’s intervention into linguistics], then it is clear we shall come to a satisfactory conclusion: ‘Unity of contradictions’, i.e. a translation both faithful and smooth. (1958: 1.16)

Thus, apparently under the illusion that contradictions would not be solved by anything like the Cultural Revolution, Loh worked within the dialectic and started to write down some solution types. So where was he coming from in terms of stylistics?



A Loh Road to China

81

Comparison in search of influence Much of Fedorov’s Введение в теорию перевода (1953) was translated into Chinese in 1955 under the title of 翻译理论概要 (Introduction to the Theory of Translation). This was just four years prior to Loh, and we have seen that Fedorov is indeed referred to in Loh. Apart from Fedorov, the only other translation scholar Loh refers to (negatively) is Tytler, albeit in the company of a very strong team of Western grammarians: Leonard Bloomfield, William Entwistle, Otto Jespersen, Henry Sweet and Zellig Harris. The weight of those names would suggest that much of Loh’s thinking about language was indeed coming from English-language theorists and the grammarians of English, rather than from Stalin, Mao or any particular scholar of translation. Other evidence suggests that Loh had a long-term concern with English rhetoric and thus probably with the comparative stylistics of English and Chinese. In the 1920s he authored a booklet with the title Thirty-Two Questions on Practical Rhetoric, published in English by the Kiangsu (Changzhou) First Middle School, Nanking. The second edition was dated 1924, a full thirty-four years prior to Loh’s textbook on translation.7 Was this the same Loh Dian-yang? The textual evidence suggests so. For example, the 1924 booklet answers questions about how to give ‘life’ to a text – the main answer is that you must use repetition – while in the 1958 textbook (volume 2) we find tips on ‘How to make a translation full of life’ – where the main answer is once again that you should use repetition. Note that in both works Loh consistently refers to ‘rhetoric’, in the English tradition leading back to the study of Greek and Latin, rather than to ‘stylistics’, as in the French tradition. Yet a similar concern with liveliness is present: Loh’s is not a linguistics of abstract systems and applied rules. By all accounts, Loh was a serious scholar with a long-term personal interest in rhetoric, particularly of English. He probably had no need to borrow heavily from Russian. At least two features of Loh’s work might nevertheless be attributed directly to Fedorov.8 The first is Loh’s reference to the overarching principle called ‘adequacy’, which he explicitly takes from Fedorov: we have learned much from A. B. Feedorov [sic], whose ‘Principles of Translation’ has, indeed, supplied us a sound system of theory on translation. Adequate translation, as suggested by the Soviet scholar, is, in fact, a translation both faithful and smooth. (Loh 1958: 1.16)

82

Translation Solutions for Many Languages

Adequate translations are certainly not word for word, but they might come close to it: ‘faithful to the original text and yet expressive in the second language’ (1958: 2.77). This notion of something combining both form and content is a key component of Loh’s approach – it is the baseline from which all other ‘ways’ are taken when necessary. The second influence from Fedorov then comes in a wonderful section where Loh’s header declares: ‘Translation Method Varies with the Quality of the Original Work’ (1958: 2.88). Here Loh names three kinds of prose texts, which are recognizably the three formulated by Fedorov (1953): 1. News reports, documents and scientific treatises 2. Political essays 3. Literary productions Each of these types corresponds to a ‘method’ of translation. In dealing with expository prose (category 1), ‘we should make efforts to reproduce exactly what is said in the original’. When translating political essays (category 2), ‘the translator should be faithful not only to the contents of the original but also to its purpose’. And in literary texts (category 3), the translator should preserve ‘all or nearly all the artistic qualities of the author’ (1958: 2.88–9). These points concord with Fedorov, and perhaps for similar reasons. In historical context, Loh, like Fedorov, was allowing that the purposes of the Communist Party could require that certain features of ‘political essays’ be altered. This was a very specific forerunner of Skopos theory. The more substantial question, for my current purposes, is to what extent the Russian influence might have extended into Loh’s actual taxonomy of translation solutions.

Loh’s principles That Loh was not borrowing from French should be fairly clear from a comparison of his categories with those of Vinay and Darbelnet (Table 4.1). At the same time, if we try to see his work as a conceptual extension of the Russian tradition (Table 4.2), we actually find that he was saying many new things.



A Loh Road to China

83

Table 4.1  Possible alignment of categories from Vinay and Darbelnet (1958) and Loh (1958) (cf. Zhang and Pan 2009: 366) Vinay and Darbelnet (1958)

Loh (1958)

Borrowing

Transliteration Symbolic Translation Coinage of New Characters

Calque

Semantic Translation

Literal Translation [Prosodic effects: Amplification/Reduction, Explicitation/Implicitation, Generalization/Particularization, Compensation]

Omission Amplification Repetition

Transposition

Conversion

Modulation

Inversion Negation

Correspondence or Reformulation (équivalence) Adaptation

Table 4.2  Possible correlations between Russian typologies and Loh (1958) Smirnov (1934)

Retsker (1950)

Fedorov (1953)

Loh (1958)

Adequacy

Equivalence

Equivalence

Transliteration Build from components

Transliteration Semantic Translation Symbolic Translation New Characters

Substitution

Analogue correspondence

Similar function

Substitution

[Permutation Grammatical Restructuring Modification Adaptation]

Concretization Development Antonyms Compensation

? Omission Amplification Repetition Conversion Inversion Negation

Since one of my purposes is to see how the different traditions can be brought together, I will compare Loh’s categories within both these frames. Here I run through Loh’s solution types, grouped in accordance with his own larger categories.

‘Nouns denoting things of foreign origin’ I first focus on just one problem: how can the translator render foreign terms for which no prior equivalent exists in the target language? We have seen that

84

Translation Solutions for Many Languages

Vinay and Darbelnet (1958) propose just two ‘procedures’ for this (Loan and Calque), while Fedorov (1953) offers three. Loh’s treatment of the same problem puts forward four main solution types: Transliteration, Semantic Translation, Symbolic Translation and Coinage of New Characters, in addition to following established convention. Further, these four can work in combination (thus giving five types in Loh’s actual presentation):9 Transliteration: chocolate 巧克力; Washington华盛顿 (Putonghua: Hua-sheng-dun) Semantic Translation: airplane飞机 Transliteration plus Semantic Translation: utopia 乌托邦 Symbolic Translation with Semantic Explanation: cross 十字架 Coinage of New Characters: oxygen 氧; pump 泵

Why this growth in the number of terms?10 The simple categories probably correspond in all three typologies: Simple Transfer is Borrowing is Loan is Transliteration, and Building from Particles is Calque is Semantic Translation, at least on the basis of the examples. The fact that Chinese uses characters would then give the added possibilities of Symbolic Translation and Coinage of New Characters, which are logical enough. And as for the combinations, it seems generally recognized that all these things can always be combined, in all the languages under consideration, suggesting that the compound categories are not in fact substantial novelties. So are there really just two categories here, plus two special ones just for Chinese? Perhaps not. There are at least three problems: 1. Transcription in the same or different script? One could imagine a choice has to be made in every translation operation that involves different scripts. For these problems, there is always the possibility of using the start-language script, for example by writing ‘McDonald’s’ or ‘KFC’ in China, using Latin letters – a possibility that seems not to have been contemplated in Loh’s China of the 1950s. In fact, what Loh’s examples indicate here is the copying of sounds, as distinct from copying script. 2. New or old solutions? When my Chinese students look at Loh’s examples today, many express surprise that some of the items should be seen as translation problems at all. What was difficult to render in 1958 has now been solved by reference to established equivalents. This distinction between old and new solutions is missing from the above classifications



A Loh Road to China

85

(although Loh does recognize old mistranslations, as we shall soon see). One might suppose that Loh’s whole initial problematic here, when he refers to ‘nouns denoting things of foreign origin’, presupposes novelty and hence the constant need to choose, just as Vinay and Darbelnet’s solutions would similarly only be worth mentioning for as long as items like bulldozer in French and fuselage in English can be seen as new or recent borrowings. The newness of oxygen as 氧 (or indeed of Sauerstoff, ‘sour stuff ’, in German) is no longer absolutely new. With time, the successful borrowings lose that newness and there is no longer any special translation problem to solve – just an old one to look back on.   The creation of new characters was clearly a necessary solution in the period when modernity was being introduced into Chinese. Yet the same solution can serve postmodern purposes as well: Dai Congrong (2010) mentions the creation of new characters as one of the ways she has translated Joyce’s Finnegans Wake into Chinese, since it allows a mash-up of meanings in much the same way as Joyce produced his portmanteau words. (Dai’s main translation solution is nevertheless a huge set of translator’s notes.) 3. Repeating old mistakes? As mentioned, Loh opens space for ‘established mistranslations’. Thus ‘Soviet’ is 苏维埃 (Putonghua pronunciation: Su-weiai), no matter how much better the transliteration could have been. And ‘Eden’ as 艾登 (Putonghua: Ai-deng) might even count as an example of established mistransliteration. Such approximations nevertheless allow terms to travel from language to language. Loh gives the example of the tropical cyclone in the region of the Philippines or the South China Sea that is called ‘tai-fung’ in Cantonese, literally meaning ‘great wind’ (大风), which in English becomes ‘typhoon’. This English word was then transliterated into Chinese as 台风 (Putonghua: tai-feng). Loh’s position in such matters is intrinsically conservative: one must accept what tradition gives us, including the Chinese term that comes back to Chinese from English. (The Chinese respect their elders.)   The question of relative novelty and historical correctness should also raise doubts about the categories for which Loh (and others) distinguish between changes that are obligatory and those that are optional. In principle, the new problems are the ones that allow several solutions (and are thus truly translation problems), even though with time the solutions become more obligatory. Or more exactly, the obligatory structures are the ones that have had time enough to gain dominance and authority.

86

Translation Solutions for Many Languages

On this view, it seems superfluous to make any systematic distinction between obligatory and optional solutions. After all, the kinds of problems translators need real help with (assuming they have learned languages) are always the optional ones. In summary, translation into Chinese certainly requires more categories for dealing with ‘things of foreign origin’, but perhaps not as many as Loh proposes – the typologies are most useful for solving new problems, those for which no equivalents are yet established.

A missing literalism? As we saw a few chapters ago, Vinay and Darbelnet list Literal Translation as a procedure, referring to situations where the syntax is not altered: The ink is on the table becomes L’encre est sur la table (article, noun, verb, preposition, article, noun, in both cases). Loh does not offer any category on this level. Much as one might assume that this is because Chinese syntax is completely different from Western languages, if it has a basic syntactic structure it would be Subject-VerbObject, as in English, and a cursory study of contemporary news reports or software prose suggests that it can more or less follow English word order (no doubt because these genres have been permeated by English). So why should there not be something vaguely corresponding to Vinay and Darbelnet’s point of departure in Literal Translation? The answer seems to lie in the way that, standing above all his specific ‘principles of translation’, Loh announces the overarching principle called ‘adequacy’, which we know he takes from Fedorov. Adequate translations are certainly not word for word, yet they might come as close to it: ‘faithful to the original text and yet expressive in the second language’, says Loh (1958: 2.77). Loh’s sentence examples do indeed tend to be as literal as possible, resorting to paraphrase or freer forms of translation only in cases where literalism does not work. When he says, for example, ‘[o]f course, it is sometimes necessary to do away with the mechanical identity in form, before we can attain adequacy in translation’ (1958: 2.79), the mode of thought is that the translator should start close to the given text and only deviate when necessary. This is precisely the mode of thought that could justify Vinay and Darbelnet’s placing of Literal Translation at the centre of their model. So why is something like this ‘adequacy’ not listed among Loh’s main principles? Perhaps it was too obvious to be mentioned. Then again, translation



A Loh Road to China

87

from English into Chinese entails a risk that is not so obvious in work between English and French. If Loh were to enshrine as his basic principle anything that could be interpreted as literalism, he could be seen as condoning English interference in Chinese grammar, and that is certainly not his ideological position. Of course, the condoning of interference was a long way from Vinay and Darbelnet’s intentions as well, but in the case of English and French the historical interference had occurred well in the past, with the invasion of Germanic and then Romance languages in the British Isles – the same word orders are possible because they are from the same language family, at base, with different historical layers imposed one over the other. Vinay and Darbelnet could thus talk about literalism in a way that Loh could not. The years since Loh have seen Chinese accept increasing grammatical interference from English, perhaps in the text genres that are the most translated. The general ideology of comparative stylistics (or rhetoric, in Loh’s terminology), with its supposition of fictitiously equal power relations between languages, is that such interference should ideally be avoided on the level of grammar. It might be necessary on the level of lexical items, since there are indeed ‘things of foreign origin’, but not with respect to the basic structures of a language’s syntax and morphology. Thus, perhaps, exact literal translation is not mentioned as such, even though there are many cases where the same syntactic order is possible in both Chinese and English.

Conversion and Inversion On the other hand, Loh’s textbook is full of sections on the different parts of speech and how they can change in translation. That is, what Vinay and Darbelnet call Transposition is ubiquitous. The name Loh uses for it is actually Conversion, expressed in examples like the following (from a 1951 speech by Stalin, before China became a nuclear power): ST: All peace-loving people demand the prohibition of atomic weapons. TT: 一切爱好和平的人民都要求禁止原子武器。 [All love peace people(s) demand to prohibit atomic weapons.] Here the English noun ‘prohibition’ becomes a verb in Chinese, which would indeed count as Transposition. There remains a problem of definition here. In Vinay and Darbelnet, a change in word class does not necessarily imply a change in word order – many

88

Translation Solutions for Many Languages

of their French–English examples actually maintain the order of the elements. And yet Vinay and Darbelnet also give examples like the following: ST: As timber becomes more valuable TT: Depuis la revalorisation du bois [Since the revalorization of timber] Here the French syntax has TIME LOCATOR, NOUN PHRASE, while the English has TIME LOCATOR, NOUN PHRASE, VERB PHRASE. The order of semantic elements in French is TIME, (NEW) VALUE, [TIMBER], and in English TIME, [TIMBER], TIME, (HIGHER) VALUE. The point is that the French nominalization of ‘become more valuable’ (actually ‘change value’) is not only a change in word class but also, in this case necessarily, a change in the order of elements. As we have seen, however, Vinay and Darbelnet offer no separate category for changes in the simple order of syntactic or semantic elements (which the Russian tradition called Permutation). It is as if such things happened as a matter of course once a change in word class is introduced, which is far from the case. This question is important for two reasons. First, Loh similarly has no separate high-order category for syntactic changes, which is puzzling when one sees phrases like the following: ‘skillful change of the word function and word order in the sentence will help render the version appropriate and intelligible’ (Loh 1958: 2.188; italics mine). Loh was clearly aware that word order would change, and it certainly happens in many of his examples, but it was somehow not worth mentioning. And second, Loh does offer separate categories for just one very particular change in syntax, Inversion, as if all the others were not worth comment. Loh actually has a fairly idiosyncratic take on Inversion. He generally understands it as the use of an inverted syntactic pattern that is marked: ‘the constituent elements of a sentence are arranged in a way that is different from the general rules of word-order of the language in question’ (1958: 2.229; italics mine).11 Since the default word order in Chinese might generally be SUBJECT, VERB, OBJECT, any change in that order could perhaps count as Inversion. Yet the syntax is far freer than that; something else must be meant. Judging by the examples, the notion only seems to concern translation in the sense that markedness should render markedness, as in the following: ST: On the meadow lie several children. TT: 草地上躺着几个孩子。 [Meadow on lie several children.]



A Loh Road to China

89

Here the marked inversion in English corresponds to marked inversion in Chinese. A more complex example is the following: ST: I don’t know his very name, let alone his ability. TT: 我连他的名字都不知道,休说他的能力。 [I even his name do not know, apart from his ability.] Here the focus marker ‘very’ is partly rendered by the underlined object-verb inversion in Chinese, given that the ‘general rule’ is verb-object (as in English), although there is also the focus marker 連 (‘even’). Note, also, that the lexical marker ‘very’ in English can be rendered by syntactic inversion in Chinese, which might make this qualify as a case of Loh’s Conversion. The one example seems to illustrate more than one solution type. The clarity of Loh’s concept is similarly obscured by his allowance that some inversions are ‘necessary’ rather than ‘optional’ (1958: 2.231). If they are ‘necessary’, how can they be marked? The above example (‘his name do not know’) is classified as ‘necessary’, even though the focus is also marked lexically. This is confusing. Of course, Loh may be speaking here as a teacher of English, since he takes pains to point out that English uses inversion in more ways than Chinese, in interrogative and exclamatory utterances. Since the latter are mostly unmarked in English, inversion in such cases would logically be necessary in English but not necessary in Chinese. But surely that is a very obvious grammatical point? Another reason for talking about necessary and optional Inversion could lie in Loh’s attention to what we might call under-determined markedness. For example: ST: I have brought my dictionary with me. TT1: 字典我已带。 [Dictionary I have brought.] TT2: 我已带了字典。 [I have brought dictionary.] Here both translations are possible, with and without Inversion, even though the English sentence does not use inversion and has no focus marker (at least in its written form). Loh prefers the added emphasis of the translation with the inversion (the first one), perhaps because he was thinking as a teacher of rhetoric who simply wants students to use all the expressive resources of their language, to make discourse lively. But in that case, the distinction should not be between necessary and optional, but between different degrees of start-text determinism.

Translation Solutions for Many Languages

90

Yet another reason for Loh’s mention of ‘necessary inversion’ might be found here: ST: He disputes about everything – even about the facts. TT: 他样样要争，连事实也要争。 [He everything disputes about, even the facts also disputes about.] Loh says there is inversion in the second part of the sentence in Chinese, where the marked word order corresponds to the focus marker ‘even’ in English. However, there is also inversion in the first part of the Chinese, where it is presumably a simple syntactic adjustment to what ‘sounds right’ and is thus ‘necessary’. Only the second inversion is rendering markedness (which is also rendered by repetition of the verb). A further explanation for ‘necessary inversion’ comes from the varieties of Chinese in play. In some cases, Inversion is systemically necessary only in Paihua (vernacular Chinese) but not in Wenyan (classical Chinese), and the classical Chinese phrases are still commonly used: ST: I’ll do all I can [do]. TT Paihua: 凡我能做的(事)我都愿做。 [All I can do, I’ll do.] TT Wenyan: 我愿尽力而为。 [I’ll do all I can do.] Here the Paihua necessarily has Inversion but the Wenyan does not. This would justify Loh’s inclusion of ‘necessary inversion’ while retaining the general relation to markedness.

Negation The case of what Vinay and Darbelnet call Modulation (basically a change in perspective) is perhaps more interesting. Loh has a high-order category of Negation, which can be seen as enacting changes in perspective. This interpretation is clear in some of Vinay and Darbelnet’s examples of Modulation: ST: shallow ST: peu profond [not deep] (1958/72: 55) In fact, negation of ‘deep’ is the only way to render ‘shallow’ (as in the shallow end of a swimming pool) in French, Spanish, Italian and no doubt other



A Loh Road to China

91

languages. The depth is the same in all languages, but it is being seen from the opposite perspective. So is this the way negation is being used in Loh? In Loh, the examples of Negation at the lexical level do indeed involve a change of perspective: ST: Yesterday I gave an unprepared speech before a big audience. TT: 昨天我在广大听众之前即席致词。 [Yesterday I big audience before improvise speech.] Here the negative value of being unprepared becomes the positive virtue of being able to improvise (if we can assume the speech worked). There is also grammatical negation: ST: They never work without helping each other. TT: 他们每逢工作，必互相帮助。 [They whenever work, definitely help each other.] The logical change from the double negative to the positive is clear enough and is due, says Loh, to ‘cultural differences’ (1958: 2.253). Loh nevertheless makes no reference to the other kinds of perspective shift mentioned by Vinay and Darbelnet. Then again, if we allow that his categories of Conversion and Inversion involve expressive changes in grammar, one might try to see all such things as shifting perspectives. I note in passing that shifting between positives and negatives is one of the truly useful tips that can be given to trainee translators, quite independently of what sounds natural in a particular language. When you are blocked by a term or a phrase that allows no easy rendition, try translating the opposite of the term or phrase, then apply a negative to that (as in the example of ‘shallow’ above). This was mentioned in Retsker (1950) and has been carried along in the Russian tradition; it is called Antonomy in Chesterman (1997), but I am not sure the term is well recognized.

Categories for achieving naturalness Loh proposes three further principles or ‘ways’: Omission, Amplification and Repetition. These appear to work in opposition: Omission contracts, whereas Amplification and Repetition expand. In Vinay and Darbelnet we find similar terms among the ‘prosodic effects’, where sentences are adjusted in terms of the following oppositions:12

92

Translation Solutions for Many Languages

Amplification/Reduction Explicitation/Implicitation Generalization/Particularization Comparison with Loh’s categories reveals interesting lacunae on both sides. First, Loh’s Omission seems to allow the translator consciously to leave some elements out, whereas Vinay and Darbelnet do not use any such term (‘omission’, ‘deletion’ or indeed ‘addition’ are not in their vocabulary in this sense). This is presumably because there was doubt as to whether such things would still count as translation. Vinay and Darbelnet were not alone with this doubt. Nida, for example, talks about Addition as something a translator could do with a text, but he immediately explains that ‘there has been no actual adding to the semantic content of the message, for these additions consist essentially in making explicit what is implicit in the source-language text’ (1964: 230–1). Similarly, what Nida calls Subtraction apparently ‘does not substantially lessen the information carried by the communication’ (1964: 233). The equivalence paradigm generally does not legitimize cases of outright addition or omission, where the translator need not point to something in the start text as the reason for what is in the target text. As we shall see, an author like Vázquez-Ayora could discuss Paraphrase and Omission as things translators occasionally do, but he issues forceful and repeated warnings against them: ‘To translate does not mean to explain or comment on a text, or to write it as we see fit’ (1977: 288; my translation). So what is Loh’s position? Considered in these terms, Loh’s examples have more to do with Explicitation and Implicitation than with anything else: ST: She covered her face with her hand as if to protect her eyes. TT: 她用手蒙脸，好像去保护眼睛。(1958: 2.98) [She used hand cover face, as if to protect eyes.] Here the double ‘omission’ of ‘her’ can easily be seen as a case of Implication, dictated by the infrequency of such possessives in Chinese. There is no actual deletion of any semantic element. Other examples are not so clear, however: ST: Give him an inch and he will take a mile. TT: 得寸进尺。(1958: 2.107) [Get an inch try for a yard.] In this case we find implicitation of the subject and probably of much more as well: the Chinese four-character idiom (actually more like ‘Get inch, try foot’) condenses what is a whole narrative in English.



A Loh Road to China

93

There should be few surprises here: Implicitation will be used when moving from a relatively hypotactic language to a relatively paratactic one (English tends to spell out the relationships between the sentence elements; Chinese traditionally does not). By the same token, movements in the other direction would tend to be marked by Explicitation, as implicit relations are made clear in English. Is this what we find in Loh? To judge by the examples, Loh’s Amplification does indeed involve Explicitation: ST: 昨天元旦，人人都很高兴。 [Yesterday New Year’s Day, everybody all very happy.] TT: It was New Year’s Day yesterday and everybody felt very happy. (1958: 2.128) Here the copula is implicit in Chinese, explicit in English; the same can be said for ‘felt’. No content is actually being added. For that matter, some of the examples that Vinay and Darbelnet give for what they call Amplification (actually étoffement) could also be seen as Explicitation: ST: the charge against him ST: l’accusation portée contre lui [the charge brought against him] (1958/72: 183) Technically, there is Amplification because there is one extra word in the French. Yet it would be difficult to argue that the verb ‘to bring’ is not implicit in the English. A second example from Loh: ST: There is no more scrambling and starving. TT: 现在不再有争抢，挨饿等事了。(1958: 2.132) [Now no more scrambling and starving types of things.] In this case the addition of ‘types of things’ seems required by Chinese discourse convention. As it stands, this example corresponds quite happily to what Vinay and Darbelnet call Amplification, as in: ST: To the trains ST: Accès aux quais [Access to the platforms] Here we are told that the French preposition à (to) needs the grammatical of the noun meaning ‘access’. The Chinese addition of ‘types of things’ would seem

94

Translation Solutions for Many Languages

to follow a similar reasoning, albeit not governed by absolute grammatical obligation. Loh’s one remaining translation ‘way’ is Repetition. This category finds nothing similar in Vinay and Darbelnet,13 and one is surprised to see it mentioned at such a high rank, if only because English and French rhetoric traditionally eschew repetition of the same (we learn to use variants). Loh’s examples are fairly clear: ST: The custom has [been] handed down generation after generation. TT: 这风俗世世代代传下去。(1958: 2.161) [The custom (has been) handed down era after era and generation after generation.] ST: Nels had it all written out neatly. TT: 纳尔斯把它写得清清楚楚。(1958: 2.162) [Nels had it all written out neatly and clearly.] In these cases, the Chinese seems incomplete without the extra ‘era after era’, which doubles the repetition already found in English. The expansion of ‘neatly’ is also for the sake of greater naturalness, albeit not obligatory (when I tried the example in class, about half of my Chinese Masters students did indeed use the repetition recommended by Loh). Loh’s examples here are into Chinese, suggesting that Chinese employs repetition more than English does. Loh nevertheless recognizes the value of repetition in English discourse as well. So why this prominence of Repetition? After all, it is no more than a particular case of Amplification. And yet Loh insists on it. As noted above, Loh had a long-standing interest in the relation between repetition and the ‘liveliness’ of discourse. Recall that his 1924 booklet claimed that Repetition is the main way one gives ‘life’ to a text, while in 1958 he discusses repetition under the head of ‘How to make a translation full of life’. This contrasts interestingly with Lu Xun’s ‘stiff translation’ and the position that ‘[t]he key to good writing is to avoid clichés and empty words’ (Lu Xun 1931/2004: 159). Where the revolutionary wanted translation to purge Chinese of its less precise resources, Loh does everything possible to ensure that those same established patterns enjoy droit de cité. One could go further here, although Loh does not. Repetition is not just a matter of mindlessly repeating the same thing. Without repetition there is no rhythm, and without rhythm the discourse has no aesthetic qualities, little conative functionality and compromised capacity to express social identity and thus motivate a populace. Loh was marked by that very Aristotelian sense of



A Loh Road to China

95

‘rhetoric’, as with the use of discourse to have an effect on people. His insistence on Repetition, as an affirmation of life and interaction, would tend toward the kind of ideas operative in Bally, and thus to the role of stylistics as a counterbalance to literalism. On another reading, this emphasis on rhythm is perhaps saying no more than that the translator should not forget to use the resources of the target language. A similar point is made by the Catalan translator Joaquim Mallafrè (1991), who emphasizes that while accuracy might be a matter for the ‘language of the city’ (llengua de polis), the power to move people comes from the ‘language of the tribe’ (llengua de tribu), with its repertoire of nursery rhymes, sayings, prosodic resources, jokes and otherwise embedded cadences. In English, for example, there is little difference in accuracy between ‘She came home safe’ and ‘She came home safe and sound’, but the latter speaks with a code of shared authenticity. That is a useful and often necessary lesson. Yet it nevertheless seems out of place in a typology of translation solutions. I suspect that the most useful basic lesson here would be ‘feel free to use all the resources of a language’, and that particular maxim only has value when opposed to a view of translation where three words in one language ideally correspond to three words in the other. Seen from this perspective, Repetition has no reason not to be a special case of Amplification. This then allows us to reduce Loh’s ways of producing naturalness to just Contraction and Expansion (or indeed any other terms that might mean making things shorter or longer). It is an unfortunate binarism, but there it is.

Categories missing from Loh I have already noted that Loh has no term for straight copying (where the foreign script is retained) and he somehow fails to list ‘literal’ or ‘adequate translation’ as a solution type, presumably because it was the base from which all other solutions were thought to deviate. Then, on the level of Vinay and Darbelnet’s ‘prosodic effects’, Loh has no formal types for Generalization (the semantic value of the translation is broader than the start-text item) and Particularization (the reverse), and he makes no special mention of anything like Compensation (where a value expressed on one level in the start text is expressed on another in the translation, perhaps in a different place). Such things can nevertheless be found in the examples; they should pose no conceptual problem for Loh’s general mode of thought.

96

Translation Solutions for Many Languages

A more significant absence is the cutting or joining of sentences as a solution type. Chinese–English contrastive linguistics suggests that English is characterized by frequent use of clauses that are longer than is the case in Chinese (Chen 2000: 9), which means that translators into Chinese are wont to cut English sentences into shorter units. Such Resegmentation features explicitly in many of the later typologies in China, starting with the influential work of Zhang Peiji et al. (1980) (see Zhang Meifang 2001 for a history of translation textbooks in China). Beyond those lacunae, however, comparison with Vinay and Darbelnet shows a few other absences of importance. In particular, Loh has nothing to say on the level of what Vinay and Darbelnet call équivalence (corresponding idioms) or Adaptation. Loh’s general approach was to stay close to the text: there was to be no large-scale cultural appropriation or domestication; the foreign text would always remain situated in the foreign culture. In this, Loh remained faithful to the idea of ‘adequate translation’ as found in Fedorov, and perhaps also to the preference for ‘accurate translation’ expressed by Mao Zedong.

Solution types in a current Chinese textbook The lasting influence of Loh’s approach can partly be gauged by looking at a textbook currently used to train translators working between Chinese and English. Ye and Shi (2009) deal with solution types in the following chapters: 1. Conversion 2. Amplification 3. Omission 4. Negation 5. Division and Combination 6. Moving components in a sentence 7. Translating relative clauses 8. Translating adjectival clauses 9. Translating adverbial clauses 10. Translating sentences with the passive voice 11. How to translate metaphors 12. How to handle idioms and other culture-bound expressions Only the first six items name solution types; the rest concern types of problems. Of those six, the first four are clearly in the tradition initiated by Loh. The fifth



A Loh Road to China

97

item, which concerns the resegmentation of sentences, is not in Loh but dates from Zhang et al. (1980). Repetition has disappeared, but it was something of an idiosyncratic hobbyhorse in Loh anyway. Interestingly, Loh’s four solution types for foreign terms find no systemic place in the contemporary textbook, although Ye and Shi do briefly mention Transliteration (2009: 16). Does this suggest that such decisions are no longer within the purview of translators? It is quite possible that the front-line creation of new terms is nowadays more likely to be in the hands of journalists, commentators, general writers, perhaps anyone writing in Chinese with a knowledge of English, or indeed government terminologists. Or is it simply that terminology is now considered a separate discipline? Fifty years ago, when knowledge of foreign languages was restricted, the front-line status of translators was quite different: their range of decision-making was possibly wider with respect to terminology, and narrower on the level of Adaptation. The second point to be made here is that the special attention to metaphors and idioms (11 and 12 in the above list) is in the territory of interest to the Vinay and Darbelnet tradition; it perhaps owes less to the narrower linguistic frame that Loh found in Fedorov. It is difficult to read any specific historical change into this. It could be that the concerns of the various national traditions have converged to some degree. In any case, the tradition of Loh seems not to be as different from the West as it once might have seemed. This impression of convergence has nevertheless been challenged from within Chinese Translation Studies.

Resistance to Loh’s typology and its likes Tao (2011) relates Chinese textbooks on translation to the country’s political history and development as a world power. She divides the evolution of textbooks since 1949 into the following phases: initial, developmental, flourishing, multidirectional and professional. It is not my concern to evaluate the virtues of those terms. The point that worries me is that, in the run-up to ‘flourishing’, Tao cites only Loh’s propagandistic examples (2011: 41), locking him into a historical location where he was clearly not very happy anyway. There is something slightly unfair in selecting only those examples and thus implicitly suggesting that Loh’s entire thought can be held ransom to that past and had no lasting consequence. As Zhang and Pan (2009) tell us and as we have seen, some of Loh’s categories were carried through to the following generations, being repeated in subsequent textbooks, albeit with a few substantial modifications.

98

Translation Solutions for Many Languages

A more serious attack on that mode of thought comes from an outright nationalization of Chinese Translation Studies. In Liu Miqing (1990/2004: 236), for example, we read that ‘the basic paradigm of Chinese translation theory should start and end with our mother tongue – Chinese’. This radical centring emphasizes the differences between Chinese and inflectional languages, concluding that: the study of Chinese translation theory should not be dominated by morphology, which emphasizes the formal relationship between two semiotic systems. Rather it should be dominated by semantics, which focuses on the operational mode of symbols, meaning and reference … It is also necessary to care less about the linear equivalence formula and more about the correspondence on the composite level between two languages. (1990/2004: 237)

This makes a lot of sense: there can be no doubt that the difference between paratactic and hypotactic structure (to characterize a major difference between Chinese and Western languages) tends to shift translation solutions to a level of abstraction higher than that which is operative between most Indo-European languages (or are lower down on the Vinay and Darbelnet table). Liu Miqing’s ideal set of translation solutions would presumably have little time for most of Loh’s categories, which envisage the possibility of some kind of correspondence on the level of expressive resources. Let it all be ‘semantic’, or ‘functional’, and let the translator assume all necessary freedoms at any level below that. And yet, is there not something strange in a ‘translation theory’ that is concerned with just one language? Surely most translations involve at least two languages (cf. Lin Zhang 1998/2004)? Can translation categories really be fashioned in the image of just one side of the translator’s work? Or is this project a reactionary attempt to convert ‘comparative stylistics’ to just one monolingual stylistics, perhaps like Bally’s ‘stylistique française’, which was basically a comparison with German but did not want to declare itself as such? There are several reasons why monolingual translation solutions should be questioned, be it in Geneva or in China. First, there are many languages involved, some closer to Chinese, others more dissimilar. To simply emphasize difference is to reveal that you are only concerned with one kind of ‘other’, vaguely named as ‘inflected’, ‘Indo-European’, but one might as well say ‘English’ (nevertheless one of the least inflected of the bunch). This is not a centring on Chinese as much as it is a wilful distancing from English. Second, these days in China, students have to be trained to translate out of Chinese, since demographics mean there



A Loh Road to China

99

are not enough L1 speakers of English to meet market demands. If your translation theories are locked into just the start language, they are not going to help with this particular directionality. Third, as can be seen in the many examples used in the textbooks, a (perhaps surprising) degree of formal correspondence is possible between English and Chinese – the syntactic differences are greater in the cases of Japanese and Korean, which would have more reason to found their own independent theories, if syntax were the only criterion (I will return to this when discussing Japanese). The ‘Chinese only’ theorization of translation is nevertheless a laudable reaction to translation practices that allow increasing external interference in the regular structures of the language, notably from English. As formal calques abound (and yet are apparently not dealt with in a textbook for translators), one might understandably seek to get rid of them by any means possible, including by playing down any kind of formal comparison. The basic rule becomes something like: ‘Do whatever is necessary on the level of form in order to sound properly Chinese.’ That is not a bad rule to have. Of course, this desire to protect the established structures of a language was at the base of the entire comparative enterprise from the beginning. It was behind Bally and Malblanc trying to protect the character and psychology of French in the face of German; it was behind Vinay and Darbelnet driving into Canada to teach the Quebecois to protect their language from English; and it was behind Loh using the kind of rhetoric he had learned from English to then teach translators to exploit the special resources of Chinese. Conservative nationalism was a feature of this entire paradigm right from the start. If comparative stylistics should now want to revert to monolingual stylistics, it sounds more like a confession of historical defeat than anything else – we tried to teach students how to keep the languages separate; our lessons didn’t work; so let’s stop teaching on that level! There is some sense in the argument. But not enough.

5

Spontaneous Combustion in Central Europe?

Chapter summary: Three theorists in Central Europe produced highly original ideas while partly drawing on more Eastern approaches to translation. Jiří Levý, in Brno, formulated a typology of translation solutions and claimed that translators exhibited ‘tendencies’ (later researchers would call them ‘universals’) whereby translations tend to be more general, neutral, logical and explicit than their start texts. Anton Popovič, in Slovakia, produced several elaborate typologies from which he derived three ‘stylistic attitudes’ of translators, drawing on analyses of texts rather than languages (thus unlike Bally). Otto Kade, in Leipzig, expressed his ‘equivalence types’ in simple arithmetic terms (‘one-to-one’, ‘one-to-several’, etc.) while giving considerable weight to the extra-linguistic situation and the translator’s subjectivity. Levý and Popovič contributed significantly to the development of Western Descriptive Translation Studies, even though that tradition, more concerned with literary history than with training translators, has shown only marginal interest in solution types.

After 1989, visiting Prague, Bratislava and Leipzig was genuinely exciting. It was somehow a less adulterated Europe, the one that had been hidden from us in the West, in what looked like crumbling ancient glory. What was once the concealed East, behind an iron curtain, could be seen as what it had always been: Central Europe. And as I attended conferences in those cities, it became very clear that nothing was going to be simple. Occasionally one heard phrases like ‘the free world’, with brave enthusiasm for the opening to the West. For the most part, though, the scholars I met had little doubt about the solid value of the intellectual traditions that supported their work on translation, and most of them appeared strangely patriotic, recently liberated from a dark transnational yoke. We from the West had come to learn. There are several competing accounts of what was happening in Translation Studies in what I shall loosely call Central Europe, defined by the withdrawn veil of Real Socialism.

102

Translation Solutions for Many Languages

The first is informed by those doses of national pride après coup. On this view, there would be schools of thought based in cities: a Czech school around Jiří Levý (who taught in Brno) would have inherited the work of the Cercle Linguistique de Prague of the 1920s and 1930s; a Nitra School would be formed around the work of Anton Popovič and František Miko in Slovak; a Leipzig school would unite the work done at the Centre for Applied Linguistics at Leipzig University. Those places would have produced original insights into the nature of translation – mainly literary in Brno, Prague and Nitra, mainly linguistic and technical in Leipzig – and those places would be separated along national lines. There is nothing inherently wrong with that view, since proximity in a city can bring a meeting of minds, and Czech is obviously much closer to Slovak than it is to German. A second account emphasizes the contacts established between scholars working on literary translation in the 1970s and 1980s: Levý, Popovič and Miko met James S Holmes from Amsterdam and Itamar Even-Zohar and Gideon Toury from Tel Aviv, along with the Flemish scholars José Lambert, Raymond van den Broeck and André Lefevere. This incipient network was initiated at a conference held in Bratislava in May 1968 (a good month for attempted revolutions) and was followed up by further conferences in Leuven in 1976, Tel Aviv in 1978 and Antwerp in 1980, as well as at meetings of the International Comparative Literature Association. By some accounts, this would be the origin of the ‘Manipulation School’, or ‘Descriptive Translation Studies’, or even, for a few, the birth of Translation Studies as a science – although the claim that the whole of Translation Studies was somehow born in Leuven in 1976 is little better than a creation myth. For this approach, it would be clearly wrong to place Leipzig in anything like the same basket as Brno, Prague and Nitra. A third account, which I have formulated elsewhere, sees the work in Brno, Prague, Nitra and Leipzig as part of the development of more general European concepts such as the desirability of analysing works of art in terms of objective patterns and the usefulness of the concept of system when applied to cultural practices. Ideas on that level can be seen as proceeding historically from Russian Formalism (and prior to that, European Symbolism), with the Cercle Linguistique de Prague being just one step in a movement of general development and diffusion. On that account, the descriptive work on translation in the 1970s and 1980s, which was broadly based on at least those two ideas, was just part of a wider movement involving the whole of Central Europe, as large ideas moved across national borders.



Spontaneous Combustion in Central Europe?

103

That third account tends to upset the nationalisms of the first account: no, say the scholars in this new Central Europe, we are the ones who invented the key concepts and the others have been taking from us. For example, it would be mistaken, says Zuzana Jettmarová (2011: xviii), to associate Levý with anything from Russian Formalism, since ‘Prague structuralism was in many fundamental ways its outright opposite’. In some respects, this is certainly not wrong, but you have to find and describe those particular respects. If you look closely at the biographies and the texts, Central Europe was a space of moving people and multiple mixed traditions. ‘Prague’ was key not so much as a city as it was a node in a network,1 and although Levý (1963/2011: 10) undoubtedly does draw on the Cercle Linguistique when he cites with approval a 1913 text by Vilém Mathesius, one of the Czech co-founders of the Cercle, on the same page he also allies himself with Roman Jakobson, another co-founder, who was a travelling Russian (and sometimes classified as one of the Russian Formalists), and then he cites approvingly from the German classicist Ulrich von WilamowitzMoellendorf, the Polish linguist Zenon Klemensiewicz and the German linguist Werner Winter. A Czech draws on a Russian, a Pole and two Germans (and much else throughout the book!). This does not seem to be a simple tale of one national school against another. And then much depends, as well, on what counts as ‘originality’. You don’t just wake up one morning and decide you are going to study literary translations in a systemic way – everyone works from what has been done previously and in relation to what you are aware is being done elsewhere. None of these three accounts is wholly wrong. But if someone offers you just three alternatives, the third is always the one they want you to accept. In the cases of Levý and Popovič, the texts themselves go to some lengths to indicate the kind of political context they wanted to be seen working within. In the early 1960s Levý (1963/2011: 4–6; 1963/9: 15–17) gives an impressive list of the studies on translation that had been published in European languages, along with the centres where translation was being researched, including Geneva, Montreal, University College London and Austin (Texas). About a decade later, Popovič (1975/2006: 3–4) gives a similar list, in which I note the addition of Saarbrücken, Leuven, Binghamton, Canberra (first time I ever heard of it!), Edmonton, Amsterdam and Leipzig, as well as Nida and the previous work in Czechoslovakia. In both lists, pride of place is nevertheless given to the Soviet tradition, which Levý (1963/2011: 5) calls ‘the most systematically active work in translation theory’, while Popovič (1975/2006: 3) concurs that ‘[t]he first position in the world is occupied by Soviet translation theory’ (in a usage

104

Translation Solutions for Many Languages

where ‘translation theory’ covers what we nowadays call ‘Translation Studies’). Both accounts then describe the split between the literary and linguistic sides of Translation Studies in Russian, with Levý admitting that the polemics ‘have been rather pointless and futile’ (1963/2011: 5). One could of course regard the praise of Soviet theory as a political gesture designed to keep Big Brother happy – this was the age of the Warsaw Pact and Real Socialism, after all. And the surprising aspect remains the pains that both theorists take to place their own work within an extensive international network of study and debate, bridging both East and West. Although their texts certainly do deal with literary translations into the Czech and Slovak languages, the frame of their thinking is by no means language-specific. They saw themselves as participating in an international project, reaching both eastward and westward. One way to test these views is to compare them on the basis of a few common criteria. The search for translation solution types enables precisely such a test, although there is no reason why what we find should then be applied across the board. All we have to do is find and compare the way these theorists describe translation solutions. We might then be able to say something about how their work relates to other formulations of the same problem. The first point to be made, of course, is that the theorists we will be looking at in this chapter did indeed classify solution types. They might have used other names and other concepts, but the basic intellectual endeavour was there. To that extent, they were certainly working within a predominantly European tradition.

Jiří Levý and the behavioural tendencies of literary translators Jiří Levý (1926–67) was born in Košice, in what is now Slovakia, near the border with Hungary. After studies in English and Czech at the University in Brno, from 1950 to 1963 he taught at the University of Olomouc, in what is now the Czech Republic, and from 1964 he held the chair for Czech Literature and Literary Studies at the University of Brno, in the Czech Republic. Levý’s first book was a history of Czech theories of translation (České teorie překladu), published in 1957. His impact on Translation Studies came through his seminal work Umění překladu, published in Czech in 1963, translated into German by Walter Schamschula (with Levý’s participation) in 1969 as Die literarische Übersetzung, then published in a second Czech edition in 1983, incorporating the additions made in the German version. The book was



Spontaneous Combustion in Central Europe?

105

translated into Russian in 1974, into Serbo-Croatian in 1982 and into English in 2011 as The Art of Translation (working from the original Czech and the German translation). Here I am reading the German and English versions. This roundabout play of original and translations might explain why many of the examples concern English and German, rather than Czech. In that respect, at least, Levý was far from being a linguistic nationalist. The book is a wide-ranging discussion of the options available to literary translators; it is not primarily designed to teach anyone how to translate. If you look for translation solution types (called ‘procedures’ in the English translation, ‘Verfahren’ in the German), here is what can be found. For the translation of names, which is used as an illustration of the general nature of translation, Levý actually plays with four terms (1963/2011: 86–7; 1969: 88–90): Copying: Straight reproduction: ‘Mr Ford’ (the car manufacturer) becomes ‘Mr Ford’. Transliteration: Change of script, as from Cyrillic to Roman script. Translation: The mystery play character ‘Jedermann’ [every-man] becomes ‘Everyman’; ‘Mr Ford’ (in The Merry Wives of Windsor) becomes ‘Herr Fluth’ [‘Mr Flood’] in Schlegel’s translation (1963/9: 88). Substitution: ‘Replacement by a domestic analogue’, as when ‘Sir Oliver Surface’ becomes ‘Herr Oliver von Obenaus’ [Sir Oliver Fromabove?] (1963/2011: 86). Simple enough, it seems, but even here the theorizing is messy: ‘Only transliteration, not copying, counts as a translation procedure’ (1963/2011: 87), says Levý, but then ‘translation’ is also described as a separate kind of solution. Translators are thus generally seen as mostly having to select between two alternatives: ‘translation per se and substitution’ (1963/2011: 105), even though substitution is one of the things translators can do. One is reminded here of the way Derrida (1985) criticized Jakobson’s three types of intersemiotic translation, of which only one was ‘translation proper’. Of course, the difficulty might lie in the slippery nature of translation itself. But it could also just be loose theorizing. The terms ‘analogue’ and ‘substitution’ here seem similar to their usages in the Russian tradition. One is also reminded of how Fedorov produced a remarkably similar typology for the treatment of ‘realia’ (when describing them I used the terms ‘transliteration’, ‘componential translation’ and ‘analogue translation’, while ‘copying’ was missing because everything would normally be transcribed into Cyrillic). This need not suggest influence (if anything, Levý’s intellectual

106

Translation Solutions for Many Languages

style seems closer to that of Fedorov’s literary opponents), but it does indicate that the same problems were being addressed in much the same way. Levý then sees lexical problems as being solved by shifts along three axes (1963/2011: 114): General vs. specific denominations Neutral vs. expressive stylistic denominations Repetition vs. variation of vocabulary These categories are followed by three axes of ‘intellectualization’, which are things that translators can also do: ‘Logicalization’ of the text Explicitation of what is only half-said (content) Explicitation of syntactic relationships (form) Levý’s general claim, based on the observation of many examples, is that, on the lexical axes, translators tend to opt for solutions that are more general, more neutral and have less variation than do the corresponding start-text items, while on the ‘intellectual’ axes (perhaps better described as ‘ideational’?) they tend to produce translations that are more logical and more explicit. These are absolutely key claims; they lie at the origin of Descriptive Translation Studies; they are at the root of any attempt to identify features that are found more in translations than in any other kinds of texts – one should not forget that the German subtitle of Levý’s work is Theorie einer Kunstgattung, ‘theory of an artistic genre’, claiming that translations are like each other in some way. These are the kinds of features that the Tel Aviv school would later term ‘universals of translation’ (see Pym 2007, 2008) and Toury (1995/2012) would bring together as ‘laws of translational behavior’. As such, they would clash head-on against the pedagogical tradition of Vinay and Darbelnet, who had strong preferences about what translators should do, over and above what they tended to do in real life. This is where one might claim that the Czech theorist was doing something completely different. Exactly how different? Given that he presents three axes rather than a series of discrete solution types, Levý could be thinking in terms of continuous variables, clines along which the translator might slide one way or the other. Is that how he is thinking? His text nevertheless does give names for what translators do, and his general discourse would appear to be based on discontinuous variables. Similarly, despite the ostensibly descriptivist stance, Levý is certainly not above expressing his preferences for some translations over others: the tendencies to



Spontaneous Combustion in Central Europe?

107

clarify and explain are seen as reducing the aesthetic function of the text, erring too far in favour of communication, in terms that are vaguely reminiscent of Walter Benjamin’s essay ‘On the Task of the Translator’ (1923/2004). And for that matter, more generally, both the descriptive and prescriptive approaches are necessarily based on the options available to the translator. So we should perhaps not be looking for radical novelty on that level. Not surprisingly, within Levý’s accounts of these axes, there are some solution types that are very much like the classical ones found elsewhere. Explicitation is indeed in Vinay and Darbelnet’s B-team of solutions, as is Compensation, which Levý includes in another part of his text, giving the classical example of using register to express the distinction between formal and informal second persons (1963/2011: 104) – it is not entirely clear how Compensation would fit into the sets of axes, though. As for the tendencies themselves, they are sometimes expressed in a way that could be read as referring to solution types. For example: Nivelization [sic]: ‘Emotionally coloured expressive means sometimes lose their stylistic value in translation if rendered by a neutral, colourless expression’ (1963/2011: 111); also called ‘regression to normalisation’ (ibid.). Intensification: The opposite of ‘nivelization’. Limited lexical variation: ‘In translation, the resources of a language in terms of synonyms for subtle differentiations of meaning are generally underexploited’ (1963/2011: 113). Where an English novel might use ‘he/she said’ throughout, other literary traditions require more variation (‘replied’, ‘retorted’, ‘explained’, etc.). The technical terms are admittedly unhappy in English, but they can certainly be construed as at least talking about solution types. So is there anything really new here? Or is this just another way of saying basically the same thing? I think Levý was using a fairly traditional organization of solution types in order to say something substantially new about translation, with long historical resonance. The idea is simple and of unsuspected fecundity; it is not something that can be reduced to prescriptivism vs. descriptivism; and it does not necessarily derive from Prague structuralism, as far as I can tell. It runs like this: the solutions used by translators tend not to be like those used by non-translators, so translational language is different from non-translational language: ‘it is the psychology inherent in the activity of translating that is behind the tendency for translators themselves to opt for generalisation, neutralisation and repetition [as opposed to variation]’ (1963/2011: 114). I am not aware of this being

108

Translation Solutions for Many Languages

said in such systemic terms prior to Levý. To be sure, Vinay and Darbelnet were complaining in the background that translators did not use the creative solutions that non-translators use, and any number of literary critics had similarly lamented the distinctive qualities of translations, sometimes called ‘translationese’. But here in Levý we find a way of approaching solution types themselves in terms that can map the specificities and divergences of translational language. That was new and insightful. The message to translators is no longer just, ‘You can do A, B, or C, and I think it is better to do C’ (because we help maintain or enrich the target language, or whatever); now it is more like, ‘You can do A, B, or C, but watch out, most translators tend toward A even though it might be better to try B or C’. That said, what appears not to be developed in Levý’s book is any explanation of why translators tend to opt for some solutions rather than others. Several factors are casually named (tradition, ease, ignorance), but it is hard to discern anything that might be a full psychological or sociological explanation. What we find instead is an account of the wider factors that translators should bear in mind: Specific translation procedures are part of the translator’s overall method and in turn individual solutions are subordinate to the overall approach. Regarding solutions of individual issues, two recurring fundamental considerations underlying translation methods have been identified: (a) the work itself and above all the interrelationship between its unique and general attributes; (b) the reader, especially his ability to comprehend unique facts and allusions. (1963/2011: 104)

The relation between the parts and the whole should sound familiar to readers of the Soviet tradition (it is one of the main points repeated in Fedorov), as is the importance of the particular receiver, which we have seen mentioned since Sobolev (1950). However, no Russians are mentioned in the vicinity of the above passage and many of the basic ideas could have been recruited easily enough from Mukařovský and home-grown Czech structuralism. Levý goes further than this when discussing theatre translation, formulating a ‘principle of selective accuracy’ (1963/2011: 162). This means the translator’s approach: involves something like a system of variable procedures, subject to the translator’s conception of the respective dramatic configurations and his notion regarding the primary objective of the performance. The translator’s approach to the text is therefore flexible; in some cases precise semantic nuances are



Spontaneous Combustion in Central Europe?

109

of paramount importance while in others style and intonation will tend to predominate. (1963/2011: 162)

As we saw when dealing with Russian theories, the appeal to the ‘objective of the performance’, along with the idea that accuracy is adjusted according to need, had also been part of Soviet theorizing since Sobolev (1950). So, too, might be the idea that there is (or should be) only one kind of translation ‘proper’. However, much of the above passage could equally have been recruited from Catford (1965), whom Levý cites elsewhere in his text. Whatever the case, despite the many axes along which the translator can drift one way or the other, Levý does not stay with any grand dichotomy such as ‘domesticating’ vs. ‘foreignizing’, or ‘dynamic’ vs. ‘formal’ (and he was certainly aware of Nida’s approach at the time). He could easily have gone down that road, but he did not. That seems not to have been the Central European way.

Anton Popovič and solutions as style Anton Popovič (1933–84) was born in Prešov in Eastern Slovakia, studied Slovak and Russian in Bratislava and did postgraduate studies from 1956 in Brno (Moravia, in what is now the Czech Republic), where he met Levý (Špirk 2009: 4) – remember that Levý’s first book came out in 1957, his second in 1963. Popovič’s first book, published in 1961, was on the reception of Russian literature in Slovakia. From 1968 he was Associate Professor in Bratislava (Slovakia) but was simultaneously associated with the Department of Literary Communication in Nitra (Slovakia), along with František Miko and several other scholars. Popovič is thus usually associated with the ‘Nitra school’. Popovič’s main work is Teória umeleckého prekladu (Theory of artistic translation, 1975), rendered into Russian, Hungarian and Serbo-Croatian in 1980 and into Italian in 2006 (I am reading the Italian). For my present purposes, though, the fragment of most interest is from the early 1968 work Preklad a výraz (Translation and Expression), where Popovič explains the stylistic differences that can be found when you compare translations with start texts (these are called ‘shifts’, a term used by Catford in 1965). The list interests me because there are seven types and they could be translation solutions.2 Here I reproduce the typology as translated by Jaroslav Špirk (2009: 8; 2014: 32), where the 1968 terms are followed by the 1976 ones in parentheses:

110

Translation Solutions for Many Languages

Levelling (‘nivelization’ in Popovič 1976): Simplification of the expressional, i.e. stylistic, qualities of the original. Intensification (emphasis): Exaggeration of the expressional qualities of the original. Transformation (inversion): Change in the expressional values of the original. Substitution (substitution): Replacement of the original expressional features by domestic ones (encompasses words, phrases and idiomatic expressions). Compensation (correspondence): Compensating for untranslatable elements, often in another place, by stylistic means unique to the translation’s language. Standardization (‘typization’): Translating by stylistic means typical of the translator’s language and literature. Individualization (individualization): Translating by stylistic means untypical of the translator’s language and literature. All these types are called ‘stylistic’, since they operate on the plane of expression in a text, and they are all operative at the ‘microstylistic’ level, since Popovič also has a lot to say about macrostylistic and sociological variables. But can they really be seen as ‘solution types’? Well, Substitution and Compensation certainly look like solution types, and ‘transformation’ might be one too, if we knew a little more about it. Standing either side of those recognizable types we then have two axes, reminiscent of Levý: Levelling looks like the opposite of Intensification, just as Standardization and Individualization are opposite sides of the same coin. What is happening here? I don’t think these two pairs can be read in the same way. The first couple, as in Levý, is regrouping a set of fairly familiar solutions such as Simplification vs. Amplification, Explicitation vs. Implicitation. The second pair, though, is saying something very different: it says, more or less, that everything depends on what is normal in the receiving context. The translator might opt for solution A or solution B (Explicitation or Implicitation, let’s say), but the import of that solution is how it relates to what is normally done in non-translational texts in the target culture. In theory, the same solution could be Standardization in one context and Individualization in another. I don’t have access to any Slovak example for this, but Chinese might do: the tag ‘某某说’ ‘[someone] said’ comes before a direct-speech utterance in traditional Chinese fiction but after the utterance in an English novel. Over the past fifty years, though, the English order has become normal in Chinese novels, which means that a Chinese translation that puts the tag after the utterance would have been ‘individualizing’ fifty years ago but is now quite ‘standardizing’ (the example is from Tian 2013: 59).



Spontaneous Combustion in Central Europe?

111

In itself, this second pair could simply be an observation about the contextual effects of certain solutions; it may not affect the solution types themselves. However, Popovič (1968: 43–4) uses analysis of the above shifts to derive three basic ‘stylistic attitudes’ that a translator can have with respect to the target literature – once again I cite from Špirk (2009: 9; 2014: 33): Zero attitude: Does not enrich the literature of the target literature; a style sometimes called ‘translationese’. Redundant attitude: A style in which the individual style (i.e. poetics) of the translator influences the style of the translation more than the original work does (also called ‘adaptation’). Discovery of a new style: An attitude whereby the translator discovers a new style, i.e. the style of the original, new to the domestic tradition of the translation, a stylistic calque. One might struggle to understand the terms here (for me, ‘translationese’ is more likely to be in a translation that is full of calques) and the deck seems very stacked in favour of what a different tradition would call ‘foreignization’. The point to make, though, is quite simple: this highly synthetic level of analysis strangely incorporates terms that look like quite traditional solution types: ‘adaptation’ and ‘calque’. One way or another, the solution types are there. They are just being embedded in a theoretical discourse that has many other things to say. So what is Popovič saying of importance? Several things. First, that style is not just in the language system (as it was for Bally and the tradition of comparative linguistics): it lies in the complex relations of the text – Popovič was using František Miko’s theory of style, which also included semantics (Špirk 2009: 20). Popovič thus has little special to tell us about the Slovak language, but he does have a lot to say about literature. Second, various levels of abstraction are possible (the micro-analysis has some solutions; a synthetic, higher-level analysis has others), which means that not all solutions have to be operating on the same epistemological level. True, Popovič is analysing literature rather than training translators, but the message is there nevertheless. This retrospectively questions the typologies where all terms are indeed presented as operating on just the one level. And third, as mentioned above, whatever is done within the text is in some way subordinate to the way it relates to the context of reception: the one solution can have different values in different contexts. Popovič had many other things to say. In addition to several complete reworkings of the above typology (see, for example, Popovič 1975/2006: 83–99),

112

Translation Solutions for Many Languages

he conceptualized the intertextual invariants of start text and translation as a ‘metatext’; he outlined a sociology of translation; he drew up a checklist for the historiography of translation; he formulated a ‘praxeology of translation’ based in part on ‘confronting the system of translation science and socio-cultural needs’, which meant studying, among other things, ‘the influence of cultural policies (the Party principle) on the translation programme’ (Popovič 1975: 282, 239, cit. Špirk 2014: 29, but I cannot see the Party mentioned in the Italian translation).3 For all of that, I remit to the work of Špirk, just as I await future translations from Slovak. In the face of Popovič’s wealth of interests, it is quite perverse to go hunting for no more than solution types in his work. Despite that, the three points that can be extracted from Popovič’s use of solution types are in themselves magnificent contributions to the ongoing historical tradition: style is in the text, not the language system; solutions operate on different levels; and their effects depend on context. These are all ideas that can be placed alongside Levý’s insight that translators and non-translators tend not to use the same linguistic patterns. Some things did indeed catch alight in Central Europe. Špirk (2009: 4) notes that the Communist Party allowed Popovič to represent Czechoslovak Literary and Translation Studies abroad – he was able to travel and was in touch with a network of scholars with similar concerns. Špirk also notes, though, that Popovič remained a marginal figure in the West, with only eight publications in English or German, since ‘the Communist Party of Czechoslovakia would not allow open cooperation with Western scholars’ (2009: 21). These things are difficult to judge. Despite all the propaganda, Popovič was obviously able to travel, he did have those eight publications in English or German (you only need one good one!) and he did participate in an international network of scholars – all of whom were also fairly marginal at the time. It is a little too easy to assume that Communism meant total closure. I remember attending a very moving homage to Popovič delivered by Raymond van den Broeck in Prague in 1995, in the presence of Gideon Toury and José Lambert. It was very clear, from the emotions of that speech, that the relations between the various members of the network had been both scholarly and affective, far warmer than a purely Cold War narrative would suggest.



Spontaneous Combustion in Central Europe?

113

Otto Kade and innovation in Leipzig Many will object to Otto Kade being placed here, in a Central European context, rather than in the following chapter on German approaches. German historians of translation theory would prefer to talk about a separate ‘Leipzig School’, just as the Czech and Slovak scholars would prefer to reconstruct their own historical tradition. I will try not to upset them gratuitously. Otto Kade (1927–80) was born in Frýdlant, in what is now the Czech Republic. He was conscripted into military service in 1943 (at the age of sixteen) and was in the Reich Labour Service in 1944. In 1945 he was held by the British as a prisoner of war. In 1946 Kade worked as an interpreter for the Národní výbor (National Committee) in Frýdlant before moving to the Soviet-held part of Germany, where he began his career as a teacher of Russian in technical schools. From 1950 to 1970 he was a practising conference interpreter, at ease in German, Russian and Czech (Salevsky 2007: 367); he was in charge of the East German translators and interpreters at Party conferences, both in East Germany and abroad. In 1956 he was a lecturer at the newly founded Institute for Interpreting at Karl Marx University in Leipzig, where he became Deputy Director. Kade held various official positions within the university, as well as adviser to the Ministry of Education and co-founder, in 1962, of the East German professional association of interpreters and translators (Salevsky 2007: 369). He developed close scientific and personal contacts with what was at that time called the Maurice Thorez State Pedagogical Institute of Foreign Languages in Moscow, with which he carried out empirical research on conference interpreting. Kade is best known for his 1968 book Zufall und Gesetzmäßigkeit in der Übersetzung, which has not been translated into English but whose title might be rendered as ‘Coincidence and regularity in translation’.4 As we noted when commenting on the receptions of Russian theories, Kade does cite Fedorov extensively, but only in Russian, without translation into German. Other Russians referred to include Sobolev and Barkhudarov, although not Retsker. There are also notes on Vinay and Darbelnet, Nida, Bloomfield and others from the West, but Kade is clearly more engaged in dialogue with ideas from the East. He has read Levý’s 1963 Umění překladu, in Czech, describing it as ‘the most significant attempt to construct a theory of literary-artistic translation’ (1968: 109), although he has nothing further to say about that significance (Kade was working on general translation, not literature). Tellingly, his examples are drawn from the German translation of a congress of the Communist Party of the Soviet Union. A Party man, he was by all accounts a believer.

114

Translation Solutions for Many Languages

Kade, like Fedorov, opposes the notion of untranslatability, which he considers an idealism that gives one language more value than another. Since a language cannot be separated from its speakers (Träger), untranslatability ‘puts us on the road to reactionary racist ideology’ (1968: 17). Kade was not pulling his punches. On the other hand, the same mistrust of idealism brings him to criticize Jakobson’s and Bloomfield’s notions of ‘effability’, the principle that anything that can be expressed in one language can be expressed in all languages (1968: 68). This facile idealism, intimates Kade, is equally amiss. He instead calls for a flexible notion of what is or is not translatable, and he criticizes Fedorov for what he sees as a lack of Marxist dialectics in this regard. This brings Kade to a basic model of relations between start and target utterances: one-to-one, one-to-several, one-to-part and one-to-none. This simple arithmetic, which could have been drawn directly from Fedorov (1953: 122–3), recognizes that translatability is a variable. That model then produces a rather more complicated typology of ‘equivalence types’ (Äquivalenztypen) (1968: 79–81), where the numerical relations are conjugated on the linguistic categories of form and content and then explained in terms of langue and parole: 1. Total equivalence: One-to-one correspondence (Entsprechung) on the levels of both form and content, as in the case of names, titles and numbers. 2. Facultative equivalence: One-to-several correspondence on the level of expression and one-to-one correspondence on the level of content. ‘An effective one-to-one relation is only possible in the domain of parole’, since the choice between one solution or another will depend on the linguistic and situational context (1968: 81). 3. Approximative equivalence: One-to-one correspondence on the level of expression; one-to-part correspondence on the level of content. The classical case here is the colour names that do not correspond to each other in different languages when considered in terms of the language systems, but which can correspond in the domain of parole, when referring to one particular manifestation of colour: ‘The one-to-one relation on the level of parole is generally reached through the context’ (1968: 81). 4. Null-equivalence: One-to-none correspondence both on the level of expression and on the level of content. This would be the case of realia where the foreign term is used but the concept is still lacking in the target culture. In principle, examples could be any use of a foreign term (such as my use of ‘parole’ here). Kade notes, however, that after an initial period the



Spontaneous Combustion in Central Europe?

115

foreign term may start to function as a normal part of the target discourse, thus paradoxically approaching one-to-one correspondence (1968: 82). Although these are presented as types of equivalence, the term Entsprechung (correspondence) was to gain more general usage as the way the German tradition described solution types, and the categories can indeed be analysed as solution types. Several observations are in order. First, all the examples here are lexical. The categories are nevertheless referred to again when Kade discusses fuller linguistic transformations (as we shall see in the next chapter), so one supposes they should apply to more than the lexis. Second, as noted, Kade theoretically has three variables: simple mathematical relations (with four values: one-to-one, one-to-several, one-to-part, one-tonone); linguistic form and content; langue and parole. Even if we discount the last variable, which is only used for explanations, the combination of the first two should give sixteen combinations (that is, sixteen solution types), of which only four are presented. Why is this? Clearly, only two mathematical relations are really in use: one-to-one and not-one-to-one (which gives the four). Third, these two values map onto the major distinction between ‘equivalence’ and ‘substitution’ in accordance with the way those terms had been used in Russian theory at least since Smirnov (1934), except that here we have two kinds of ‘equivalence’ (1 and 4) and two kinds of ‘substitution’ (2 and 3). For that matter, they could also be mapped back onto Vinay and Darbelnet’s distinction between direct translation (1 and 4) and oblique translation (2 and 3). So what is really new here? Not much in the actual divisions themselves. Fourth, Kade is occasionally praised (for example in Salevsky 2007 and SnellHornby 2006: 30) for having applied ‘communication theory’ to translation, thus apparently helping Translation Studies to move beyond a narrow fixation on linguistics. That claim might not be obvious from the categories presented above. Kade does, however, give a considerable role to the terms ‘situation’ and ‘context’, which become synonymous with his description of what is possible on the level of parole. This, I suspect, is the way ‘communicative theory’ might feed into the analysis of solution types: the choices made in types 2 and 3 ‘depend on the situation’, which is another way of saying that the reasons for selecting them cannot be theorized in advance. This may have been something passably new. Then again, Popovič, also in 1968, was similarly recognizing the active role of the translator, who could accept or change the prevailing stylistic options. On both sides, in the same years, these theories constituted in nuce an active

116

Translation Solutions for Many Languages

recognition that the translator’s subjectivity was obliged in these situations to make choices that depended on factors so complex as to be untheorizable in terms of guidelines or preferences. The original title of Kade’s work referred to ‘objective and subjective’ factors in translation, and it is in the appeal to situation that we find the subjectivity. I fail to see how that constitutes communication theory, or indeed why that insight cannot be expressed in linguistic terms. At one stage (in Pym 2010/14) I collected theories of translational equivalence and started to put them into two boxes: theories of ‘natural’ equivalence on one side, where the translator was supposed to go out and find a pre-existing optimal solution, and theories of ‘directional’ equivalence on the other, where the translator was seen as actively creating equivalents, which could theoretically be fresh each time. And then I came to Kade. Which box should his theory of equivalence be put in? Hard to say: types 1 and 4 use the ready-made, while 2 and 3 require creative choices. When you find something that does not fit into your beautiful pre-established boxes, especially when it is as simple as a four-term typology, there is a good chance it is saying something new.

The politics of a middle empire My Czech and Slovak friends complain that the historical work in their tradition is always overlooked, and now they will certainly complain that I have put them in bed with an East German and a Party man at that, when apparently everyone else was an anti-Communist at heart. Much as the lamentations are understandable, I do not think the productive ideas of these three theorists have lacked impact. Levý’s identification of ‘tendencies’ was to become a huge area of research and debate within Translation Studies; the active role attributed to the translator in Popovič and Kade radically altered the way equivalence was theorized; and the influence of Czech structuralism, if that is what is at stake, found multiple other outlets not only in linguistics (Prague phonology was the showcase success story of international structuralism; Roman Jakobson carried the lessons as he went ever westward) but also in literary studies (notably through the work of René Wellek within American Comparative Literature). Central European theories have certainly had influence. Yet, admittedly, there has not been excessive active recognition. I suspect that Cold War politics had their part to play. In principle, in the Communist regimes of Central Europe in the 1960s and 1970s, any full professor was a member of the Communist Party and political leadership was supposed



Spontaneous Combustion in Central Europe?

117

to come from the Soviet Union, in all spheres of life. On the up side, cultural policy was also based on the maintenance of linguistic diversity at the national level, which meant that there were quite massive state-sponsored translation programmes (hence Popovič’s ‘praxeology’, to check on how the programmes were applied). Further, there was broad application of Marx and Engels’ idea that the proletariat had to inherit bourgeois culture, which meant not only translating high literature for the masses, but also selecting texts to be translated in order to combat rival ideologies, notably from the Church (see Pokorn 2012 on the role of translated children’s literature in Socialist countries). For all these reasons, work on translation actually did quite well under Communist regimes, quite independently of whatever opinions were held by the theorists themselves. And translation theory was ultimately there to check on the quality of the huge amount of translating that was being done. It was there to perform a social service to the state apparatus. The French and Canadians liked certain solutions because they were interested in defending the wealth of a large language, French, which was their idea of social service. In Central Europe, on the other hand, translation theory had little to say about protecting language systems but much to say about the values of disseminating high culture, and that, I suspect, is where their politics played out. When Levý’s ‘tendencies’ set up an opposition between ‘communication’ and the aesthetic values of the (great literary) text, the thrust is essentially the same as Popovič’s defence of literature against ‘commercialism’. In both cases, the higher values are legitimated by an entity that, for the theorists, should have no particular concern for clear explanations or competitive commerce: the state. Even when Popovič complained that about a quarter of the translations into Slovak were of ‘low quality’ and that one of the causes of this was the lack of competition between translators (1975/2006: 141),5 his proposed remedy was not to introduce any competition in the market sense: his more immediate aim was to develop a sophisticated translation criticism that could raise awareness of what was going on. In that respect, he remained faithful to his political mission. Yet these were not static thinkers. Our three theorists illustrate Central Europe as a space of mobility: Levý was born in Slovakia but worked in Brno; Popovič was born in Slovakia, studied in Brno, then returned to Slovakia; Kade was born in what is now the Czech Republic and moved to Leipzig. Similar stories can be told for the Cercle Linguistique de Prague. The movements can also be seen in the books on translation written in the 1960s and 1970s. Levý’s main book went from Czech to German and back to Czech; this and some of the other books mentioned here were translated into Hungarian, Russian and

118

Translation Solutions for Many Languages

Serbo-Croatian well before reaching Italian, and only recently English – most of us reached this space very late. That zone of multidirectional movement and exchange now occupies a large middle ground between East and West, a zone of political ambivalence, an area of multilingualism and, necessarily, of translation. The translations are there largely to help minority languages and cultures survive, as recognized by Levý in his early work on Czech theories of translation (České teorie překladu, 1957).6 And the theorization was presumably there to help the translations. But the movement of ideas as such does not have as its aim the maintenance of smaller cultures. That relation is far too simple. We might more productively observe that thinkers in smaller cultures are forced to be polyglots; they thus have wider horizons and a certain propensity to translate and to appreciate translations; they thus tend to develop ideas about translations. The theorists we have been looking at drew on thought from well beyond their immediate traditions; they wanted to be seen as players on the international stage (no doubt in order to enhance their local status); and then they did indeed exert influence within that wider frame. How should we understand the relation between the maintenance of the local culture and the larger-scale work of intermediaries? In this case, the networks loosely map onto a composite political space, the Habsburg Empire, from which Leipzig was admittedly always just to the north (but then it was long a significant node for trade its own right, on the intersection of the Via Regia and the Via Imperii). The composite middle kingdoms, as cultural buffer zones, can play tricky political roles. Bear with me: the first written document in French is said to be a translation, in fact of the Strasbourg Oaths of 842. The oaths were sworn in French and German, hence their translational status. They were sworn by two brothers – one in what is more or less France, the other in what is more or less Germany – in order to exclude a third brother, Lothar, who had the crown of Francia Media. When two gang up on one, the one loses: Francia Media disappeared from history as a state. But that middle ground, the place where translations were performed between the major players, went underground. It reappears, in part, if we map the places where translators come from and where their task is theorized. The smaller countries, not just in the Netherlands and Flanders (the northern parts of Francia Media) but also in Habsburgian Central Europe, tend to act as middle kingdoms of translators. They go underground, then resurface as spaces of mediation. It is not altogether surprising that some of the most dynamic translation theories of the twentieth century came from the smaller cultures of composite middle empires.



Spontaneous Combustion in Central Europe?

119

Descriptive Translation Studies as a project without solution types The coming together of scholars working in smaller countries, with smaller languages (Czech, Slovak, Dutch, Hebrew), created a network with various names: from the ‘Manipulation School’ to ‘Descriptive Translation Studies’ (as in Toury 1995/2012). That network also created a conceptual map for Translation Studies as a discipline, dating from Holmes (1972/2005) and subsequent commentaries, in apparent oblivion of calls for something similar in Russia at least from 1958. Such has been the historical success of that network that it is not infrequently seen as originating the whole of Translation Studies, which would seem to be a gross simplification. The strange thing, though, for a group that saw its work as being the descriptive comparison of texts is that Descriptive Translation Studies, beyond the initial work of Levý and Popovič, has left us with no concerted typology of translation solutions. This merits some thought. Let me qualify the claim. Some will immediately retort that a complex model of solutions was developed by Kitty van Leuven-Zwart (1989, 1990), where translation shifts are indeed categorized on many levels, in an ostensibly bottom-up way (working from small shifts to larger ones). The very complexity of that model, however, means that there are theoretically no limits to the number and type of shifts that can be observed. And this complexity makes it very wieldy in the teaching situation and virtually unreadable in every other situation: once you start to note down the shifts incurred by translation, you quickly gather so much data that it is hard to make much sense of it. Devoid of reductive top-down categories (that is, without a typology), the model has not prospered. Others will reply that James S Holmes himself developed an explicitly top-down typology of solution types when categorizing the way translations can deal with a foreign verse form: mimicking it, eschewing it and three options in between (see Holmes 1970/2005). The result has a certain systemic beauty, approaching that of Torop (1995/2000), who also theorized the options for rendering foreign poetic forms. Yet it, too, is rarely cited and has spurned no continuation or further development. The reason for this relative absence is to be found, I suggest, in the literary approaches of the scholars concerned and their concomitant lack of concern for the problems of actually training translators. What we call Descriptive Translation Studies was originally intended to write a certain kind of literary

120

Translation Solutions for Many Languages

history; it had very few pedagogical interests. In fact, when Gideon Toury (1992) does address the issue of training, it is provocatively to suggest that students be taught to ‘break norms’: find out what the dominant norms are (there is no theoretical limit to their number), then stimulate creativity by challenging people to do the opposite. Indiscriminate revolutions need no careful typologies. The legacy of Descriptive Translation Studies has been countless studies, many of them doctoral dissertations, with many observations about translations past and present. These studies occasionally claim that their findings will be beneficial for training. Yet a stable metalanguage for solution types has not been one of those benefits.

6

Cold War Dalliance with Transformational Grammar

Chapter summary: From the late 1960s, solution types were seen as ‘transformations’ in both the United States and the Soviet Union, broadly following the development of transformational linguistics. This trend sought to connect with the early work on machine translation, especially as presented in Revzin and Rozentsveyg, at the same time as the mathematical metaphors offered a sense of scientific certainty and measurable progress. For the early Nida, the hope was that a message could be broken down into ‘kernels’ and then regenerated, with the various transformations in fact being solution types. Kade and Komissarov nevertheless saw transformations as happening between languages, not just within them. This interlingual transformational approach might speak to the much later developments in statistical machine translation.

I pedal back a little to consider the term ‘transformation’. We saw the Russians using it, sometimes more than ‘technique’ or ‘procedure’, when describing their solution types. Yet it is not a word we find in Fedorov’s 1953 book; it was not a word or concept that Loh would have picked up through Russian. Talk of transformations became frequent in the 1950s and 1960s, following the work of the American linguists Zellig Harris (whose Methods in Structural Linguistics appeared in 1951) and Noam Chomsky (whose Syntactic Structures was published in 1957). We thus find Harris cited in Loh (1958), later most significantly in Nida (1964) and then by Russians like Barkhudarov (1975), with ‘transformation’ taking on various meanings at several other pit stops along the way, including Kade (1968). Here I propose to follow that development with an eye on how it affected thought about solution types, particularly in the work of Eugene Nida. I am also interested in various ways transformations were interpreted, particularly with respect to the development of machine translation systems. I suspect that the

122

Translation Solutions for Many Languages

subsequent development of statistical machine translation can tell us something about the relation between structure and usage, and that this in turn might help retrieve some transformations from the scrap heap of old translation theories.

Why transformations? The historical fascination with the term ‘transformation’ was clearly due to the prestige of what became known as ‘generative-transformational grammar’, which at one stage promised to revolutionize not just linguistics but every human science that had recently been reshaped by structuralism. The approach nevertheless held a special appeal for theorists of translation. There are several reasons for this. First, most obviously, the development of computing capacity held out the prospect of usable machine translation, which was of extreme interest on both sides of the Cold War. Chomsky was funded by the US military1 precisely in the hope that his work would connect with machine translation, and one presumes that similar interests were pursued in the Soviet Union. An influential transformational approach was proposed in Revzin and Rozentsveyg’s Основы Общего И Машинного Перевода (Fundamentals of general and machine translation) in 1964 and is widely cited in the general translation theories that followed. Note the title ‘general and machine translation’, suggesting that what happened in machine translation might supply models for the rest. The citations of this work in the early 1970s map more or less onto the years when, as we have noted, the Russian term ‘equivalence’ stopped being used for word-for-word certainty – as it was used in Shveytser (1973/87) and Retsker (1974/2007) but not in Barkhudarov (1975) – and started to be applied as a feature of all solution types. In Retsker (1974/2007) there is a discussion of the two senses in which ‘equivalence’ was being used at the time: one sense in which a unit can have many possible equivalents, and a second in which there is only one authoritatively established equivalent. Retsker insists on the second sense.2 In Barkhudarov (1975) we find a more open discussion of the ‘new’, wider usage of the term (which could also have been drawn from Catford 1965). These discussions would seem to locate the terminological change with some precision. Second, transformational grammar posited that a finite number of rules could be applied in order to generate an infinite number of different sentences. This idea of ‘rule-governed creativity’ could be found in Chomsky in the mid



Cold War Dalliance with Transformational Grammar

123

1960s (1966: 27) and is implicitly referred to in Kade (1968).3 The idea should also have had a particular resonance in the Soviet context. If we consider that the Russian schools of thought in the 1950s had been split by debates over linguistic rules and literary creativity, with the literary side claiming that the rules killed creativity, now an approach based on creative rules offered a way the two sides could perhaps be brought together. It was at least worth looking at. Third, the early notions of transformation, carried across from mathematics, offered a sense of certainty in an age informed by the spirit of critical scepticism. These were the same years in which Quine announced the indeterminism of translation, arguing that the one utterance could be rendered in several different ways ‘which stand to each other in no plausible sort of equivalence relation however loose’ (1960/2013: 27). A similar challenge might be seen behind the theories that privileged text type or communicative purpose, since they also posited an initial plurality of possible renditions. Any theory that could restore a degree of unitary certitude, even at the expense of not dealing with semantics, was worth several looks (as indeed should have been Quine’s untheorized assumption that translation should somehow be ‘equivalent’). To grasp that appeal, consider the sense that ‘transformation’ has in algebra, where one side of an equation can be transformed into the other. For example: x(2 + y) = x∙2 + x∙y This kind of mapping produces two pieces of language that look different and involve different operations, but they clearly have the same overall value. What is more, the mapping process can go from one side to the other and then back again. That promise of reciprocal equivalence between conceptually different things should constitute a substantial challenge to indeterminism. There were two main ways of using this idea. Initial uses of the transformational approach tended to involve an idealist grasping of the thing to be translated: the start-text unit could be disambiguated and reduced to its constitutive units (sometimes known as ‘kernels’), these could be dragged across to the other language and then transformations on the target side could generate an acceptable equivalent translation. We thus find countless diagrams where the start-side process works down to easier, clearer forms, and target-side processes then work up, adapting those units to criteria of target-side usage. That approach emphasized the generative aspect of generativetransformational grammar. It was more in tune with Harris’s linguistics, which did not separate semantics from syntax (also in keeping with the tradition of Bloomfield’s structuralism). When Nida made use of transformations, it was

124

Translation Solutions for Many Languages

in this sense: the identification of ‘kernels’ was as much semantic as it was syntactic, and it offered the illusion of certitude. A second kind of usage, less idealistic, emphasized the transformational aspect. This was more in keeping with Chomsky’s insistence that he was dealing with syntax only, without semantics. On this view, transformations can be productive of new values (rather than repeat the same value) and can occur between languages (rather than simply within one language, either before or after the moment of transfer). This approach emphasizes the transformational side of the grammar, rather than the idealist generation of language from a fixed underlying form. This is the approach that was picked up in Kade’s commentary on transformational grammar. This second approach to transformations, which does not assume reduction to a ground, is suggestively similar to something we have met previously: Bally’s theory of ‘functional equivalence’. I offer the citation again: Pieces of a grammatical system can replace each other [peuvent s’échanger] because of their shared function, even when their semantic and stylistic values are not identical. Compare the following ‘functional equivalents’: the house my father owns, of which my father is the owner, belonging to my father, owned by my father and finally, my father’s house. (1932/65: 40)

Bally, I think, was analysing simple transformations, albeit without the term. And he was not doing so in order to express the certainty of any one ‘kernel’ version. On the contrary, he was celebrating the creativity of language, found in the way the affective values of the speaker could be expressed by choosing any one of these variants. As I argued above, all Bally had to do in order to produce a theory of translation was to see transformations as operating from language to language. He did not do that, but did anyone else?

Eugene Nida and the illusory certitude of kernels In 1964 the American linguist and Bible translator Eugene Nida proposed that generative-transformational grammar could help translators make decisions: A generative grammar is based upon certain fundamental kernel sentences, out of which the language builds up its elaborate structure by various techniques of permutation, replacement, addition and deletion. (1964: 60)

Nida’s early work had been as a structural linguist; his first book was on morphology. He reached transformations as a linguist, rather than on the



Cold War Dalliance with Transformational Grammar

125

back of previous theories of translation. As the above citation indicates, his thinking about translation solutions is directly derived from the generative part of transformational grammar. Nida proposed that the translator could extract the kernels from the start text, then apply the same or different techniques (‘transformations’) to generate the target text. The transformations could be like the ones commonly used in English (passives from actives, questions from statements, and so on), but they could also include the solutions that translators commonly use to achieve a ‘natural’ sounding text: changing the order of elements (called Permutation), changing word class (Replacement), explicitation (Addition) and implicitation (Deletion). The solution types are in fact the main models of transformation, and this seems to be the only context in which Nida offers a clear typology of what translators can do. The types were not pulled out of a hat: they came from linguistics (as indeed was the case of Transposition and Explicitation for Vinay and Darbelnet, which were drawn from Bally’s unannounced transformational grammar). In a later text, Nida and Taber seem to be offering a linguistic reason why there are only a handful of solution types. They have clearly bought into the argument that the ‘deep structures’ of sentences point the way to the universals underlying all languages, such that ‘languages agree far more on the level of kernels than on the level of the more elaborate structures’ (Nida and Taber 1969: 39). From there, it is a short step to dreamt keys to the human mind: ‘in all languages there are half a dozen to a dozen basic structures out of which all the more elaborate formations are constructed by means of so-called “transformations”’ (Nida and Taber 1969: 39). If we can get the kernels and the transformations, all translation will be easy. The actual kernels analysed as such in Nida turn out to have more to do with Harris’s syntactic semantics than with pure syntax. They look like simple structures that are basically used for disambiguation: the phrase ‘the fat major’s wife’ is not ambiguous if it is derived from the two kernels ‘The major has a wife’ and ‘The wife is fat’ (Nida 1964: 61) – presumably rather than ‘The major has a wife’ and ‘The major is fat’. So a simple transformation can remove the ambiguity and translation is then skating freely. One wonders why this is not simple componential analysis or basic pre-editing, but at the time it might have looked like a revolutionary theory. Despite our current doubts, Nida’s evocation of transformational grammar was a reasonably elegant way of explaining what was being translated: ‘kernels’ could partly replace problematic idealist assumptions about ‘ideas’ or ‘messages’. It also located solution types within a linguistic framework: the ‘procedures’

126

Translation Solutions for Many Languages

described by Vinay and Darbelnet are now seen as the ‘techniques’ or ‘adjustments’ needed for the transformations, and there was a reason for their number and variety. Unfortunately, the beautiful ideas published in 1964 were promptly quashed by Chomsky’s pronouncement, published the following year, that deepseated universals did not ‘imply that there must be some reasonable procedure for translating between languages’ (Chomsky 1965: 30). Linguistics was not going to support this foray into translation – strangely, like Bally, Chomsky didn’t want to go there. Why Chomsky wrote that sentence remains a point of conjecture. For George Steiner (1975/88: 111), the pronouncement recognizes the arbitrariness of the linguist’s construal of any structures and was thus covering over a major weakness within Chomsky’s approach at the time.4 For Gentzler (2001: 50), Chomsky’s pronouncement is more like a wise word of caution about what cannot yet be done with a developing linguistics. For Melby and Warner (1995: 179), more interestingly, it is a claim about the non-reversibility of transformations: while there can be a reasoned generation from deep structure to surface structure, there is no guarantee of a rational movement in the opposite direction. For example, ‘the fat major’s wife’, as a surface structure (which is what the translator is looking at), remains ambiguous because there is more than one possible underlying structure. If you have no prior knowledge of what is meant, the attribution of disambiguating kernels does not in itself remove the ambiguity. If you do know what the underlying structure is, then you can go from there to the surface, but if you only have the surface, you cannot reach the underlying structure with any certainty. This non-reversibility returns us to the problem of indeterminism, showing the transformational approach to be an idealism with very limited applicability to translation. In Nida’s later work (especially in Nida and Taber 1969: 39–40) we find insistence that the kinds of ‘kernels’ of interest to translators are actually not at all deep-seated. In fact, they come to be recognized as simple componential analysis. The translation theorists were trying to use transformational grammar, but it just was not going to work. The great adventure in transformational linguistics was very short-lived. The appeal to transformations was then picked up by Vázquez-Ayora (1977), who set out to unite the ‘American school’ of transformational linguistics (basically Nida) with the ‘Franco-Canadian school’ of translation theory (basically Vinay and Darbelnet). The practical result, as we shall see, was a comparative stylistics very much in the Vinay and Darbelnet idiom, with very



Cold War Dalliance with Transformational Grammar

127

little trace of anything being drawn from a newer kind of linguistics. For a more consistent engagement with transformational linguistics, we have to return to Russian.

Russian transformations and a new sense of ‘equivalence’ The Soviet translation scholars of the 1970s consistently used the term ‘transformations’ to describe translation solutions (for example Retsker 1974/2007; Barkhudarov 1975). Shveytser (1973/87: 51) claims that they were not borrowing from American transformational grammar, which apparently only dealt with syntax, whereas the Soviet typologies paid serious attention to meaning as well. The prime reference for the Russian theorists, as we have noted, was Revzin and Rozentsveyg (1964), not Chomsky. The only problem with Shveitser’s argument is that in the work of Zellig Harris, prior to Chomsky, we also find semantics dealt with from a transformational perspective, and there are cases of Russian theorists working from Harris, in fact dealing with the same kind of problem as Nida. When presenting Addition and Omission as solution types, Barkhudarov (1975: 221) certainly cites the work of Harris. Harris’s transformational grammar elegantly explains Implicitation as ‘[t]he zeroing of redundant material. [This transformation] drops words from a sentence, but only words whose presence can be reconstructed from the environment’ (Harris 1972: 133). The definition works clearly enough for syntax, as in the case of ‘He plays violin and she [plays] piano’ (Harris 1972: 300, 681), where the second ‘plays’ is redundant and can be elided in English but can easily be recuperated, both in English and in translations into other languages. Working in reverse, Explicitation can be seen as the transformation whereby such ‘zeroed’ elements are recuperated. Harris, however, claims that a similar process also allows the Implicitation of semantic elements, and this is actually the example that Barkhudarov cites. I give the whole citation for the sake of reminding us all of the kind of prose that the search for certainty once entailed: The Xap of a particular word in a structure is the member (or members) of X which is the main co-occurrent of that word in that structure, for the given subject matter … In many circumstances … the Xap can be eliminated; other members of X cannot. There is no loss of information, for the absence of the X which is required in the A X B structure (whose presence is evidenced by the remaining

128

Translation Solutions for Many Languages

A B), points to the Xap which is determined (up to local synonymy) by the individual words of the Ai Bi. Thus from violin-prodigy we generally reconstruct violin-playing prodigy and from violin-merchant we generally reconstruct violinselling merchant. In any case, the grammatical reality of Xap lies in the fact that it and no other X can be zeroed in this way (or that Xap is the only X that occurs in the given position). (Harris 1972: 134–345; cf. Nagal 2008: 15)

Note that the term ‘grammatical reality’ is here associated with ‘it and no other X’, which is mathematical certitude. Grammar, it seems, deals only in 100 per cent certainties here; there is no room for lesser probabilities; there is no concern for the quantities of usage. I find this very problematic. In fact, I have elsewhere claimed (Pym 2005) that the term ‘explicitation’ (and consequently ‘implicitation’) should be reserved for grammar words only, the stuff of syntax, since the rest is probabilistic, a question of being ‘in the know’. The semantic kind of ellipsis analysed in Harris and Barkhudarov is decidedly less convincing than in the case of strict syntax: the elision (or implicitation) of the verb, or of cohesion markers in the case of Blum-Kulka (1986), can be seen as rule-governed; those rules can be considered more or less co-extensive with the language in question; their acceptability can be based on the judgements of people assumed to be competent language users. However, that seems to me not to be quite the same thing as claiming that the verb ‘to play’ is somewhere within the deep structure of ‘violin-prodigy’, since the recuperation of that semantic value requires a more individual act of interpretation: one has to know beforehand the kinds of situations in which this particular term is used; it is not possible to recuperate the value by applying a rule learned on the basis of quite different terms in quite different situations. For this kind of semantic implication, one needs local knowledge: at best it could involve a ‘local grammar’.5 The analysis of ellipsis is important because a tendency to explicitation has been posited as a translation ‘universal’ (since Blum-Kulka 1986): translations tend to explicitate more than non-translations do. If a start text says, ‘He plays violin and she piano’, translations of that text will tend to restore the elided ‘plays’. Similarly, according to Barkhudarov, the phrase ‘violin virtuoso’ can be translated as ‘master violin player’ or some such (most translations into Chinese do indeed include a character meaning ‘play’).6 These two processes are nevertheless not the same. The first is formally stylistic – it produces an elegant text; the second requires the translator both to interpret the phrase and to make assumptions about the text-user’s experience and local interpretative capacity – it has to be construed for a particular kind of use. The difference between these processes is hidden in Harris’s passage above in the slide from



Cold War Dalliance with Transformational Grammar

129

‘co-occurrent of that word in that structure, for the given subject matter … we generally reconstruct’ (hence based on the ‘general’ frequency of occurrences for a particular language user) to ‘the grammatical reality’ and ‘the fact that it and no other X can be zeroed in this way’ (presumably now for all language users, as a part of grammar). It is dangerous to assume that ‘we generally …’ means ‘all speakers must …’. But this is precisely what the semantic side of transformational grammar is given to do. One simply cannot be sure that all language users will recuperate the same ‘hidden’ semantic values. This means that the ‘decoding’ side of translation does not have the same degree of certitude as the second ‘recoding’ side – in fact it means that the whole code metaphor is undermined by hermeneutics. And that is once again the root of many problems in the application of transformational grammar to translation. So we return to the problem we had when reading Bally: exactly what is meant by a ‘rule’? If the initial promise of transformational linguistics involved some kind of certitude, to what extent was it able to deliver? Perhaps no better than Chomsky was able to help the US military with its understanding of the world.

Two proposals by Kade When Otto Kade writes on transformational grammar in 1968, he is aware of Harris, Chomsky, Lees and Saumjan, and he comments on the way Revzin and Rozentsveyg (1964) had applied transformational grammar to machine translation. But he has a rather different proposal to make. Kade maintains that the transformations are not just within one language but between languages, and that the Russian theorists of machine translation were thus wrong to tie transformations to the identification of invariants. He might have made the same criticism of Nida. The example he uses to make this point is simple: within one language, a negative sentence can be seen as a transformation of a positive one, but this does not hold in the case of translation, ‘since the transformation contradicts the concept of translation’ (1968: 99). This sounds a little silly, since Kade is playing on different uses of the term ‘transformation’. For the kind of algebraic equation that we began with, the formation of a negative would be an incomplete mapping operation, although a double negative could restore the balance in English: x = − (−x)

130

Translation Solutions for Many Languages

For most of transformational linguistics, though, the formation of the negative is a productive application of a rule, just as there are rules for the formation of nouns from verbs, tenses from infinitives and so on. You apply rules to produce new forms, and the transformation is considered successful for as long as the relevant linguistic values remain the same (in the case of negation, all other relations except the negated item should ideally remain the same). Yet translation, says Kade, is contradicted by such a change in basic meaning. It follows that there must be two kinds of transformations: those that are productive within a language (where meanings can be changed) and those that maintain equivalence between languages (where meanings cannot be changed). The translation theorists tend to agree: if you want to use negation as a solution (as we have seen in Chinese examples), it usually has to involve a double negation of some kind. That distinction would remain trivial if it did not lead to the more systematic idea that transformations can be applied directly from one language to another. This is where Kade says that Revzin and Rozentsveyg went wrong, although I’m not entirely sure Kade gets it right. Here he offers a few clues as to how transformational grammar might be applied to translation: When there is previously existing semantic-functional equivalence, then translation can be seen as a case of interlingual substitution. When there is no such equivalence, then the translation can first be approached through a 1:1 transfer of linguistic elements, in which case one can speak of transformations that can take place within the start language. When the transformed material is then in a 1:1 correspondence (and that must be the goal of the transformation), this can provide the basis for interlingual substitution, which can be achieved through transformations in the target language … (1968: 100)

There is a certain beauty in this, if and when you can remember what Kade’s categories mean. In the first case, equivalence exists already and no transformation is necessary: ‘five’ is ‘fünf ’ is ‘cinco’ is ‘五’, and that all happens between languages. In the second case, I think you are going to use transformations to get the start text to a point where it can be transferred in a simple one-to-one way. Let’s see. To pick up the classical example, this is what Kade might have in mind (here using simple grammatical labels instead of any formal logic): 1. He swam across the river. 2. [subject masculine] [verb swim, past tense] [object] [across]. 3. [subject masculine] [across in verb, past tense] [object] [verb swim as adverbial]. 4. Il traversa la rivière à la nage.



Cold War Dalliance with Transformational Grammar

131

The transformations of interest here are in the passage from 2 to 3: ‘across’ is expressed in the verb and ‘swim’ is expressed in the adverbial – classical transposition, of the chassé-croisé variety (Vinay and Darbelnet 1958/72: 105–6). And once you are there, you just fill in the blanks with your favourite target-language words. It is not hard to see all the solution types as transformations of this kind. My problem with the analysis, though, is that it imitates precisely what Kade criticizes in the early theorists of machine translation. It proposes a traditional V-movement (start text to essential structure to target text, also recognizable in appeals to ‘deverbalization’). It seems to me more psychologically plausible that the ‘pre-existing equivalents’ enter the scene as early as possible, if only to help trigger the syntactic transformation: 1. He swam across the river. 2. Il [verb past] la rivière [across]. 3. Il [across as verb past] la rivière [swam as adverbial]. 4. Il traversa la rivière à la nage. This is entirely compatible with Kade’s proposition and it might even be what he was getting at. The difference between the two sequences is perhaps minor (in theory the replacement of one-to-one equivalents could happen at any stage, both for human translators and for machines). But the resulting picture, in the second sequence, is far more like something that is actually happening between languages, recognizing the stage of what second-language acquisition theorists call an ‘interlanguage’, and visibly distinguishing translational transformations from the kind that are considered operative within single language systems only.7 This identification of translational transformations leads to a key argument that is rather closer to my theoretical concerns. Kade, we have seen, considers that the point of the transformations is not to reach any deep structure, underlying truth or definitive banishment of indeterminism. Far from it: the only goal of the transformations is to reach a point where straight translations become easy. This is an important de-idealization of transformational grammar. From there it follows that ‘the semantic content of a structure does not form an invariant core by transformation, but more through the regularity with which a certain end structure can be derived from a certain start structure under certain conditions’ (1968: 99, emphasis mine). This is a major reservation. Kade recognizes that the rules only go so far: the rest is for the regularities of use. If there are rules, they come from the Gesetzmässigkeit of translation, as in the title of Kade’s book, which I have translated as ‘regularity’ even though the German word also has ‘rule’ (Gesetz, indeed ‘law’) embedded within it.

132

Translation Solutions for Many Languages

So Kade (1968) formulates two useful proposals: transformations can operate between languages, rather than just within them, and they can be based on regularities of use. Those ideas would turn up elsewhere.

Transformations and probabilistic usage Vilen Komissarov (1972) explains why he calls all solution types ‘transformations’: It is extremely interesting to examine translation as a transformational process of a certain kind. In studying the formal and contentual relations between the original and the translation, one cannot help but notice their resemblance to relations between intertwined structures as analyzed by transformational grammar. (1972: 8, trans. Douglas Robinson and Svetlana Ilinskaya)

Fair enough. Komissarov then notes the use of ‘kernels’ in American transformational grammar and starts to reinterpret solution types in terms of applied transformations. In doing so, he refers only to Russian sources – this was the Cold War – and his prime frame of reference is machine translation, citing Revzin and Rozentsveyg (1964). Yet his view of linguistics does not stop there: By examining some formal units of the original and translation as intertwined transformations, the theory assigns a significant role to the comparative study of various linguistic forms between which relations of translational equivalence can be established. The study of types of correlation between such elements allows us to uncover absolute and statistically most probable correspondences or ‘modes [способ] of translation’, which are marked in the translated text wherever certain of the original’s formal units are present. (emphasis mine)

So here we have the two ideas found in Kade (apparently not cited here): specifically translational transformations, and rules being based on use. Komissarov’s probabilistic approach need not have been taken from Kade: it was, after all, in keeping with Retsker’s initial premise of basing the linguistics of translation on ‘regular correspondences’ right back in 1950. What is new here was the opening to statistics in the context of machine translation. In hindsight, Komissarov’s take on transformations was quite prophetic.



Cold War Dalliance with Transformational Grammar

133

Translational transformations and machine translation For many decades, the linguists working on machine translation developed transformational rules, often of the kind that we have seen described by Kade. In the ‘transfer’ model of machine translation, the rules can be for straight replacement (for simple lexical items) or for transformation (particularly at the points where syntactic systems do not correspond). That model was necessarily restricted to language pairs. If I write rules for English–German translation, most of them are unlikely to be of great help when developing a programme for English–Arabic machine translation. Then again, the ideological promise of transformational grammar was that deeper and deeper structures would apply to more and more languages and a universal grammar would result. Yet machine translation has not waited for that to happen. The most significant advance in machine translation has come not from rules but from mathematics. Statistical methods were developed by IBM from 1993 and, as most of us are very aware, have been used by Google’s free online service from 2007, particularly thanks to the work of the computer scientist Franz Joseph Och. Statistical machine translation is based not on grammatical analysis but on the frequencies with which mathematically identified ‘phrasemes’ are associated in a database of paired segments: the mathematics calculates the probability with which, on the basis of prior translations, a particular solution is used to solve a particular problem. This does not mean there is any need to throw away the rules that have been written and that work; but it does mean that new probabilities are being derived from what translators do, rather than from what linguists can formulate. In short, rules are being derived from usage. To give an idea of how statistical machine translation works, use Google Translate to render the text ‘duelos y quebrantos los sábados’, which describes what Don Quijote ate on Saturdays. Put the words in one by one: Duelos: duels (along with suggestions of mourning and grief, since several ‘kernels’ are possible).

This is not much good. ‘Duels’ might be the most statistically frequent translation, but you can’t eat duels and the machine doesn’t know this – it lacks semantics. We very probably have the wrong disambiguation here. But keep going: Duelos y: duels and

134

Translation Solutions for Many Languages

‘Y’ means ‘and’. It is a straight replacement, but a very small solution. Keep going: Duelos y quebrantos: duels and losses

Again, this is straight dictionary-type matching; it is not really helping. This particular phraseme is not being transformed correctly and it is hard to envisage a rule that could get us from here to anything edible. But wait, add just one more word: Duelos y quebrantos los: scraps on

What happened? The plural masculine definite article (‘los’) has been translated by the preposition ‘on’ (what?) and the three words ‘duels and losses’ have shrunk to just one: ‘scraps’. This makes no sense in terms of any kind of linguistic transformation, but it does make sense in term of Don Quijote’s miserable cuisine. What has happened is that the four words in Spanish only appear once in the bilingual database, and that one occurrence is from a previous translation of Don Quijote. It may not be the best possible translation, but it is a lot better than what the transformational rules were doing. Further, as an admirable paradox, this instance of statistical machine translation gives us a solution visibly invented by a human translator. In statistical systems, the solutions not only come from usage, they come from translational usage. By the way, as a counter-test, if you feed ‘Il traversa la rivière à la nage’ into Google Translate (in July 2015), you get ‘He crossed the river to swim’, and a similar literalism appears when you try it going from English to French. Why does the system fail in this case? Well, if you do a search for the French sentence, virtually all the hits are from Vinay and Darbelnet’s example. The utterance is somehow not in current usage (except in the linguistics of translation) and so the statistics fail. But Vinay and Darbelnet never set out to explain what translators do – their concern was with what translators should do. Their descriptions, to that extent, are all counterfactual. If we now look back at what theorists like Kade and Komissarov were saying about transformational grammar, their comments do indeed seem prophetic. The use of transformations between language systems, rather than within them, is something like what we are finding. And then, the dependence of such rules on regularities in usage, and thus on analysable quantities, is precisely what allows statistical machine translation to work. The only problem is that what actually happens in those movements between languages depends on probabilities and thus on continuous data. The kind of categorical thought required by



Cold War Dalliance with Transformational Grammar

135

transformational grammar certainly applies to the linguistic changes, but it can then no longer pretend to represent cognitive processes.

A place for transformations? Writing a few years later than our references in this chapter, Wilss (1977: 231–2) did not consider transformational linguistics to be the way forward. He gave three main reasons: (1) transformational linguistics had not described movement between language systems (which was largely true, although Kade tried to see a way to do precisely this); (2) the simplicity and elegance of transformational trees is simply not possible in cross-language operations (which may not necessarily be true); and (3) there is no proof that different linguistic expressions are actually generated from the same schema (which is surely a problem both within a language and between languages). As I noted with respect to Nida’s use of ‘the fat major’s wife’, transformations only work if the translator already knows exactly what has to be translated – they can solve problems of expression, but not of understanding. In fact, when you look at it, this is true of all the typologies we have been looking at: they all assume exact knowledge; they do not address situations of cognitive doubt. I nevertheless suspect there is still a place for transformations in measuring degrees of cognitive effort (as we shall see later) and in translator training, specifically with respect to syntax. Indeed, it is in a textbook for students that transformations most clearly flourish into solution types (as we shall see now).

7

Forays into Romance

Chapter summary: The most direct legacy of Vinay and Darbelnet was in the work of Gerardo Vázquez-Ayora, who in 1977 adapted their categories to translation between English and Spanish. His preferences were clearly for ‘exactitude’, following closely the intentions of the text while allowing for differences in the stylistics of text types and languages. On the level of theory, VázquezAyora claimed he was integrating comparative stylistics with transformational grammar, and a later text of his opens up to early poststructuralism. VázquezAyora’s relatively eclectic approach might be contrasted with the 1977 work of Scavée and Intravaia on French–Italian translation, where the preferences of the Italian language are attributed to three ‘complexes’ illustrated by great Italian writers of the past. The Italian nationalism has happily had as little impact as did Vázquez-Ayora’s attempt at an integrated typology.

So we have Vinay and Darbelnet working from Bally in 1958, a minor excursion in Malblanc, then Fedorov leading to Loh, also in 1958, initiating a certain lineage in Chinese. Prior to all that, things were happening in Soviet cultural spaces, then in Central Europe. In the meantime, the American Nida picked up the transformational approach, which independently influenced Soviet theorists in the 1970s. So where does the story go from there? I return to the beginning: Vinay and Darbelnet are driving along. They did indeed reach Montreal and there was a certain development of their ideas on their adopted home turf, in Canada, with articles by Bart and Clas. Yet the immediately following node in that developing network would appear further south, in Washington and in Spanish, under the pen of Gerardo Vázquez-Ayora, who was also trying to be a transformationalist.

138

Translation Solutions for Many Languages

Gerardo Vázquez-Ayora attempts integration Gerardo Vázquez-Ayora’s Introducción a la traductología: Curso básico de traducción (Introduction to Translatology: A basic course in translation) was published in Washington by Georgetown University Press in 1977. It is rarely commented on beyond publications in Spanish. The text explicitly works from Vinay and Darbelnet’s categories and applies them to a multitude of Spanish– English examples. As such, it is eminently pedagogical in purpose. There is, however, a strong theoretical strand in Vázquez-Ayora’s thought, much of which comes out in his slightly later article on ‘semiostylistics’ (1979). Here I comment first on the textbook, then on the theorization. Frustratingly little is known about Vázquez-Ayora himself. A scarcely reliable source (a lone Facebook site) indicates that he worked in Ecuador and as a ‘reader/reviser of translations at an international organization’, and we know that his textbook was published at Georgetown, where a Division of Interpretation and Translation had been established in 1949. Vázquez-Ayora is said to have been active in the Fédération Internationale des Traducteurs, but their records do not show this. His publications indicate a serious interest in literary translation, specifically of Latin American authors into English, while his opinions on accuracy and the limits of translation sound like someone who has spent a lot of time revising junior translators. Beyond that, he remains an international man of mystery, so far. The Introducción of 1977 comprises some 400 pages and has two parts. The first is a veritable course in general linguistics, of both the sentence and the text, as was being developed in the United States at that time, with special emphasis on generative-transformational approaches and the semantics of the day (the big names are Chomsky, Fillmore, Katz and Fodor). The second part includes a last chapter on the ‘general translation procedure’, dealing with the mechanics of reading the text, using dictionaries and so on – it is full of sound advice but of little consequence here. What is left in this second part is 100 pages or so of ‘technical procedures’, very much in the Vinay and Darbelnet sense, arranged into categories that should look familiar: Literal Translation Transposition Modulation Equivalencia Adaptation



Forays into Romance

139

Amplification/Condensation Explicitation Omission Compensation There are sub-sets under each of these heads and countless examples, making this a rich and engaging work in the tradition of the day (although how anyone could use it as a textbook in class is hard to say). Looking at the above list, it is easy to spot the Vinay and Darbelnet terms, with the last four of them (from Amplification downwards) having been elevated from ‘prosodic effects’ to the level of main categories, as indeed we have found in the Chinese tradition. The categories of Borrowing and Calque have disappeared, since Vázquez-Ayora considers that they do not properly concern translation – they are the work of terminologists, apparently. This might be true if you work in a large international organization where there are separate departments for such things. There are nevertheless plenty of translators left standing on the frontline of linguistic battles and who confront terminological problems on a daily basis. Like Vinay and Darbelnet, Vázquez-Ayora starts from the concept of Literal Translation. This is not just out of convenience: it is the necessary first step for the application of transformations. The progressive transformations of literalism (going down the list) are together termed ‘oblique translation’ (traducción oblicua), as in Vinay and Darbelnet. When writing in English a few years later, Vázquez-Ayora uses the rather more sexy term ‘decentered translation’ (1979: 206), coined in the midst of references to French poststructuralism and probably not deserving to have been forgotten so quickly. As you use solutions from further down the list, you are ‘decentering’ the assumptions of literalism. On the concept of Literal Translation, Vázquez-Ayora is disarmingly opinionated: Practicing translators very commonly believe there are two kinds of translation, one ‘literal’ and the other ‘free’. We have seen that Literal Translation is one of the translation procedures. But ‘free’ translation most definitely does not exist. The first criterion for all translations is exactitude; if it is not accurate, it is not a translation. Versions that are amplified, corrected or commented, or works of exegesis, are all quite different things from translation. (1977: 265; my translation here and throughout)

This is where Vázquez-Ayora sounds like someone who has spent years revising novice translators: accuracy is everything and don’t try to think otherwise. Translators do not have the ‘capacity’ (‘facultades’) to attempt anything more

140

Translation Solutions for Many Languages

or less than accuracy, he assures us, which could mean they lack either the skills or the authority, or probably both. At the same time, however, generalized literalism is itself considered ‘false translation’ (1977: 251), so the translator has to know precisely when and where to deviate from it. Decentred translation (traducción oblicua) should thus be understood as the gradual application of procedures that allow the translation to be ‘faithful to the intention of the author and the genius of the language’ (1977: 290), presumably praying that the two coincide. Note here the double reference to author and language, both of which are implicitly personified. Stylistics, in Vázquez-Ayora, pays ideological dues to the study of an author and ‘intentions’, as in traditional literary stylistics, but by no means abandons the notion of each language having its own preferences, if not intentions. When it comes to the actual solution types, the fact that the examples are isolated phrases and sentences means that the personified language inevitably wins out over any other ideological factor – if ever there were a real desire to look at an author’s style or intentions, Vázquez-Ayora’s mode of exposition allows no space for it. As in Vinay and Darbelnet, Vázquez-Ayora struggles to insist that stylistics is the domain of options, rather than of the obligations of grammar. In theory, this means there is no need to mention obligatory transpositions such as the default position for adjectives being after the noun in Spanish (and before the noun in English), or classical traps for anyone on a romantic holiday in places Hispanic: ST: I like you. TT1: [Yo] te quiero. (‘I you want’, meaning ‘I love you’) TT2: Me gustas. [To me you are pleasing.] Any reasonable speaker of these two languages will render ‘I like you’ as ‘Me gustas’ (TT2) and there should be no question of alternatives or stylistics. At the same time, the relations between these two phrases can clearly be described in terms of solution types and it is hard not to do so. Come to think of it, there is so much going on here that first-timers might be advised not to risk falling in love: Spanish has no active verb with the meaning ‘to like’, English has no active verb with the meaning ‘gustar’, so the alignment of the two is destined to look like an active–passive transformation. Then the expression of the first person in Spanish is optional (since it is expressed in the verb declension), so its surface-level absence is bound to look like implicitation. And of course, the slide between wanting and loving is a semantic problem that concerns more than linguistics.



Forays into Romance

141

The simple fact that the solution types can indeed explain both the obligatory and the optional variants invites a constant crossing of any line that might separate one domain from the other. It is hard to keep them separate, and Vázquez-Ayora effectively forgets he was ever trying to separate them. The best way to show this is to look at each of the categories in turn. Transposition is little different from usages in Vinay and Darbelnet. Some examples are explicitly optional and unproblematic. In the following case there are two possible translations, one literal and the other with transposition (the adjectival noun phrase ‘of interest’ can become a verb), and both are fine: ST: Of interest to Latin America TT1: De interés para América Latina TT2: Que interesa a la América Latina (1977: 271) [That interests Latin America] The obvious question to ask is why anyone would want to produce the second transformation, which requires more work (transformations require conceptual effort) for an uncertain gain. The concepts of work and gain are unfortunately not in evidence in Vázquez-Ayora – we simply learn about what is possible. Other examples involve interesting degrees of sophistication in the transformations themselves: ST: He was never bothered again. TT: Nadie volvió a molestarlo. [No one repeated to bother him.] (1977: 271) Here Transposition is found in the adverbial ‘again’ being expressed in the verb ‘volvió’, and perhaps in the adverbial ‘never’ becoming the noun ‘nadie’ (no one). It might be grammatically possible to try something literal like ‘No fue molestado de nuevo’, which more or less follows the English word order but would certainly not be in frequent usage. There is, moreover, a more general transformation here of the passive into the active, which might technically count as Modulation, with the focus shifting from the person bothered to the potential botherers. Some examples are intriguing cases of what Vázquez-Ayora calls Crosstransposition (for Vinay and Darbelnet’s chassé-croisé), where semantic elements change both place and word class: ST: The door flew open. TT: Se abrió la puerta con el viento. [Opened itself the door with the wind.] (1977: 282)

142

Translation Solutions for Many Languages

A whole lot of different things are happening here, including a precarious explicitation of the verb ‘flew open’ (how does Vázquez-Ayora know it was the wind and not a violent intruder?). Cross-transposition is also seen in the following example, where the sense of the final adverbial ‘through’ in English is expressed in the fronted verb ‘abrirse’ in Spanish: ST: He elbowed his way through. TT: Se abrió paso a empellones. [He opened way by elbow pushes.] (1977: 282) Here is the explanation offered for these examples: Each language dissects and analyzes objective experience in its own way. Spanish considers the main point to be the action, which it therefore entrusts to an important element, the verb. This same point is only a secondary aspect in English, which English relegates to the particle … (1977: 282)

This is the kind of explanation one finds in the tradition of comparative stylistics (cf. Vinay and Darbelnet’s example of ‘He swam across the river’). Can one really say that Spanish is more focused on actions (and thus verbs) than English is? Taken at face value, these examples seem more likely to ensue from the asymmetric distribution of linguistic resources: English has phrasal and prepositional verbs (Spanish does not); Spanish uses optional inversion very freely (English does not); Spanish can have an implicit proform as subject (English cannot). And some of these differences might in turn be derived from the distinction between a language with strongly inflected verbs and another with weakly inflected verbs – they have different ways of solving the same problems. Does that mean Spanish likes actions and verbs more than English does? Vázquez-Ayora’s explanation of Cross-transposition in terms of the ‘genius’ or ‘spirit’ of a language (1977: 85–6, 292) does seem to be a conceptual leap of faith. Modulation is a change of perspective or of metaphor, without altering the referent (1977: 291). It provides another occasion for trenchant opinion: ‘The translator who does not use Modulation is not a translator; the efficient use of Modulation tests the translator’s imagination, sensibility, expressive capacity and ingenuity’ (1977: 293). I note that Vinay and Darbelnet had more moderately described Modulation as the ‘touchstone of a good translator’ (1958/72: 233; 1958/95: 246). Some of Vázquez-Ayora’s examples are intriguing, as in the replacement of one body part with another:



Forays into Romance

143

ST: To brush shoulders [to have known each other casually in the past] TT: Codearse (1977: 297) [To be at each others’ elbows – to be in proximity] Without clear context, one wonders if these two expressions really do refer to the same thing. Other examples are even more worrying in their assumption of an indisputably common referent: ST: To have second thoughts TT: Cambiar de idea (1977: 294) [To change opinion] ST: I took the job from my friend. TT: Mi amigo me cedió el trabajo. (1977: 298) [My friend ceded the job to me.] The first example here changes the idea, I suggest, since ‘to have second thoughts’ surely means ‘to have doubts’? And in the second example, the two sentences need not have the same meaning: the actions are carried out by different agents, and in English the job could have been taken without the friend’s consent. This further adds to the growing suspicion that perspective and metaphor are not neutral with respect to meaning. Other examples here would belong to Loh’s Negation. The first one might be familiar to visitors to Spain, surprised when they are asked to ‘tranquilize themselves’: ST: Don’t get so excited. TT: Tranquilízate. [Calm yourself] (1977: 299) ST: Blind flying TT: Pilotaje sin visibilidad [Piloting without visibility] (1977: 296) And then there are examples included here that simply presume too much or could easily belong to other solution types: ST: That’s the answer. TT: Esa es la solución. (1997: 295) [That’s the solution – but surely it could be a respuesta, an ‘answer’, to many kinds of questions?]

144

Translation Solutions for Many Languages

ST: Allowable error. TT: Error admisible. (1997: 297) [Allowable error – why is this not just Literal Translation?] ST: East Coast TT: Costa del Atlántico (1997: 302) [Atlantic Coast – surely a case of Explicitation and nothing more?] In short, this is where the wheels start to fall off. The examples indicate so many different things, sometimes including a remarkable absence of transformations, that it is hard to imagine any learner leaving these pages with a clear idea of what ‘Modulation’ means. Equivalencia is Vázquez-Ayora’s version of Vinay and Darbelnet’s équivalence. Citing Malblanc, he describes the category as ‘a Modulation that is lexicalized’ (‘que se lexicaliza’) (1977: 314), which might be helpful if anyone has understood what ‘Modulation’ means in this system. Here it appears to refer to standardization in the form of idioms. As such, it is a question of degree rather than of category. The relative lack of distinction is clear in the following example, presented as equivalencia but with a change of body parts that looks pretty much the same as the shoulder/elbow example above: ST: To pull somebody’s leg. TT: Tomar el pelo a alguien. (1977: 316) [To take someone’s hair] In short, we might want to get rid of this category altogether. Adaptation is as in Vinay and Darbelnet: the colour white means purity in the West, death in the East. Vázquez-Ayora seems quite keen on gaining naturalism in this way: ‘Beyond the fields and cases in which it is necessary to retain a “foreign element”, each missing adaptation obliges the reader to enter a reality that is strange and false [una realidad extraña y falsa]’ (1977: 330). So things that are strange are necessarily false? Vázquez-Ayora cites examples from Nida, but he tries a few from his own brew: Strange and false is certainly the impression that Latin Americans have when they read a North American novel where the babysitter is rendered indiscriminately as a niñera. The father and head of the household is thus romantically engaged not with a young university student but with a live-in maid who looks after children. The North American concept here should have been handled with wide-ranging Adaptation. (1977: 325)



Forays into Romance

145

So what might acceptable Adaptation look like in this case? Vázquez-Ayora does not tell us. The historical solution has actually been to adapt the Spanish language itself to the needs of rising middle classes in Latin America, often by using babysitter as a Spanish word, perhaps with an explanation on first usage when in a text – but why would that be Adaptation? It would seem to be precisely the solution that Vázquez-Ayora wanted to avoid! The solution recommended by the Real Academia Española (2005), similarly systemic, is indeed to use niñera/niñero, adding the phrase ‘por horas’ (paid by the hour) in cases of possible misunderstanding – an elegant solution, but surely just a case of Explicitation? Yet another solution has been found in Spain, where the earlier development of middle classes in need of childless nights-out actually gave rise to a new word for babysitters – canguro (yes, kangaroo). But should that wonderful metaphor be classified as Adaptation? Amplification/Condensation: Vázquez-Ayora is aware that translation from English into Spanish commonly involves an explicitation of semantic values and syntactic relations, making for longer texts (1977: 336). His examples, though, seem rather petty, boiling down to an optional word or two: ST: To speak aloud. [Better, ‘to speak out loud’?] TT: Hablar en voz alta. (1977: 337) [Speak in loud voice] ST: To endanger. [Could be ‘to put in danger’?] TT: Poner en peligro. (1977: 339) [To put in danger] ST: The night express for Birmingham. TT: El expreso nocturno con destino a Birmingham. (1977: 342) [Could be ‘a Birmingham’.] [The night express with terminal in Birmingham.] As I have tried to show between square parentheses, none of these examples justifies a systematic difference between English and Spanish. Vázquez-Ayora also allows that there can be Condensation (the opposite of Amplification) when going from English into Spanish. The examples (1977: 346) are nevertheless English bureaucratic prose that can easily be shortened in English, rendered in language that could easily be made much more embellished in Spanish. There are no rules here, just some bad writing. Explicitation: This is the same as in Vinay and Darbelnet, but the examples are not eloquent:

146

Translation Solutions for Many Languages

ST: We’ve seen the great art. TT: Hemos visto las grandes obras de arte. (1977: 352) [We have seen the great works of art.] The use of ‘grandes obras’ (‘great works’) is clearly a case of Explicitation, but it might also be Amplification. And for that matter, there seems to be no reason why the inclusion of ‘con destino a’ in the above case of the Birmingham train should not equally be Explication rather than Amplification. Indeed, Explicitation can be found all over the place. Omission/Over-translation: It seems strange to find these terms here. Vinay and Darbelnet did not use the term ‘omission’; Loh did use it, but in the sense of ‘implicitation’. So does Vázquez-Ayora, a stickler for accuracy, really allow the translator to leave things out? Far from it. The explanation here refers to adjustments to what is ‘natural’ in different languages, so it could just as easily be contraction or implicitation. One example is a piece of verbose bureaucratic prose: ST: The failure to act on the part of the committee. TT: La comisión dejó de actuar. (1977: 362) [The committee stopped taking action.] This example is a nightmare. First, the English phrase is potentially ambiguous: it could mean the committee failed to act, or that someone failed to act on behalf of the committee. Second, the Spanish is also ambiguous, but in a different way: the committee could have stopped taking action, or it could have refrained from acting. And third, it is not hard to envisage Spanish bureaucrats producing prose that is just as verbose as the English: ‘La falta de medidas por parte de la comisión’. Opaque prose is a bane in both languages and one would hope it could be tackled as such. Vázquez-Ayora has admirable stylistic preferences in this regard, but they are ill served by his overlapping categories. Compensation: The description here is great: ‘What is lost in one place or part of the text can be recuperated or compensated for in another’ (1977: 376). Yet some of the examples are perplexing: ST: The atmosphere in the big gambling room had changed. It was now much quieter. TT: El ambiente había cambiado por completo en la gran sala de juego que ahora se encontraba más tranquila. (1977: 378) [The atmosphere had changed completely in the big gambling room, which now was quieter.]



Forays into Romance

147

The intensifier ‘much’ in the English text has its value expressed in the intensifier ‘por completo’ in Spanish. This is compensation, apparently, because the intensification is qualifying the degree of quietness in English and the degree of change in Spanish. But why was the transformation necessary? Surely a literal version would have worked? Vázquez-Ayora does indeed give a literal version, which he dismisses as ‘too weak’, and his aesthetic judgement is probably correct, for someone. However, the English version itself is very likely to be ‘too weak’ – personally I would prefer to say that ‘the atmosphere had changed completely’, if only the English start text had been written that way, even though I can understand it well enough as it is. We are left, I suspect, with a category where Vázquez-Ayora’s stylistic preferences risk requiring unrealistically complex linguistic efforts from translators, with little appreciable gain in return. In sum, the very abundance of examples, which are supposed to make the categories clear, finishes up obscuring the few areas of relative clarity inherited from Vinay and Darbelnet. You might object, of course, that I have cherry-picked a few bad examples from about 100 pages of illustration. But no, not really: almost any of Vázquez-Ayora’s examples could lead into quandaries like the few mentioned above. He was fundamentally unable to deal with the extreme variation that characterizes relations between languages. Each category opens a can of worms (some of which escape into the neighbouring category); each concept is like an attempt to repair a leaking roof from the inside (the water will always drip through elsewhere); or better, to speak with Bally, the life of languages exceeds this attempt to put it into rational boxes. Or was Vázquez-Ayora simply better at revising translations than he was at constructing theory?

What did Vázquez-Ayora think he was doing? The worrying thing about Vázquez-Ayora is that, as Gil-Bajardí has observed (2003: 65–6), ‘it is difficult to distinguish the strictly generative-transformational features from those that are merely contrastive-structural’. In other words, all the new linguistics that he invokes make little actual difference to the categories inherited from tradition. So how does Vázquez-Ayora position himself in linguistics? On the one hand, he explicitly claims he is integrating what he calls the ‘Franco-Canadian School’, naming Bally, Panneton, Malblanc, Bart, Clas, Rey and others (1977: 5). On the other, he also seeks to draw on the ‘American School’, which is where he

148

Translation Solutions for Many Languages

finds Chomsky and his partial use in Nida. The real problem is that these two sides are not integrated: the ‘American School’ supplies the linguistics and the ‘Franco-Canadians’ give the translation categories. When discussing Transposition, Vázquez-Ayora claims there is some useful input from transformational grammar: ‘transformations do not change the content since they are different ways of saying the same thing’ (1977: 268). That, as I have suggested, is a questionable article of faith, a time-bound belief in stable semantic content that is not only essentialist but here is drawn out of context: Chomsky’s groundbreaking work in generative-transformational grammar concerned syntax, not semantics, and for that matter Transposition should be purely an affair of syntax. Historical accuracy notwithstanding, it is interesting to note the way Vázquez-Ayora firmly believes that the (then) recent linguistics would bring out substantial advances, to the extent that he has no real interest in older linguistics. In this case, he makes no reference at all to Bally’s method of studying the ‘equivalents’ internal to a language, or the way Bally placed alternative expressions one on top of the other such that, visually if not entirely conceptually, it is rather like a series of transformations. Bally’s schematization was at the root of what any linguist does with language; the prime difference is that the early transformationalists thought that the linguists’ operations were actually within some kind of linguistic mind. In a conference paper published in English two years after his textbook, Vázquez-Ayora (1979) does indeed claim that the metalinguistic work of stylistics enacts transformations that, ideally, reflect the generative transformations by which translations are produced: By means of a reversal of the process of generation of discourse (back-transformation) … the pioneers of generative stylistics … attempted to show that some readers and critics’ inferences about an author’s style had some underlying markers in the history of transformations. (1979: 204)

The peculiar thing here is that all the theorists referred to in this passage were working in literary stylistics, and the reference is very clearly to the style of particular authors, not to the ‘genius’ of particular languages. Vázquez-Ayora goes on to claim that literary studies is the correct place to analyse how a text works, or how translation criticism should be developed, and here he cites Kristeva, Greimas, Barthes, Eco and Culler as indicating the importance of generative approaches within structuralism. None of those names are cited in the textbook on translation, and one frankly wonders how much of the French authors Vázquez-Ayora had actually read. The level of ‘literal translation’ is here



Forays into Romance

149

revamped as ‘zero-degree translation’, as an overt allusion to Barthes’ theory of the ‘zero-degree of writing’ (1953); ‘oblique’ becomes ‘decentred’, as noted above; the level of macrotextual adaptation and compensation is linked to Van Dijk’s text linguistics of the day;1 and Kristeva’s Semiotikè (1969) is cited but there is no inkling of the extent to which her generative reconstruction is actually psychoanalytical, referring to Freud and Lacan. In all, one gains an intriguing glimpse of the integrated poststructuralist translation theory that might have been – especially if someone had followed up the idea of transformations acting to repress or censor the unspoken. As Ortega y Gasset had seen, ‘each social group silences some things so as to be able to say others’ (1937/83: 444, my translation). Further, underlying Vázquez-Ayora’s now clearly literary approach, all transformations are seen as being based on interpretations: ‘in any translation function, interpretation is an integral operation … Translation is an act of interpretation’ (1979: 205). This, of course, is something that had been integral to hermeneutic approaches since the nineteenth century, was expressed clearly in Fedorov (1927/74) and was waiting to be miraculously revealed by Venuti (2013), yet it remains an insight sadly lacking from almost all the other approaches to translation solutions. Vázquez-Ayora may have been an inveterate name-dropper, a theorist who never really achieved the integration he claimed to have attempted, but he did have the great virtue of envisaging some of the more interesting connections that could have been made within the space of translation solutions.

Postcard from a different world: Italian glory revived? Vázquez-Ayora was pointing in interesting directions, even if he failed to get there. In the same years, though, it was entirely possible to take comparative stylistics down a far more restrictive path. In 1979 Pierre Scavée and Pietro Intravaia published their Traité de stylistique comparée. Analyse comparative de l’italien et du français, which purports to analyse ‘collective style’ as ‘the preferential choices of a whole collectivity that privileges just some of all the possible affective expressions, in accordance with a particular sensibility’ (1979: 14). While ungenerously declaring Humboldt’s theory of language systems as worldviews to be ‘outdated, impressionistic and metaphysical’, they deploy a bizarre combination of Leo Spitzer’s literary comparativism and basic structuralism to characterize the Italian language in terms that themselves seem remarkably impressionistic, metaphysical and wilfully dated:

150

Translation Solutions for Many Languages

We speak of a ‘Saint Francis complex’ to designate the clear traces of a sentimental humanism that marks Italian sensibility, of a ‘Benedetto Croce complex’ to name a tendency to conceptualism and abstraction, and of a ‘Pietro Bembo complex’ for the atavistic penchant for the sonority of periods, rhythms, redundancies and bombast, etc. (1979: 22)

The task of translators is then to maintain these acquired ‘collective styles’, transforming expressions where necessary. Each of the ‘complexes’ is illustrated by translation pairs like the following: È difficile tirare avanti in queste condizioni. Ce n’est pas facile de vivre dans ces conditions. L’europeismo è andato indietro.

L’idée européenne a régressé. L’argomento è di ardua trattazione.

Le sujet est difficile à traiter. When the French verb ‘to live’ becomes the Italian ‘to pull ahead’ (in the first of these examples), the shift is supposed to reflect the humanist optimism of Saint Francis. Nice to know! When the French ‘European idea’ is expressed as ‘Europeanism’ (in the second example), it is the abstractionism of Croce that is shining through. And when an argument is ‘difficult to deal with’ in French but ‘of arduous treatment’ in Italian, this is supposed to be the legacy of Bembi. Such a characterization of a language, not to mention the strange assumption that French is a neutral point of reference, is a throwback to the German grammarians and literary historians of the nineteenth century, in fact to positions prior to Bally. But there it was, in 1979, happily marginalized by an emerging discipline called Translation Studies.

8

Meanwhile Back in German

Chapter summary: German-language approaches to translation solutions traditionally give greater weight to the more foreignizing options, as one might expect from the history of German as a relatively stateless language seeking to incorporate resources from the outside. The early systematization of previous categories was nevertheless challenged in the 1980s by Skopos theory, which posited that solutions depended first on the purpose of the translation situation, rather than on particular texts or languages. In the 1990s this approach was integrated into the traditional solution types in the work of Michael Schreiber, who distinguishes between high-level translation ‘methods’ and lower-level translation ‘procedures’, the latter usually operative at sentence level. Schreiber also includes categories for editing and adaptation (Bearbeitung), which allows the work of translators to include more than translation.

Our knowledge was once limited not only by the languages we didn’t know, but also by the books we couldn’t find. In 1979, having decided I wanted to write a Master’s thesis on translation theory (and thinking it would be no more than a Master’s thesis!), I found myself in Europe looking for books. Prior to that, my knowledge of translation was limited to Nida, Steiner, Lefevere, Mounin, Meschonnic, Levý in German translation and cited fragments from Vinay and Darbelnet, with the last-mentioned being, to my mind, by far the least interesting. There was not a lot more to be found, at least not in English and French. So, travelling in 1979, I bought some reasonably fresh publications: Ladmiral (1979) and Koller (1979/92), with the latter pointing me to Reiss (1971/2000). In Koller, I found news of a whole body of theory that I had never heard of: the Leipzig school, especially Otto Kade (but strangely very little about Fedorov or any other Russian theorist). Since I wanted to work on something like a political economy of translation, Kade sounded like he would offer a helpful socio-historical approach. So there I was, in East Berlin, the very day the Soviets

152

Translation Solutions for Many Languages

invaded Afghanistan (24 December 1979), going through bookshops looking for Kade, while snow fell and troops were moving all over the place. I didn’t find him – in fact I was not to lay eyes on Kade (1968) until January 2014. But that is another story. The message of the anecdote concerns not so much what I missed, but what was available in German at the time, and the role of German as a knowledge filter. Without German, it was not easy to know what had been happening in Russian or Czech (yes, there had been some reviews of Fedorov, but not the actual texts); without German, the projects of the Leipzig School were unavailable; without German, it was not really possible to follow the debates over Skopos theory in the 1980s. And even with German, as I have just tried to indicate, it was not always easy to find the texts. Here I will try to indicate how those diverse strands of German-language theory came together in one particular typology of translation solutions, published by Michael Schreiber in the 1990s.

Why German translation theory has been about languages I seem to have left German theory at a point where Malblanc was citing von Humboldt (1820), attempting to make comparative stylistics speak in terms of the ‘worldviews’ of different languages. That was in spite of the pains Bally had taken to distance his stylistics from the ‘impressionistic’ aspects of the German tradition. There was, however, rather more at work in German translation theory, and it had been there for a considerable time. Wilhelm von Humboldt (1820/1945) had indeed talked about languages as expressing worldviews. Yet he was also a translator from Greek, and as a translator he saw his mission as being not so much to maintain separate worldviews as it was to enable one language to help another language develop. For him, the most important aim of translation was ‘to expand the significance and expressive capacity of one’s own language’ (1816/1963: 81), an idea that finds its roots in the mediaeval hierarchy of languages (Pym 2000). In von Humboldt’s case, structures from Classical Greek were supposed to ennoble the rough-hewn qualities of German. This required a relatively literalist view of translation, although not blindly so. The general view of translation as actively developing the target language can be traced from Schleiermacher (1813/1963) and Goethe (1819/1963) through to a strong nineteenth-century tradition. It became what Berman (1984/92) saw as the central concept of Bildung, the cultural formation,



Meanwhile Back in German

153

education and development that was to ensue from a national concern with translation: to translate was to incorporate the other into the development of the self. The importance of language systems in this relationship was overdetermined by the fact that German as a national language did not correspond to a unified state until 1871 – development of the language acted as a surrogate for political unity. Germanic reflections on translation through to the twentieth century thus tend to talk more about relations between languages than about people or texts. Indeed, this tradition can be seen at work in Walter Benjamin’s 1923 essay on ‘The Task of the Translator’, which is mostly about how languages relate to each other, no matter how many postmodern interpretations it is subjected to. Partly thanks to this tradition, towards the end of the nineteenth century it became commonplace for European linguists to assert that (major national) languages had distinct ‘personalities’ or ‘geniuses’, and we have seen this as a major starting position for Bally’s notion of comparative stylistics. For Bally and the Francophone tradition, translation was likely to be a cause of interference in those established systems, and was thus to be mistrusted. In the German tradition, however, those same interferences could have a positive value, as a cause of development and improvement. So where Bally sought to correct literalism and never overtly integrated translation into foreign-language teaching, the German tradition was far more accepting of literalism and used translation extensively in the learning of languages. Not by chance, the teaching method known as ‘grammar translation’ was developed in Prussia (Howatt 1985: 131; cf. Pym et al. 2013). The two traditions had quite different ways of evaluating their national languages, hence different views of the role of translation and correspondingly different approaches to the teaching of translation. Yet they both needed typologies of translation solutions.

Attempts to systematize solution types The Germanic passion for system may be a stereotype, but in the case of translation solutions there appears to be some substance behind the rumour. As we have seen, Kade (1968) did not refer to Vinay and Darbelnet but did develop a simple logical approach to solution types, based on the categories of one-to-one, one-to-several and so on. That mode of thought, developed in Leipzig, was to have an impact on the translation theorists of what was then West Germany, but only after they negotiated the schemata that were coming from French and English.

154

Translation Solutions for Many Languages

Wolfram Wilss’s Übersetzungswissenschaft of 1977 is a compendium of Translation Studies to that date. It gives respectable space to typologies of translation solutions, including Vinay and Darbelnet (respected because, as Wilss notes, their typology had at that time been in use for some 20 years). There is nevertheless something slightly comic in the way Wilss immediately imposes a systemic hierarchy on the ‘list’ model inherited from comparative stylistics (1977: 121; my translation): Translation procedures

Literal translation (Substitution)

Non-literal translation

Loan, calque Word-for-word Literal translation  Transposition Modulation

Well, he tried. But the hierarchy is not entirely happy: the term ‘literal translation’ appears on two different levels; ‘word-for-word’ has been introduced from somewhere else (it is not in Vinay and Darbelnet); ‘équivalence’ and ‘adaptation’ have been forgotten about (although they could be fitted in); and the term ‘Substitution’, imported from Kade (see Wilss 1977: 125), is positioned in precisely the opposite place from where the Russian theorists had put it. As Wilss immediately recognizes, the hierarchy is of little benefit unless it can help identify what translators do in practice, and everyone more or less agreed, even then, that translating translators tend to mix and match their solutions. Wilss further claims that any resemblance to actual practice has been finessed by Vinay and Darbelnet’s initial distinction between grammatically obligatory solutions (toward the left of the diagram) and the optional ones (to the right) (1977: 122–3), although I am personally not convinced that things were ever that neat. He then initiates a rather interesting semantic slide, almost inadvertently: the minor adaptations required by ‘servitude to grammar’ (supposedly on the left) are described in terms of ‘one-to-none correspondence’ (when a value is expressed elsewhere or in a quite different way) and ‘none-to-one correspondence’ (when something appears to have been added). Those terms were from Kade; that mode of thinking about solutions was to become important elsewhere, in a different kind of theoretical language. Like Wilss’s overview, Koller’s Einführung in die Übersetzungswissenschaft (1979) was a substantial summary of what had been done in Translation Studies to that date, presented as a textbook for the several thousand students enrolled in the German university-level translation schools at the time.



Meanwhile Back in German

155

Although Koller does include Vinay and Darbelnet in his list of references, his general comment on their solution types is considerably more disparaging than was Wilss’s: The ‘procédés techniques de la traduction’ described in the Stylistique comparée developed by Vinay and Darbelnet and Malblanc seem to me to be so heterogeneous and bewilderingly unsystematic [verwirrend unsystematisch] that I decline to reproduce them here. (1979: 235; my translation)1

One suspects Koller found the categories ‘bewilderingly unsystematic’ partly because Wilss had tried to make them bewilderingly systematic, and had visibly failed. Koller also seems to respond to Wilss when he presents his own categories for solving translation problems, expressed in terms of the following ‘correspondence types’ (Entsprechungstypen): One-to-one: The number five translates the number fünf. One-to-several: River is translated as fleuve (if it flows into the sea) or rivière (if not). Several-to-one: The reverse of the above. One-to-none: When there is no corresponding item, transcription, loan translation, description or adaptation can be used. One-to-part: When none of the semantic boundaries match (as in basic colour terms in different languages), various simplifications or combinations are possible (for example, yellowish, blue-green, blood red). These terms clearly draw on those elaborated in Kade (1968): East Germany entered the West. And if you look hard enough, you might recognize something like the division introduced by Retsker (1950) between equivalence and ‘analogue’ correspondence. But no one was particularly looking that far to the east. And in any case, the careful application of systemic linguistic categories was soon to be blown apart.

The apparent Skopos revolution In 1984 two books appeared in German: Grundlegung einer allgemeinen Translationstheorie (Foundation for a General Theory of Translation) by Katharina Reiss and Hans Vermeer, and Translatorisches Handeln (Translatorial Action) by Justa Holz-Mänttäri. Together, these works presented translation as a goal-driven activity that was specific to each situation. They thereby radically

156

Translation Solutions for Many Languages

questioned all previous theories that had calmly set up mapping operations between pieces of language. If there was a foundational principle, it might be this: ‘The dominant factor of each translation is its purpose [Zweck]’ (Reiss and Vermeer 1984: 96). A more elaborate statement was offered by Vermeer a few years later, when the Greek term Skopos had been adopted in order to say ‘purpose’: Each text is produced for a given purpose and should serve this purpose. The Skopos rule thus reads as follows: translate/interpret/speak/write in a way that enables your text/translation to function in the situation in which it is used and with the people who want to use it and precisely in the way they want it to function. (Vermeer 1989a: 20; translation from Nord 1997: 29)

These are now quite obvious points, I think, and we have seen that they were actually made by Sobolev right back in 1950 (although no one in West Germany seems to have been aware of the fact). Skopos erupted thanks to astute academic marketing and the unsubtle art of making old ideas sound new; it generated fierce debate in Germanic circles, probably because it concerned the status of Translation Studies as an academic discipline, attempting to carve out an institutional space separate from the study of Modern Languages (and thus separate from both linguistics and literary studies). That is, the reason could have been good old academic politics. Was there any intellectual content to the debate? Yes, there was some. Skopos theory, if read in terms of the above pronouncements, was not telling anyone to translate in a different way: if extreme literalism is considered appropriate to the purpose, then that is the way to go; if free adaptation is appropriate, then it should be done; and if the start text needs to be rewritten rather than translated, then the translator should offer to do that. The key difference, as might be clear here, was not in what translators should do, but in the range of textual activities that translators could legitimately engage in: translators, as writers of new texts, could be asked to produce more than translations. In this way, the adepts of Skopos theory opened a space for decisions that were effectively much wider than those envisaged by the typologies of translation solutions. Instead of saying what was in the translator’s toolbox, this theoretical discourse tried to locate the factors determining why the translator should turn to one type of tool or another. One result of this was the complete absence in this camp, as far as I can tell, of anything like a typology of translation solutions – the question was somehow below their sphere of interest. Skopos theory had nothing to say in this particular history. A second result, deriving from the portentous certitude of



Meanwhile Back in German

157

theoretical pronouncements like the ones I have cited above, was that there was little left to discover. Skopos theory did not spawn much in the way of research projects; it offered few intellectual adventures to be completed in the future; it generally failed to produce a second generation. For all that, some trace of the Skopos moment can be found in the more recent typologies of translation solutions.

Schreiber integrates Skopos Michael Schreiber’s Übersetzung und Bearbeitung (Translation and Text Improvement, perhaps) (1993) effectively shows the import of Skopos theory in that it categorizes not just the traditional solution types, but also varying degrees of correction and adaptation – all of the latter coming under the head of Bearbeitung, which might indeed be ‘text improvement’ but could also cover editing, revision, post-editing and probably more. This additional space for wider translational action is a clear inheritance of Skopos theory, at the same time as the use of the term Bearbeitung cleverly defuses the debates over the limits of translation (which can then remain more or less as it was, with minimal effect on the traditional categories themselves). A little later, Schreiber (1997) presented a list of translation ‘procedures’ that were clearly intended for pedagogical purposes, and those were then reworked in the influential Handbuch Translation, published in 1998, where Schreiber is responsible for the entry on ‘translation types and translation procedures’. Here I concentrate on the types listed in the 1998 Handbuch. The extreme interest of Schreiber’s work derives not just from the Bearbeitung part but also from the way he draws on many previous German-language typologies of translation solutions, thus implicitly developing a model that should work for many languages (almost always in combination with German). Indeed, he does the kind of work I am attempting to do in this book, and his results deserve to be better known in English. Schreiber makes an initial distinction between ‘translation methods’ (Methoden) and ‘translation procedures’ (Verfahren), where the former depend on the whole text and vary in accordance with the text type and the intended function of the translation, while the latter concern smaller text units and vary in accordance with the ‘translation methods’ and the language/culture pair concerned (1998: 151). Schreiber notes that the classifications developed within the stylistique comparée tradition tend to assume that there is just one

158

Translation Solutions for Many Languages

translation ‘method’ within which the different ‘procedures’ are used (this is explicit, for example, in both Fedorov and Vinay and Darbelnet). Nevertheless, he says, some German scholars, apparently including Reiss, Nord and Schreiber himself, believe that this single ‘translation method’ does not correspond to the plurality of translation practice (as Cary had indeed argued when dissenting from Fedorov). Schreiber thus subordinates the ‘translation procedures’ to a typology of different ‘translation methods’.2 I am not sure to what extent Reiss and Nord would actually go along with the categorical part of this argument, but the reference to large-scale ‘translation methods’ certainly resonates all the way back to Schleiermacher (1813/1963) and his binary description of ‘foreignizing’ and ‘Germanizing’ as macro ‘methods of translation’. Schreiber nevertheless offers an innovative and genuinely useful typology of such text-level approaches, which actually looks rather more like general types or degrees of translation: 1. Text-restricted translation (Textübersetzung), which prioritizes invariants3 that are internal to the text, on the level of either form or content. 2. Context-sensitive translation (Umfeldübersetzung),4 which prioritizes invariants that are external to the text, notably the originally intended meaning (in the case of corrections to the text) or the original text function (in the case of culturally domesticating translations). 3. Interlingual adaptation (Interlinguale Bearbeitung), which does not prioritize invariance at all, instead placing value on variants or ‘intentional modifications’. To summarize: some translations work on what is in the text; others work on making the text work within a context; and still others more radically alter the text to suit a new context. So much for the ‘methods’.5 Schreiber then lists the solution types that might frequently be used under each of these categories. Since his work addresses many of the issues that have been found problematic in other typologies, and given that there seems to be no prior translation of his categories in English, I take some time here to put all the procedures into a table (working mainly from Schreiber 1998, with an eye on Gibová 2012: 35). Schreiber draws his examples from a range of German-language translation scholars who had worked on translation solutions over the years (Bausch 1968/81; Albrecht 1973; Koller 1979; Thome 1981; Wotjak 1985; Schmitt 1987; Zimmer 1990; Neubert 1991; Henschelmann 1993; among others), in effect bringing together a small tradition and indicating that these categories should work between German and a range of other European languages. Here I adapt some of the examples so that they might be accessible to readers with no German.



Meanwhile Back in German

159

Table 8.1  Schreiber’s model of translation procedures Text-restricted translation

Lexis

Grammar

Solution type

Explanation

Example

Lexical Borrowing

Taking-over of a lexical unit

En. ombudsman (Koller 1979/92: 233)

Lexical Substitution

Substitution of a SL lexical unit by a TL lexical unit

En. table, Fr. table

Lexical Restructuring (Strukturwechsel)

Change in word class

Fr. Construire un bâtiment … [To construct a building … ] En. The construction of a building …

Word-for-word Translation

With word-count, word class and word position retained

En. Where is he? Fr. Où est-il?

Permutation

Reordering of sentence elements

En. Where is he? Fr. Il est où? [He is where?]

Expansion/ Reduction

Increase/decrease in word-count

En. He will be able to … Fr. Il pourra … (Thome 1981: 311)

Change within a grammatical function (Intrakategorialer Wechsel)

The same idea is expressed through a change within the same part of speech, e.g. a generalized substantive takes the article in French but has no article in English

En. Progress leads to a crisis … Fr. Le progrès engendre une crise … (Henschelmann 1993: 59)

Transposition

Change in the part of speech

It. Il professore S. muove dall’idea … [Professor S. moves from the idea that … ] Ger. Professor S. geht davon aus [Professor S. goes out from that … ](Bausch 1968/81: 282)

Transformation

Change in syntactic construction

Es. La sangre derramada en la segunda guerra mundial [The blood shed in the Second World War] En. The blood that was shed in the Second World War (Wotjak 1985: 31)

Translation Solutions for Many Languages

160

Semantics

Helps

Solution type

Explanation

Example

Semantic Borrowing

Verbalization of the same content features, e.g. with turns of phrases or idioms

En. He saw red Fr. Il a vu rouge (Zimmer 1990: 40)

Modulation

Change in the point of view, through the verbalization of different content features

En. His failure to feel excitement about … De. Er war gar nicht scharf darauf … [He was not at all keen on … ](Neubert 1991: 37)

Explicitation/ Implicitation

Increase/decrease in the degree of explicitation

Fr. Avant que le professeur de gymnastique n’envoyât le coup de sifflet final … [Before the gymnastics teach gave the final blow of the whistle … ] En. Before the instructor’s whistle … (Albrecht 1973: 42)

Mutation

Change of the denotative content for the sake of another invariant, as when rhyme dominates in verse translation

En. The Price is Right De. Der Preis ist heiß [the price is hot]

Help Procedures

Translators’ notes, forewords, afterwords, glossaries

En. IRB De. Irisch-Republikanische Brigade (Schreiber 1993: 234)

Correction

Reconstruction of ‘what is meant’, by correcting an error in the start text

If the ST refers to a ‘327-meter freeway’ between two cities, the translator might want to correct it (Schmitt 1987: 2)

Adaptation

Adaptation to the target culture by using ‘situational equivalence’

French text: ‘From 1 October 1983, new bicycles must have two orange reflectors’ German text: ‘From 1 January 1986 all bicycles must have four orange reflectors’ (Schreiber 1993: 153)

Intentional changes in the amount of information offered

Summaries ‘Moralizing’ commentaries added to children’s stories (Schreiber 1993: 292)

Context-sensitive translation

Interlingual adaptation Additions/ Omissions

Source: Adapted from Schreiber (1998: 152–3).



Meanwhile Back in German

161

Schreiber’s classifications have several laudable points. First, since there are many languages in play, there is no suggestion that the translator is making adjustments for the ‘stylistic’ preferences of different languages – the priorities of situational purpose clearly make the analysis of languages a rather secondary concern. (This is all the more remarkable since it marks off this approach not only from Francophone stylistics but also from much of the weight of German-language translation theory, which had largely been about relations between different languages.) Second, the division between large-scale ‘methods’ and smaller-scale ‘procedures’ seems useful, not only in that it indicates that decisions are made on very different levels and with different priorities, but also because it suggests that the levels of description could be on a sliding scale: just a few large categories for some problems, then a range of smaller categories for others. I am not sure, though, that the subsequent distinctions between ‘lexis’, ‘grammar’, ‘semantics’ and ‘helps’ actually rise above the level of good housekeeping (partly inherited from Vinay and Darbelnet). Third, the category of Permutation fills a strange gap that appears to have gone unnoticed in Vinay and Darbelnet. Fourth, some serious attempts have been made to maintain clear and reasonably testable criteria, as in the distinction between ‘text’ and ‘context’ on the level of the methods, or between Transposition and Modulation among the procedures. Fifth, the extension of the field into Bearbeitung (a term that is not a synonym for ‘translation’) clears up many previous ambiguities with respect to terms such as ‘addition’, ‘omission’, ‘explicitation’ and ‘implication’ – here, at last, we are sure that Addition really does meet adding something new, and that Explicitation really does not. That said, there are aspects about which one might still carp. For example, the fact that the table apparently covers both obligatory and optional solutions leaves the classification without any firm cognitive base, at the same time as it removes one of the most important points to be made when teaching (the need to choose) – does ‘Change within a grammatical function’ really need to be mentioned as a solution type? In this respect, the example given for Transformation does not make things very clear (‘blood shed’ becomes ‘blood that was shed’, presumably to avoid sounding like the noun ‘bloodshed’?) and as such remains hard to distinguish from Explicitation. Or again, if I want to explain what the siglum ‘IRB’ means (given here as an example under Helps), I could either spell it out in the text or add the information in a footnote. In the

162

Translation Solutions for Many Languages

first case it might be Explicitation, in the second it is categorized as an added Help. But are these two alternatives really so different? Is the explicitation in a footnote really on a par with added forewords and postscripts? Beyond the nit-picking, the initial distinction between text and context, which underlies the distinction between the two ‘methods’ labeled Textübersetzung (‘text-restricted translation’) and Umfeldübersetzung (‘context-sensitive translation’), would not hold up in a court of hermeneutics or cognitive studies, since people are almost always construing textual sense in terms of contextual elements. Of course, you could say that, when deciding between alternative viable solutions, one kind of translation is swayed more by the text and another more by the context, but such a view of competing priorities is little in evidence here. One might seriously question whether there are indeed separate ‘methods’ operative on the whole of the text. Despite these doubts, Schreiber’s typology remains a highly significant contribution. Its inclusion of an ‘extra-translational’ dimension, for things that can and often must be done beyond a narrow concept of translation, responded to Skopos theory in its day and is now of even greater relevance, addressing some of the issues raised in the field of localization, where our technologies are separating straight translation practices from the higher levels of cultural adaptation (‘localization’ ideologically means ‘adapting to a locale’). Those issues still have to be explored. As we venture into Comanche territory, some of the work done in German can serve as a guide.

9

Disciplinary Corrections

Chapter summary: The development of Translation Studies as an international academic discipline in the 1990s meant that many of the typologies of translation solutions tended not to respond to particular languages or national language policies. Instead, they became part of a Western academic discourse. This can be seen, for example, in the way Peter Newmark’s categories informed the Thinking Translation series of textbooks for many language pairs, and also in the relative cross-cultural stability of the terms used in the French and Spanish manuals. Some of the multilingual typologies that appeared around the turn of the century then curiously returned to Vinay and Darbelnet, while others offered typologies that are unsystematic or simply too complex for pedagogical use. The result is a degree of variation that makes progress hard to identify.

At the beginning of our story, translation scholars were looking at the dominant linguistic theories, at various explicit or implicit language policies and in some cases at their immediate translation cultures. In such contexts, the theorization of translation solutions was embedded in some highly ideological political considerations, touching on issues of language policy and international relations. Increasingly, though, scholars found their points of departure in other scholars: Malblanc applied Vinay and Darbelnet to German–French; Zhang et al. reworked Loh, Vázquez-Ayora purported to adapt Vinay and Darbelnet to American transformational grammar, and so on. This embedding of scholarship within scholarship effectively enabled the lists of solutions to be successively adapted to new language pairs and historical situations, and in some cases to be explicitly ‘improved’, at least in the eyes of the newer theorists. After a few decades, particularly at the beginning of the 1990s, this process of internal academic development reached the point where the prime context of thought became not so much a particular language pair or the immediate translation culture, but the field of Translation Studies itself, as an academic discipline.

164

Translation Solutions for Many Languages

As this happened, ideas were separated from the ideologies and specific linguistic contexts in which they were developed – they were called upon to address new sets of problems. A corollary, however, is that the academic field itself thereby gained the illusion of some kind of ideological or political neutrality, as if the dirty histories had been washed away, leaving only pure ideas. That is an illusion to be challenged. Or have we really reached a state where problem-solving is a merely technical, academic affair? The process of academization was something that many of us fought for at the time, especially in the 1980s and 1990s, calling repeatedly for an independent Translation Studies, then justifying its shortcomings in terms of a discipline that was ‘still in its infancy’. The new discipline would be independent of Linguistics and Literary Studies – as envisioned when the Soviets were arguing about science vs. literature – but could call on conceptual tools from near and far, developing its status as an ‘interdiscipline’ – a term proposed by Gideon Toury (see Snell-Hornby 2006: 71) and showcased at a landmark congress in Vienna in 1992. As befits a ‘field’ in the sense defined by Bourdieu (for example, 1993), the discipline itself became the place where scholars fought for prestige, accumulated symbolic capital from having ‘their’ words disseminated and occasionally reached consensus. The downside, of course, is that the ongoing historical development of Translation Studies has thereby entangled nodes where translation scholars only cite other translation scholars, and where students have done undergraduate and postgraduate studies only in Translation Studies. The outcome is a certain parochialism, even a degree of in-breeding. Younger scholars cite the established works like Vinay and Darbelnet, with scarcely an inkling of where the ideas came from or how they might be related to other disciplines. True, there are occasional imports of biggish ideas from larger and more prestigious disciplines. Yet some of us sorely miss the days when each scholar brought in significant expertise from neighbouring fields, in addition to an awareness of the practicalities of translation. You might presume that the creation of a relatively independent Translation Studies would have enabled the solution types to be progressively refined, improved, made applicable to a wider range of languages and situations, within a discipline able to facilitate corrections. At the same time, ‘correctional facility’, I believe, is another name for a prison. Let us see how it was formed.



Disciplinary Corrections

165

Newmark and English pragmatism The English translation critic Peter Newmark was very much his own man, hardly a likely candidate for the foundation of a discipline. Yet he has to be fitted in somewhere; he developed the Vinay and Darbelnet tradition; and his various appeals to ‘authority’ were not unlike those of a schoolmaster, who did indeed see a certain school grow around him. Newmark presented a lapidary short list of ‘translation procedures’ in Approaches to Translation (1981: 30–2) and then a whole chapter on them in A Textbook of Translation (1988). As is clear in Table 9.1 (in part drawn from Gibová 2012: 33), the 1981 list is very heavily indebted to Vinay and Darbelnet but adds a few things here and there. Table 9.1  Translation solutions according to Newmark Solution type

Example or explanation

Transcription

détente; démarche

One-to-One Translation

la maison, the house

Through-Translation, Loan-Translation

People’s Chamber, Volkskammer

Lexical Synonymy

‘Translation by a close target-language equivalent.’ ‘A bath’ can be translated literally, even though different cultures have different sizes and styles of baths.

Componential Analysis

Comparison of a start-language word with a target-language word that has a similar meaning, by means of semantic components. Ein Greis, a very old (aged) man

Transposition

according to my friend … mein Freund meinte (my friend thinks/opines … )

Modulation

Lebensgefahr (life danger) ≈ danger de mort (danger of death)

Compensation

‘Loss of meaning or sound effect or metaphor in one part of a sentence is compensated for in another part.’

Cultural Equivalence

Baccalauréat ≈ A-level

Translation Label

‘An approximate equivalent, sometimes proposed as a collocation in inverted commas, which may later be accepted.’ promotion sociale ≈ social advancement

Definition

Descriptive noun-phrase or adjectival clause

Paraphrase

‘An amplification or free rendering of the meaning of a sentence’

Expansion

taste of, avoir le goût de (to have the taste of)

Contraction

science anatomique, anatomy

Recasting Sentences

Complex sentences as coordinate sentences: splitting of complex sentences into two or more sentences

Rearrangement, Improvements

Removing mistakes, misprints, idiolect or clumsy writing in defective texts

Translation Couplet

Transcription of a term followed by its translation: Gemeinde (German unit of local government)

Source: Adapted from Newmark (1981: 30–1).

166

Translation Solutions for Many Languages

The list has seventeen solutions, which can be seen as Vinay and Darbelnet’s big seven (in more or less the same order) plus a few of their ‘prosodic effects’ (Compensation, Expansion, Contraction), then some added procedures: Componential Analysis, which is really a model of how Lexical Synonymy or Modulation is produced rather than a solution type in itself; Recasting Sentences, which I suggest is better called Resegmentation when it means splitting and joining sentences, and is highly justified (as we saw when in Chinese territory); and then there are items that might be from the terrain that Schreiber would later describe as Bearbeitung: Improvements (it is not clear why Rearrangement should be next to them) and Translation Couplet (which is just one step away from a translator’s note). As it stands, Newmark’s list contains a good many things that one would like to teach translation students; it has serious pedagogical virtues. At the same time, though, it is not based on any data, it comes from nowhere and it has no clear purpose, simply being introduced by the laconic sentence, ‘The translation theorist is concerned with every type of translation procedure’ (1981: 30), somehow suggesting that these might be all the types possible. Really? This lack of an aim actually leads to mixed functions: first, the list recalls some terms that students might remember, but Newmark then gives in to the non-pedagogical temptation to add as many tricks of the trade as come to mind; second, it confuses the solutions themselves with the way they are produced (Componential Analysis, as mentioned, concerns process and has no real reason to be here); third, several categories actually track the way foreign terms enter a language, which could be something different again. This last-mentioned function would be the work done by Transcription, Translation Label, Translation Couplet and indeed Naturalization (the adapting of a term to target-language pronunciation), which was the main addition made in the 1988 version and still does not seem a happy label (‘naturalization’ can mean several hundred other things as well, as indeed can ‘translation label’). All the solution types involved in this third function are doing much the same thing and can be used in combination, but they are in different parts of the table, as if they were somehow essentially different. Somme toute, the table is doing too many different things at once. If students can follow it, they are doing very well. Smaller gripes: Calque has disappeared insofar as it was related to syntax; Compensation has no reason to be limited to the sentence; there is no category for changes that only concern word order (what the Russians and Germans called Permutation); Lexical Synonymy recognizes the different



Disciplinary Corrections

167

sizes and shapes of baths in different cultures (so the literal translation of ‘bath’ may only have some overlap with the baths in the start culture), so why should this synonymy not apply to the noun ‘house’, classified as One-to-One Translation? In fact, surely there is a bit of synonymy everywhere (except in the sense of technical, regulated ‘equivalence’, as formulated in the Russian tradition)? These, however, are mere details when compared with the political frame in which the solution types are presented. Newmark, unlike the French theorists and in full-frontal opposition to the thrust of Skopos theory, has nothing but praise for solutions that are close to the text, particularly those that draw on componential analysis, and he offers little but scorn for the rest. His general warning is that ‘you only deviate from literal translation when there are good semantic and pragmatic reasons for doing so’ (1988: xi), and there are many lesser warnings about the perils of other paths. For example, ‘[t]he use of synonymy … is the ruin of accurate translation and paraphrase is even worse’ (1981: 31). Indeed, paraphrase is ‘the translator’s last resort’ (ibid.). And as for ‘improvements’, they are ‘only justified if a) the [start] text is concerned mainly with facts or b) the writing is defective’ (1981: 32). Otherwise, great texts must be rendered as closely as possible. This highly asymmetric distribution of values underlies Newmark’s distinction between ‘semantic’ and ‘communicative’ translation, where the deviations from the text are only justified in the latter. If you look closely at his writings, Newmark’s ideology is systematically based on praise for what he calls the ‘authoritative text’: ‘I begin by reminding you that you have no right to improve an authoritative text’ (1988: 204). So who wrote these great texts? Newmark’s lists are curiously limited to great wartime heroes of liberal democracies: ‘Pericles, Jefferson, Lincoln, Churchill, de Gaulle’ (1981: 49), ‘Lincoln, Churchill, de Gaulle, Pericles’ (1981, 58), ‘Churchill, de Gaulle’ (1988, 162) and ‘de Gaulle’ (almost throughout). Wartime struggle marked out Newmark’s very clear areas of textual respect and these were the authoritarian politics that then informed his presentation of solution types. Newmark was one of the very few writers in English who actually followed what was happening in translation theory written in other European languages (mainly French and German). He tracked it, cited it and dissented from it on several serious counts, not only with respect to Skopos theory (which he saw as allowing the translator too many deviations from the authority of the text) but also Descriptive Translation Studies (which he thought similarly legitimated historical deviations). A man of strong opinions, he is much missed.

Translation Solutions for Many Languages

168

Perhaps because of his clear positions, Newmark had strangely little effect on later attempts to solve these problems. A few of his ideas did nevertheless filter through to a successful set of textbooks, which is where the academization might start.

Thinking Translation in many languages Sándor Hervey and Ian Higgins co-authored Thinking Translation: A Course in Translation Method: French to English (1992; the second edition appeared with a modified title in 2002), which includes a rather sketchy typology of translation solutions that is vaguely based on Newmark. This would probably not be of much importance if the authors and their editors at Routledge had not hit on the brilliant idea of using the same theoretical approach, with virtually the same solution types and with Hervey named as co-author,1 for coursebooks on translation from English into German (2003/6), Spanish (1995/2009), Italian (2000) and Arabic (2002), thus implicitly claiming that the one typology of solutions is suited to many language pairs – and selling quite a few books in the process, if the number of second editions is anything to go by. This could put paid to all the concerns about translation theory being language-specific, in theory.2 For the basic model, I take the first edition of the coursebook for French (1992). Hervey and Higgins offer the term ‘cultural transposition’ for ‘various degrees of departure from literal translation’ (1992: 28), which would cover the work that Vinay and Darbelnet’s categories purport to be doing. The visual schema used here (Figure 9.1) is adapted from Newmark (1981: 39). The terms put into it, however, seem to derive from no single tradition that we have looked at so far: the sliding scale from Exoticism to Transplantation clearly strives to avoid linguistics-based terms, stressing instead that the decisions all concern issues of culture. Then again, Calque does indeed come from Vinay and Darbelnet (and Bally), and Communicative Translation was Newmark’s term. So things appear to be a bit of a mix.

Source-culture bias Exoticism

Calque

Target-culture bias Cultural borrowing

Communicative translation

Cultural transplantation

Figure 9.1  ‘Degrees of cultural transposition’ (from Hervey and Higgins 1992: 33)



Disciplinary Corrections

169

This confusion multiplies as you consider the terms in detail. Exoticism covers anything that ‘imports from the ST into the TT with minimal adaptation’ (1992: 30), but then Calque is when an expression ‘is unidiomatic in the TL because it is modelled on the structure of a SL expression’ (1992: 33), which is surely the same as Exoticism? Further, ‘calque is a form of literal translation’ (1992: 33), at the same time as all these categories are supposed to be for ‘degrees of departure from literal translation’. Cultural Borrowing is then described as being different from Exoticism and Calque, because it does not involve adaptation of the start-language expression – but if it is the form with zero adaptation, why is it not at the extreme left of the diagram? And then Communicative Translation is this huge category for the use of corresponding ‘clichés, idioms, proverbs, etc.’, which means lumping a whole set of Vinay and Darbelnet solutions into a group that is more on the level of the translation method (remember that Newmark opposed ‘communicative translation’ to ‘semantic translation’ as two translation ‘methods’). The students at St Andrews, where this typology was developed, must have worked hard to make sense of the divisions. The basic instability of the categories is revealed in some of the later additions to the book series. For example, Exoticism and Calque are lumped together as one solution type in the editions for Italian and Arabic. And then Deletion is inserted in penultimate position in a version for Spanish (second edition 2009). None of these changes can be explained in terms of language specificity, as far as I can tell. In all these editions, the authors’ discussion of the above visual diagram is followed by a section on Compensation, which becomes a whole chapter of its own after the 1992 version. ‘Compensation’ here does not have the sense that it had in any of the previous usages (from Vinay and Darbelnet onwards); it applies to the following situation: where any conventional translation (whether literal or otherwise) would entail an unacceptable translation loss, this loss is reduced by the freely chosen introduction of a less unacceptable one, such that important ST effects are rendered approximately in the TT by means other than those used in the ST. In other words, one type of translation loss is mitigated by the deliberate introduction of another. (1992/2002: 43)

Again, this makes little sense in terms of a well-constructed theory, not just because it somehow excludes any possibility of gain (to offset a loss) but more simply because the logic of the trade-off could be applied to any translation decision beyond those that refer to pre-established equivalents: all the degrees

170

Translation Solutions for Many Languages

of ‘cultural transposition’ should also be instances of Compensation, surely? The awareness of competing risks is nevertheless instructive, as is the term ‘compromise’, used here as a synonym for ‘compensation’. The example given, from Molière’s Tartuffe, is also intriguing (1992/2002: 37–8): ST: N’oubliez pas: il n’est pire eau que l’eau qui dort. TT1: Don’t forget: the worst water is water that sleeps. TT2: It’s the quiet people you’ve got to watch. TT3: Just remember: still waters run deep. TT4: Remember the saying: ‘Water asleep is water too deep.’ Hervey and Higgins know that the literal translation TT1 does not mean much in English and that a non-literal translation like TT2 gives the right idea but loses the fact that a French proverb is being used. The dictionary proposal TT3 gives a standard alignment of proverbs, but in this case the alignment attributes positive values to quiet people, whereas the French is doing the opposite. The happiest solution (TT4) is to add a reference to ‘the saying’ (this would be Explicitation in other systems) and to invent a new proverb, quite successfully in this case. Compensation here refers to both those things: the explicitation (a proverb is coming) and the invention of a saying that did not previously exist. The theory dithers, but this is good translating. Some of the language versions in the series list types of Compensation, others don’t. The second Spanish edition (1995/2009: xi) notes that those categories have been removed because they ‘proved overly fussy and confusing for practical use’. This is a telling statement, I think. The whole approach in these books seems encumbered by the very attempt at Cartesian categories; it works best when balancing numerous considerations in elegant prose, hedging a way forward pragmatically and ultimately teaching by example. There is nothing particularly wrong with that rather English way of getting things done (even when deployed by the Hungarian Hervey when teaching in Scotland). The real shortcoming, though, is when the schemata in this series are then picked up and applied as some kind of sure guide. This happens in one of the few typologies for Arabic–English that I have been able to locate: As-Safi (2011) directly adopts types of Compensation straight from Hervey and Higgins (1992), using the fussy and confusing terms as an apparently authoritative theory. In the meantime, other parts of the world were heading in different directions.



Disciplinary Corrections

171

Repetition as consecration Ricardo Muñoz Martín (1998) has compiled a table of the way the solution types appear in a series of French and Spanish translation manuals from 1989 to 1997 (Table 9.2). Here I have translated the terms into English (the only halfproblematic one was ‘Ampliación’, which I have rendered as ‘Amplification’). Table 9.2  Translation solution types in several French and Spanish translation manuals (square brackets indicate secondary solution types) 1989

1990

1990

1991

1992

1997

Chuquet– Paillard

Hardin–Picot

Elena García

Pascua– Peñate

Rajaud– Brunetti

López– Wilkinson

[Loan]

Loan

Loan

Loan

[Calque]

Calque

Calque

Calque

[Literal translation]

Literal translation

Literal translation

Literal translation

Literal translation

Transposition

Transposition

Transposition

Transposition

Transposition

Transposition

Modulation

Modulation

Modulation

Modulation

Equivalence

Equivalence

Equivalence

Adaptation

Adaptation

[Amplification]

Amplification

[Explicitation]

Explicitation

Amplification

Modulation Equivalence Adaptation

[Omission] [Compensation] Condensation Recomposition

Compensation

Compensation

Condensation Recomposition

Source: Adapted from Muñoz Martín (1998: 4).

Without going into the manuals, the immediate message of the table is that the space of thought remained relatively homogeneous over the decade and across the two Romance languages. Newmark, who had published at the beginning of the 1980s, had no visible effect on these authors, and there is no sign of any disruption from Skopos, both of which points should indicate the surprising extent to which pedagogical translation practices were still language-specific. Some authors do not really want to enter the realm of terminology (as was the case in Vázquez-Ayora); others do away with the confusions of ‘equivalence’; and there is considerable incertitude at the bottom of the table, in the swirling mists of prosodic effects. Those, however, are minor differences, enacted for reasons about which we are told very little. Many authors have simply used

172

Translation Solutions for Many Languages

a cut-and-paste from Vinay and Darbelnet (this might be the case in Elena García 1990; Pascua and Peñate 1991), assuming the status and reliability of the French linguists, and recycling the categories, unthinkingly, in accordance with a pedagogy of mindless authority. Many more manuals could be added in this vein. But what would be the point?

A multilingual terminology The collective work Terminologie de la traduction, published under the direction of Jean Delisle in 1999 – and thus at the end of the time frame presented in Table 9.2 – was a valiant and well-intentioned attempt to unify the terminology of Translation Studies. Although restricted to European languages (French, English, Spanish and German), it received input from ‘[n]early twenty translation teachers and terminologists from universities in eight countries (Canada, France, Germany, Spain, Switzerland, United Kingdom, United States and Venezuela’ (publisher’s blurb). The work lists the following ‘translation procedures’ (even though Delisle elsewhere stated they were not really procedures), taken here from one of the tables at the end of the book: Adaptation, Amplification, Borrowing, Calque, Compensation, Concentration, Concision, Denominalization, Dilution, Economy, Equivalence, Explicitation, Implicitation, Interchange, Modulation, Paraphrase, Recategorization. (1999: 204)

That makes seventeen terms, which is more than two handfuls, and the alphabetical order makes it hard to see any firm hierarchies by which the number could be reduced. Since the project seems to have been initiated in Canada, it is not surprising to find Vinay and Darbelnet’s terms surviving more or less intact. In fact, the only one to have disappeared is Transposition, reborn here as Recategorization. Here is a checklist: Loan (emprunt) is now Borrowing, as in the English version of Vinay and Darbelnet (1958/95: 31, 340), although it is hard to fathom why this should be the only -ing form in the list. Calque is unchanged. Literal Translation becomes Word-for-Word Translation, with examples that are faulty as everyday translations. The notion of literalism being a starting point has been lost.



Disciplinary Corrections

173

Transposition becomes Recategorization: ‘Since the term “transposition” has been applied to a number of different translation procedures, the term recategorization provides a meaningful alternative’ (1999: 171). Possibly. But surely here we have lost the reference to word classes? Modulation is where the translator introduces a ‘change in the point of view or a clarification’ (1999: 161). The ‘point of view’ part is the same as in Vinay and Darbelnet; the ‘clarification’ seems not to be. There is no example to clarify what clarification might entail. Équivalence: As in the English translation of Vinay and Darbelnet, here we find the third sense of ‘equivalence’ explained as ‘rendering a set phrase from the source language with a set phrase from the target language that expresses the same idea, although in a different way’ (1999: 138). This is a very useful explanation, but the fact that ‘equivalence’ has three very different meanings is decidedly less helpful. Adaptation is as in Vinay and Darbelnet, with a nice example: the pair ‘Dupont et Dupond’ in the French Tintin become ‘Thomson and Thompson’ in English, ‘Hernández y Fernández’ in Spanish and ‘Shultze und Schultze’ in German (1999: 114). The other ‘procedures’ in the glossary are all as in Vinay and Darbelnet (as translated into English), with the apparent additions of Denominalization, which is surely a special case of Transposition (sorry, ‘Recategorization’), and Paraphrase, defined as ‘replacing a word from the source text with a group of words or a phrasal expression that has the equivalent meaning in the target text’ (1999: 167), which looks like a loose version of Correspondence (sorry, ‘équivalence’). All in all, the only ostensible improvement to Vinay and Darbelnet here is the term Recategorization, which seems not to have much going for it – categories are everywhere.

Molina and Hurtado-Albir In a remarkably influential article published in 2002, Lucía Molina and Amparo Hurtado-Albir give a critical review of the main prior classifications and attempt to impose some order (as indeed would be the desire of most of us). They score many points by arguing that solution types should be (1) recognized as being based on products, rather than cognitive processes, (2) ‘options open to the translator’ rather than obligations imposed by grammar and (3) independent of whether a solution

174

Translation Solutions for Many Languages

is valid or not, since this always depends on the text and the situation in question (2002: 511). These are three great principles, although the first one seems not to be helped by calling these things ‘techniques’, a term that is then given a processbased definition: ‘the way micro-units of the text are translated’ (2002: 508).3 The strange thing is that, having announced these killer ideas that should have upset the whole applecart, Molina and Hurtado-Albir then resolve to ‘maintain the most commonly used terms’ and to ‘formulate new techniques’, presenting the traditional decontextualized phrases and proffering a never-to-be-remembered list of nineteen categories, arranged alphabetically so as to defeat any attempt at organizational hierarchy. After a brilliant beachhead, they basically capitulate. Without reproducing the whole list, here are some of the problems I have with the categories and their descriptions: Amplification: ST: ‫شهر رمضان‬ [month Ramadān] TT: Ramadan, the Muslim month of fasting Description: ST: Panettone (Italian) TT: The traditional Italian cake eaten on New Year’s Eve So what is the essential difference between these two examples? Is the presence or absence of the not-so-foreign name really enough to distinguish between two major categories?4 Established Equivalent: ST: They are like two peas [in a pod]. TT: Se parecen como dos gotas de agua. [They are as like as two drops of water.] Linguistic Amplification: ST: No way. TT: De ninguna de las maneras. [By none of the ways] Again, what is the essential difference? Do we really want to believe that ‘peas’ and ‘drops of water’ are metaphors, while ‘ways’ are not? (Think about it.) Transposition: ST: He will soon be back.



Disciplinary Corrections

175

TT: No tardará en venir. [He will not delay in coming.] Modulation: ST: ‫أبا ستصير‬ [Will become+you+MALE a father] TT: You are going to have a child. Sorry, but both cases look like a change in perspective, don’t they? Generalization: STs: Guichet, fenêtre, devanture TT: Window Particularization: ST: Window TTs: Guichet, fenêtre, devanture The point here is that French has three possible translations for the English ‘window’. But hang on, didn’t Molina and Hurtado-Albir say they were not going to consider the obligations of comparative linguistics? In short, a little further order could be added to those who sought to correct all prior attempts.

Chesterman’s strategies as memes Andrew Chesterman’s book Memes of Translation (1997) deserves to be recognized as an intelligent, practical, full-blown theory of translation, hiding under a misleadingly modest title. It has a chapter on ‘translation strategies’, honestly glossed as ‘conceptual tools of the trade’ (1997: 87). The sense of the term ‘strategy’ here is far from clear, but I note that ‘communication strategies’ are described as ‘ways of solving communication problems’ (ibid.), so it is hard to see why all Chesterman’s strategies could not be called ‘solution types’. They are assumed to be behavioural (not neural), linguistic (used when manipulating texts), goal-oriented (they are solutions), problem-based (they start from trying to solve a local textual problem, not on the level of the whole text), potentially conscious (they are not routine, background activities), intersubjective (they can spread through a community of translators, which is why they are ‘memes’) and they all start from the most basic and general of translation solutions:

176

Translation Solutions for Many Languages

‘change something’ (1997: 92). Chesterman then draws on prior theories of translation procedures and shifts (many of the ones I have mentioned here) to draw up a list of things that seem to fit these criteria. He claims that the list is not language-pair-specific, although he draws his examples from the German and English texts found in an Austrian Airways in-flight magazine, which could be as good a place as any to locate the universal. Chesterman distinguishes between three classes of solution types: syntactic, semantic and pragmatic. Since there are ten items under each head (the neatness suggests good housekeeping rather than a truly decimal world), the total number of solution types is thirty – rather more than the two handfuls that I am looking for, and perhaps more than a student can remember for an exam, but nevertheless a useful number for opening horizons in a series of practical classes. They are worth reproducing (I add a few notes to the items that may not be immediately clear): Syntactic: 1. Literal Translation 2. Loan, Calque 3. Transposition 4. Unit Shift (resegmentation of phrases, clauses and sentences) 5. Phrase Structure Change 6. Clause Structure Change 7. Sentence Structure Change 8. Cohesion Change (different cohesion devices for the same effect) 9. Level Shift (a phrase becomes a sentence, etc.) 10. Scheme Change (rhyme and rhythm schemes, as in verse) Semantic: 1. Synonymy 2. Antonomy (translate a negation of the negative) 3. Hyponymy (translate with a more general term, or vice versa) 4. Converses (shifts in perspective – ‘modulation’) 5. Abstraction Change (concrete to abstract, or vice versa) 6. Distribution Change (saying the same thing with more or fewer words) 7. Emphasis Change 8. Paraphrase (rephrasing where some semantic components are disregarded) 9. Trope Change (mainly metaphor for metaphor, or non-metaphor) 10. Other semantic changes (modulations affecting sense of direction or space)



Disciplinary Corrections

177

Pragmatic 1. Cultural Filtering (naturalization, domestication, to conform to target norms) 2. Explicitness Change (includes implicitation) 3. Information Change (addition of new information) 4. Interpersonal Change (formal, polite, friendly, colloquial) 5. Illocutionary Change (statement to request, direct to indirect speech, etc.) 6. Coherence Change (relations between paragraphs units and above) 7. Partial Translation (only rendering part of a text, or summarizing it) 8. Visibility Change (possible foregrounding of the translator’s presence) 9. Transediting (radical re-editing of badly written start texts) 10. Other pragmatic changes (formal layout; use of language varieties) Later we also find Compensation, which seeks to maintain values of the level of the whole text and is thus logically free to move across all the above categories. A long list like this is able to include almost everything that was omitted from Vinay and Darbelnet: word-order change, resegmentation and the voice of the translator, for example, although one wonders why something as obvious as translators’ footnotes is not visibly included under Visibility Change.5 Now for my doubts. It is not abundantly clear to me why these three macro categories are necessary. I recall that Vinay and Darbelnet had each of their solution types able to operate on three different levels (lexis, collocation, message), so why should there now be separate strategies on each level? When you look at Chesterman’s definitions of the three levels, the differences are far from clear, and one need not be too punctilious to see it. For example, ‘syntactic’ strategies ‘manipulate form’ at sentence level and below (but doesn’t form affect content?) and yet they include not only Loan and Calque, which can surely operate on the lexis rather than the syntax, but also Level Shift, where there are actually changes between the levels of ‘phonology, morphology, syntax and lexis’ (1997: 99) – so it obviously is not all syntax – as well as verse forms (Scheme Change), which surely have nothing to do with sentence-level syntax at all? ‘Semantic’ strategies ‘manipulate meaning’ and include something of what Vinay and Darbelnet called Modulation, although changes in semantic perspective are not hard to find among the ‘syntactic’ team.6 They also have ‘aspects of clause meaning such as emphasis’, despite the fact that such things are also dealt with under syntax as Level Shift. One of these semantic strategies is Paraphrase, where ‘semantic components at the lexeme level tend to

178

Translation Solutions for Many Languages

be disregarded, in favour of the pragmatic sense of some higher unit such as a whole clause’ (1997: 104) – so why should this one not be a ‘pragmatic’ strategy? ‘Pragmatic’ strategies then concern the translator’s ‘selection of information’ in the translation, primarily in view of a prospective readership (1997: 107). This subsumes such things as Cultural Filtering, covering mundane lexical examples like ‘Familienname’ rendered as ‘Surname’ (rather than ‘Family name’). So if you go for the apparent calque (‘family name’), is your strategy then ‘syntactic’? And if you don’t, is it ‘pragmatic’? Surely this is a simple lexical problem, with the decision-making process, no matter which way it falls, on the same level? I don’t know. Anyway, ‘pragmatics’ is where we also find Explicitation and Implicitation, even though those things perhaps necessarily also involve Abstraction Change and Distribution Change, listed under ‘semantic strategies’. Here too we have Coherence Change, which concerns things like paragraph breaks, apparently – which is rather like Cohesion (‘syntactic’) but with respect to bigger units (and therefore necessarily ‘pragmatic’?). As for the inclusion of Partial Translation and Transediting, some would claim that these are no longer forms of translation (cf. Schreiber 1993). They are certainly solutions, and translators can certainly use them, but should we then also include alternative choices such as ‘refusing the job’, ‘giving the client advice’, ‘demanding more money’ or ‘passing the job to a junior colleague’? One senses that such things belong to a different league. Perhaps the most telling aspect of Chesterman’s macro division into three is the way the first block includes a zero-degree option (Literal Translation) whereas the other two do not (but surely they could?). Instead, there they have boxes for ‘others’, for the loose ends that may not fit elsewhere. The general impression is that there is no hierarchy or underlying organizational principle at work. These could simply be the things the researcher found in the previous literature, then identified in the in-flight magazine, perhaps during the two-and-a-half flight from Helsinki to Vienna. Chesterman appears to be eminently practical, even pragmatic: he simply tells us what he found, then puts it into some kind of convenient order. He disciplines the past, in the nicest possible way. So is there any ideology at work here? Yes, of course. To see it you just have to play with a few simple questions. First, there is an oddly silent meme at work in Chesterman’s presentation: each successive set of solutions is determined on the basis of textual and non-textual elements that increase in scope. In other words, the hypothetical decisions concern more things as you move further down the list; they involve



Disciplinary Corrections

179

more conceptual work. Chesterman does not say this, but he does it, instinctively, since this has been the dominant mode of presentation ever since Vinay and Darbelnet. He maintains this progression even when it would be logically advisable for him not to do so. Why else should there be a zero-degree solution (Literal Translation) for the first major block but not for the others? Chesterman then attempts to explain why translators should choose one solution type or another. It is, he says, because they are trying to conform to one or several of four competing norms: they want to do what is expected of them, what they should do with respect to the author and client, what will make the message clearest and what is appropriate for the particular start text. Chesterman’s translator is going to bend over backwards to submit to as many masters as possible, it seems. In none of this explanation, though, does the theorist ever mention the amount of work involved, or the simple idea that a translator might prefer solutions on the smaller levels because they will not involve so much hard work. Why such a strange absence? At one stage Marx was trying to understand why Aristotle saw economic value only in terms of exchange (‘one house = five beds’, in an age when houses were cheap, it seems) and not as being based on the embodiment of labour. The reason, suggests Marx, is that Aristotle did not see labour, work, as a basis for exchange: ‘Greek society was founded upon slavery and had, therefore, for its natural basis, the inequality of men and of their labor powers’ (1867/1959: 1.60). In this analogy, Chesterman plays the part of Aristotle, which I’m sure he will enjoy, and academic translators, the ones who do it for art or pleasure, might be Greek society: they do not see that translation is basically hard work. Instead of work, we find a conservative pressure to conform, on as many levels as possible and to as many norms as possible. This appeal to conformity does not sit well with later sections of Chesterman’s book, where there is a laudable appeal to ‘emancipatory translation’ (1997: 189ff.), but there it is. On the other hand, Gideon Toury, a theorist’s theorist, spent many years translating for money. When discussing the role of norms in translator training (1992), his approach is quite the opposite from back-bending conformity: he proposes we teach translators to break norms, not to obey them. Further, one of Toury’s proposed laws of translational behaviour (the ‘law of interference’) hypothesizes that translators tend to change things at the lower, smaller levels but not at the higher ones, where they save work by accepting start-text interference (1995/2012: 310–15): they tend to change sentence structures but not paragraph divisions or chapter organizations, for example. The key, in Toury,

180

Translation Solutions for Many Languages

is not the degree of translatorial conformity as such, but the amount of effort involved. You don’t have to be a genius to grasp that point, but you might have to be an academic in order not to see it.

Simple and complex decision-making The notion of the translator investing effort when choosing between solution types has been connected with a more psychological approach. Mayoral Asensio and Muñoz Martín (1997: 166; cf. Muñoz Martín 1998) propose a decisionmaking model for what they term ‘translation strategies’. This is based on the relative ‘space of the problem’, which basically entails counting the number of variables involved before a decision is reached. Their model takes the form of a flow chart that goes from simple to complex, illustrated in Figure 9.2 for ways of dealing with ‘culturally marked segments’ (what other theories ‘realia’). The model comes with many caveats: it is speculative, requiring confirmation in empirical studies; the order of decisions may alter according to the text type, the translator’s individual style and the degree of experience of the translator; and the decisions are assumed to be non-binary: ‘many potential solutions come to us and we often select one on the basis of aesthetics or intuition’ (Mayoral Asensio and Muñoz Martín 1997: 163). The resulting model is nevertheless of interest, at least as an intellectual exercise, since it has the potential to connect various solution types in a logical way. In fact, instead of each type being a transformation of a start-text structure, here each is seen as a modification or superior replacement of the previous type. There is work in this model. The model nevertheless begs many questions, and not just on the merely formal level (if the decision-making is not binary, why all the Yes/No questions?). The initial question, the assumption of relevance, is rather pleasing: it recalls the first step in George Steiner’s ‘hermeneutic motion’ (1975/88), where the translator trusts that the text is meaningful. It might also be attached to Grice’s ‘cooperation principle’ for conversation (1975): the translator must assume relevance in order to begin this conversation with the text. After that, however, we enter a dark and tangled flow chart. For example, the second question, the search for a ‘precedent’, could involve innumerable factors, indeed a complex search operation, which scarcely augurs for a simpler cognitive process than might the question of whether the start-language segment can be understood without translation. Surely Loan and Calque, placed here down at number



Disciplinary Corrections

Is the start segment completely irrelevant?

181

Yes

Omission (1)

Yes

Established translation (2)

Yes

Functional formulation (3)

Yes

Paraphrase (4)

Yes

Loan/Calque (5)

Yes

Combine 4 and 5 (6)

No

Is there a precedent for translating the segment? No

Does the TL have means to cause the same effect? No

Is the potential TL solution free of restrictions? No

Can the SL segment be understood in the TL text? No

Does the TL have means to cause the same effect? No

Create new segment (7) Figure 9.2  Translation strategy flow chart for‘culturally marked segments’ (realia), from Mayoral Asensio and Muñoz Martín (1997: 166; Muñoz Martín 1998)

182

Translation Solutions for Many Languages

5, could legitimately stand at number 2? Then we see the translator asking if the translation is ‘free of restrictions’ – drawing on a whole theorization of ‘constrained translation’ (Mayoral et al. 1988) – which we might grossly simplify as follows: if there is no room, opt for the shortest translation (which is sound practical advice, but surely it applies to all the solutions, at all the levels?). Lower down, the struggling translator, having realized that there are space constraints (on level 4), is asked to entertain a combined solution at level 6, which must surely take up more space? Further, the last question – ‘Does the target language have means to cause the same effect?’ – could surely have been asked at the very beginning, since ‘effect’ could actually be a much simpler variable to grasp than are degrees of relevance, precedents, constraints or understandability. In sum, this was a brave attempt at an impossible task, perhaps even a fool’s errand. There are so many individualized variables involved that a stable hierarchy like this seems bereft of psychological verisimilitude. Yet it does take account of effort. And that could be worth building on.

Disciplinary progress? I have received anonymous feedback on an article I co-authored in which the typologies of Loh (1958) and Vinay and Darbelnet (1958) were put to the test in the translation class (as we shall see in Chapter 11). The feedback is not wholly negative, and I am all the more relieved and grateful, as always. One reviewer is nevertheless worried that I used two clearly retro typologies in my classes (Loh and Vinay and Darbelnet) when those dated terms have apparently been ‘reconceptualized, debated and improved often over the ensuing (almost) six decades’. Really? My hidden reader openly regrets that I have not focused on what they call the ‘real historical advances in the discipline’, and then assumes, perhaps rhetorically, that some progress must have come from ‘cognitive data or other empirical tests’. The above pages have shown, I suggest, that the development of Translation Studies as a discipline has not produced any clear consensus around any new set of terms, and there is no trace of any resoundingly new approach to the question of translation solutions. Quite the contrary, what we find has more to do with uncritical repetition of old modes of thought, plus a few minor terminological forays based on occasionally shaky conceptual architectures. More worrying, there seems to have been no substantial contribution to the typologies (that I know of) from empirical studies of translation processes.



Disciplinary Corrections

183

Instead, we have a proliferation of rival terms and approaches, reaching such degrees of confusion that progress, were it there, would be difficult to recognize. Curiously, the same reviewer regrets, among much else, that I actually used old words, since it seems that ‘many scholars now use Recategorization for what was called Transposition’. Those ‘many scholars’ must be in the small coterie around Delisle’s Terminologie de la traduction (1999), which is the only place I can find where anyone is talking about Recategorization. Yes, the discipline is so small and so fragmented that each scholar’s terms are like a thumbprint and no one is really anonymous. And there are few signs of progress.

10

Going Japanese

Chapter summary: Given the many differences between English and Japanese, the application of Vinay and Darbelnet’s categories to that language pair requires additional attention not just to different scripts (as is the case for many languages) but also to specific modes of resegmentation, different tolerance levels for formal redundancy and the use of politeness formulas. The general discussion in Japan has nevertheless been on the level of comparative stylistics, based on the assumption that English and Japanese embody two completely different modes of thought. Such cultural and literary comparisons have tended to overshadow careful attention to what translators actually do. In the meantime, East Asian languages are bending themselves toward English, riding roughshod over the supposed essential differences and deploying an impoverished range of translation solutions, as indeed was predicted by the ideological warnings back in the days of Vinay and Darbelnet.

For some years I have been teaching solution types in my Translation Practicum course at the Middlebury Institute of International Studies at Monterey, albeit mostly for just two hours each year. Part of the constant joy of that class has been the opportunity to explore translation processes between English and not just French, Spanish, German and Russian, but also Chinese, Korean and Japanese, all in the one class. That group of students has pointed me toward some of the pedagogical shortcomings of the Vinay and Darbelnet typology and has enabled me to test to what extent it is suited to work with European languages only – as will be reported on in the next chapter. Most of the testing has been with English–Chinese translation, since the majority of the students work with Chinese. Some of the more fascinating problems, however, come from Japanese, which can be further away from English in terms of syntax and conceptualization processes. Here I want to focus on that apparently more extreme otherness.

186

Translation Solutions for Many Languages

Perhaps surprisingly, Japanese translation scholars seem not to have developed their own typologies of translation solutions – or at least, none has been reported to me. Hasegawa’s Course in Japanese Translation (2011) reproduces the seven fundamental Vinay and Darbelnet ‘techniques’, to which she interestingly adds Omission, Addition and modes of text re-organization, which we will look at below.1 In discussing these additional solution types and indeed in her suggested class exercises, Hasegawa nevertheless incorporates established studies that compare Japanese and English on the level of grammar and stylistic preferences. There are several such studies, but the ones with the most relevance for translation solutions seem to be by Tetsuo Anzai, a translator into Japanese (of Shakespeare, among others). Anzai offers very practical observations on the way the structure of English concerns translation (Anzai 1982a, 1982b, 1983). He has also written with the Japanese–English translator Edward Seidensticker on the translation of the Japanese texts into English (Seidensticker and Anzai 1983). In the second half of that collaborative book, Seidensticker comments with colour and irony on his own translations into English. The work is intriguing not just because it teams up two literary translators going in complementary directions, but also, in the present context, because it uses translation to illustrate wider stylistic tendencies, without any formalization of translation solutions. In this, the general approach is strangely reminiscent of Bally – translation is there, but the main thing is to compare the languages. All in all, the presence of an actual typology of translation solutions for Japanese remains fairly anecdotal: Hasegawa’s coursebook uses Vinay and Darbelnet as a convenient way of organizing a section, and little more. What interests me here is the possibility that, in the case of Japanese, the linguistic differences might be so extreme, requiring so many transformations or reconceptualizations, that the very idea of a typology falters, perhaps because isolated solutions are simply too simple. The Western typologies could be adapted to Chinese; they seem not to have been of much interest for Japanese. I would like to know why. The occasion to explore this question came in 2013 when Russell Scheinberg, a Master’s student in the Japanese programme at the Monterey Institute, needed a thesis topic to work on. I suggested that he might like to test whether the Vinay and Darbelnet model could be applied to Japanese. Most of this chapter draws on what Russell found. Another occasion came a little later in 2013, when I was external examiner for a PhD thesis by Akiko Sakamoto (2014), part of which analysed interviews with eighteen translators working between Japanese and English, focusing on what kind of theorization they used. I also came across



Going Japanese

187

Bode (2009), who attempts to illustrate Chesterman’s ten ‘syntactic/grammatical strategies’ with examples involving Japanese. More consistently engaging than these, though, has been the work of Anzai (whom I read through Scheinberg’s reports) and Seidensticker, who published enough in English to be the object of a small case study here. So I have benefited from a small range of sources, especially Scheinberg, although the words (and ignorance of Japanese) are mine. I necessarily write here as cultural outsider, starting out on a small adventure. Here I will present an encounter between Japanese and the basic Vinay and Darbelnet categories, plus a few extensions suggested by the above sources. I will then consider some of the basic political positions involved. But first, let me start from some of the specific features of Japanese, focusing especially on the writing systems.

The uses of script Chinese, Korean and Japanese use logographic characters. This difference in script marks out the most obvious problems that are unlike those between languages that use the same script. Here the written word looks very different. Then again, Japanese also has syllabic scripts, and these three Asian languages are all able to represent sounds, each in a different way, so the common distinction between characters and words is by no means absolute. Japanese and Korean can spell out the sounds of co-ca-co-la (コカコーラ, kokakōra in Japanese, or 코카콜라, ko ka kol la in Korean) while Chinese can at the same time rework the semantics: in Chinese Coca Cola is 可口可乐 (kekou kele), which word-for-word is ‘can mouth can joy’ – joy in your mouth! Further, the Westernization of these cultures is now at a point where the foreign term can be written and accepted in the Latin alphabet, which now solves many an apparent problem (the bottles of Coca Cola have the words written exactly as in the United States, alongside the versions in logographic script). When it comes to solution types, the different scripts offer further layers of alternatives. We saw that Loh (1958) recognized five or so solution types for rendering foreign terms, adding Symbolic Translation and Coinage of New Characters to what he might have picked up from English or Russian, and his tendency was to list even more (to the extent that mixed solutions were added to the list). The fact of script, as a visible complexity, is easily swept up into assumptions of a more extreme otherness, even when the problems themselves are not of any great complexity. I suspect that script was mainly

188

Translation Solutions for Many Languages

important for solution types at a time when equivalents had to be invented for huge quantities of foreign terms, and that it is now not such a problem in other contexts (it strangely plays no key role in the contemporary textbooks for training translators in Chinese and Japanese). A simple experiment might illustrate this. If script were now a significant source of problems when going from English into Asian languages (independently of standardized transcription systems), it would surely be just as much a problem when moving the other way. But it is hard to imagine enormous theoretical acrobatics being performed just for the few terms making the trip: kung fu, typhoon, chaebol and the like. In the case of translation into Japanese, the theoretical mistake too easily triggered by imaginary absolute foreignness is to assume a pristine target language that is going to be upset by the presence of the foreign. Japanese draws its lexicon from indigenous, Chinese and Western sources, with each more or less being expressed in its corresponding script: hiragana for indigenous Japanese words and grammatical particles, kanji for words of Chinese origin and katakana for words of Western origin, to which we can now add the use of Latin script, either borrowed directly as in a brand name, or as a Latinized form of Japanese (rōmaji), mostly for translation into English. In a sense, this plurality of scripts pre-empts the need for the translator to make radical decisions about them – the fundamental choices have generally already been made, or better, they have been made at the level of the script system rather than in the translation situation. Then again, script can be a stylistic option in some cases: cultural products for children tend not to use kanji, for example; words that are normally in katakana can be written in hiragana to make them sound more feminine, apparently in comic books for young girls; and katakana can be used to mark a foreigner speaking Japanese with an accent (Gyogi 2015). These are cases where script provides the translator not just with more constraints and problems, but also with additional expressive resources. When problems occur in the more literary texts or in subtitles, the Japanese, Korean and Chinese scripts can also have recourse to rubi characters (furigana in Japanese), which are added in smaller type above or to the side of a character that requires explanation. In Japanese this is usually done when a Chinese character is unfamiliar and some indication of its pronunciation is needed. But rubi can also be used to give information about the meaning of a foreign term, to add nuances, to indicate phonetic values (particularly to explain foreign puns) or more indeed to offer a second, destabilizing reading



Going Japanese

189

(cf. Wakabayashi 2006). For example, in Murakami’s translation of the short story ‘The Proxy Marriage’ by Maile Meloy (Murakami 2013), a pun on the name of the character Bridey (代理花嫁) is partly explained by addition of the rubi characters プロクシー・ブライディー, spelling out the English pronunciation.2 This effectively becomes something like the insertion of a translator’s note in the text, or the presentation of two complementary translation solutions, perhaps like Newmark’s category of Translation Couplet (1981: 31). Japanese scripts involve initial complexity, yet they also help to solve some basic problems and offer novel resources for further solutions. None of this finds any place in the Vinay and Darbelnet typology, for obvious reasons.

Transporting Vinay and Darbelnet Beyond the problems of script, why not apply Vinay and Darbelnet to English– Japanese translation? Are the differences between the languages really so enormous? And even if the application fails, can it reveal new problems and perhaps lead to new categories? Here is what broadly happens when you try to force European categories onto translation between English and Japanese.

Borrowing and Calque Japanese is full of visible borrowings, both old and new, and the pertinence of Borrowing as a solution type could scarcely be doubted, if and when we can relax constraints on use of the same script. Examples abound in the textbooks, involving degrees of reassignment that are at once syntactic and semantic. A nice case is the word アベック (abekku), which is a phonetic representation of the French avec, used not as a preposition (as in French) but as a noun to refer to a couple in love (Scheinberg 2014: 40). Borrowing can also involve a degree of mixing and matching, as seen in some of the word formations. The Japanese word for semiconductor, 半導体 (handotai), uses Chinese characters to express Latinate roots: han (semi-, ‘half ’), do (conduct-) and tai (-or, ‘actor, thing’) (Scheinberg 2014: 18). These are examples of words that have come into Japanese through translation; they cannot be taken as illustrations of the problems that translators have to solve on a daily basis. Borrowings nevertheless present interesting problems when translating from Asian languages into English. In some genres you often come across an

190

Translation Solutions for Many Languages

expression that you know came from English, so logically you just ‘return’ it to English. One of my former students, Scott Myers, coined the term ‘Loan Returned to Sender’ for this solution type (actually when translating international news reports from Chinese) and it might be remembered as the Scotian Return. The trap, of course, is that there are often deviations en route. Scheinberg (2014: 38–9) gives the example of the Japanese タイミング (taimingu), which is visibly borrowed from the English word ‘timing’ but can be used in different ways. When it appears in a patent application in a phrase that translates literally as ‘an arbitrary timing’, it makes little sense to return the loan directly: ST: 任意のタイミングで所望の商品の関連情報を … 提示するため … [In order to present … related information for a desired commodity at an arbitrary timing … ] TT … at any time Calques, understood as a borrowing of syntax or morphological construction, are mentioned far less frequently in the literature. Scheinberg (2014: 18) comments only on calques from Chinese; Bode (2009: 17) finds no Japanese example to illustrate the solution; they are not mentioned as a solution type in Sakamoto (2014); they are only very rarely listed by my Japanese students when they analyse their translations. Could this simply be because the syntactic system operates in a very different way? Or could there be something else at work here?

Word for what? English is basically a Subject-Verb-Object language (‘Mary sang a song’); Japanese tends to be Subject-Object-Verb (‘Mary song sang’), perhaps in the way that Yoda speaks English in Star Wars (in the Japanese versions his grammar is perfectly normal). English has ‘prepositions’ because they come before (pre-) their complement (‘to the station’); Japanese has ‘postpositions’, which come after (post-) the complement (‘station to’); English likes to place qualifiers before the noun; Japanese is also able to place qualifiers after the noun, with considerable freedom, and can stack long adjectival phrases before the noun; English can place important elements at the beginning of a clause; Japanese likes to put them at the end. That much is easily said and can be accounted for in the translation process. But then there are trickier things: Japanese, for instance, can omit the subject fairly freely; it has no relative pronouns; and it uses markers



Going Japanese

191

and particles in order to indicate grammatical functions (as do the inflections in inflected languages), which means that its word order does not have to do this and can thus express stylistic values. These features present obvious problems for any concept of ‘word for word’ Literal Translation between Japanese and English. Not only are there many words in English that find no grammatical counterpart in Japanese, there are also grammatical particles in Japanese that correspond to no particular words in English. Further, given the syntactic freedom of Japanese and its different basic pattern, there are cases where the order of grammatical units can be precisely the opposite from English. Scheinberg’s way around this problem is to suggest that the translation solutions are not word-for-word but ‘chunk-for-chunk’, where a ‘chunk’ is basically identified at phrase level but can be modified to suit the needs of each analysis.3 A simple example is borrowed from Anzai (1983: 201): ST: Venice is built on a number of sand islands. TT: ヴェニスは、たくさんの砂地の島の上に立っている。 Venice WA [topic marker] many island of sand on [above + locative particle NI] stands. Over and above the clear shifts in element order, Scheinberg (2014: 16) analyses this translation in terms of a series of quite minimal transformations and non-transformations: Venice = Venice is built = stands on = on a number of = many sand islands = sand island Seen in these terms, it is indeed possible to talk about something like Literal Translation, not as a word-for-word process but at the level of phrases, more or less, and allowing that the syntax marches to a different drum in the two languages.4 Perhaps surprisingly, a kind of non-syntactic Literal Translation can thus survive as a basic solution type. There are, however, many other problems in search of solutions.

192

Translation Solutions for Many Languages

Transposition As might be imagined, a semantic value that is expressed in one grammatical form in English can often be expressed in a different grammatical form in Japanese. So there is considerable Transposition, and this is where the linguists enter into speculation on the tendencies underlying the two languages. The general observation is that English prefers nouns where Japanese prefers verbs, so there are many corresponding shifts. Scheinberg (2014: 47ff.) suggests ‘denominalization’ as a specific type of Transposition when translating from English into Japanese. Hasegawa (2011) nevertheless gives class exercises where students are instructed to translate Japanese sentences ‘first with a verbal construction and second with a nominal construction’, pointing to the optional nature of these solutions. The following is an example from Seidensticker and Anzai (1983: 54), commented on by Scheinberg (2014: 50): ST: 長い間端座の形を崩さずにいたので、努力しなければ立ちあが れなかった。 [For a long interval, had not changed his kneeling position, so without straining would not have been able to stand up.] TT: The long hours of unrelieved kneeling had so paralyzed his legs that he could pick himself up only with a special effort. Here the verb phrase ‘[not] changed position’ (形を崩さず) has become the adjective ‘unrelieved’, and the verb values ‘to make an effort’, ‘strive’ and ‘strain’ (努力しなければ) have become the adverbial ‘with a special effort’. Anzai argues that in the English translation, ‘things corresponding to nouns are extracted from the situation as a whole and the relations between the extracted nouns are shown with prepositions and verbs’ (trans. Scheinberg 2014: 50).5 With all due respect, it does not seem that simple, since this example also uses Explicitation: the English has explicitated the noun phrase ‘legs’ and the verb ‘paralyzed’. There are numerous other types of Transposition at work in translation between Japanese and English, and most lend themselves to hypotheses of general systemic tendencies. Without going into the details, let us accept that Transposition is a pertinent solution type, perhaps even a dominant one, to the extent that ubiquity makes it an invisible norm. The problems kick in when one tries to separate it from other solution types.



Going Japanese

193

Modulation If we accept that Modulation involves a change in perspective, thus seeing the same object from a different point of view, then it too appears to be fairly ubiquitous in Japanese–English translation. But is it easy to separate from Transposition? Some examples are deceptively easy. The following is from Kanaya (2004): ST: I see Mt Fuji TT: 富士が見える [Fuji ga mieru] [Fuji is visible] This looks like a passive transformation with deleted subject, a clear change in perspective, without Transposition because it is technically verb-to-verb: the transitive ‘to see’ is replaced by the Japanese verb 見える (mieru, to be seen, be visible), which indicates a process without an agent. As mind-bending as that may seem, it puts no spanner into the workings of the basic solution types.6 Things become rather more difficult, however, when you consider all the movements between noun-focused and verb-focused structures, which technically count as Transposition but are in most cases clearly also instances of Modulation. This can be seen in the ‘kneeling’ example above, where the Japanese does not state the subject and thus obtains a perspective that is more immediate and experiential, as opposed to the more external perspective of English. This kind of perspective shift is frequent (it is also in the ‘Fuji’ example above), making it difficult to locate examples of Transposition that are not also examples of Modulation. The conceptual distinction can still be made (two functions are allowed to co-habit the same solution), but it becomes hardly worthwhile. An endlessly analysed example of Modulation is Seidensticker’s rendering of the first sentence of Snow Country by the Nobel Prize winner Yasunari Kawabata (here I am adapting the more literal translation given in Maynard 2007: 57): ST: 国境の長いトンネルを抜けると雪国であった。夜の底が 白くなった。 [Country border long tunnel when pass snow country. Bottom of night turned white.] TT: The train came out of the long tunnel into the snow country. The earth lay white under the night sky.

194

Translation Solutions for Many Languages

Kanaya (2004) sees this as an example of the way Japanese adopts an ‘on the ground’ perspective: in the Japanese, the reader is placed on the train, while the English places the reader as an observer outside the train. For Scheinberg (2014: 60), ‘when no subject is supplied, the missing grammatical subject acts like a vacuum that “sucks” the reader into the action’. The translator into English, though, apparently needed a subject and thus supplied the train, which is not named in the Japanese. There are Transpositions and Explications happening all over the place here, but the most prominent result is clearly Modulation. Many similar observations can be made concerning the opposed preferences of Japanese and English: a ‘participant perspective’ vs. a ‘universal perspective’; a ‘process orientation’ vs. an ‘actor orientation’; ‘passive view of nature’ vs. ‘active human agency’; and so on. Intriguing as they are, these differences would nevertheless explain the reasons for Transposition and Modulation, rather than constitute new solution types in themselves. They also run into a major theoretical problem: if the cognitive and cultural differences are somehow embedded in the language systems, are the translation solutions then optional or obligatory? I will come back to this (and to Seidensticker’s translated snow country).

Idioms and adaptation? There are cases of idioms that happily correspond between Japanese and English: 猫に小判 (‘gold coins to the cat’) renders ‘pearls before swine’. The far more common case, however, is of expressions that do not correspond and thus require some degree of Adaptation. This particularly concerns cultural values. Seidensticker and Anzai (1983: 128) find a beautiful example of non-corresponding cultural values in the word すなお (sunao,‘docile, amenable, yielding’), which is a positive virtue in Japan but is not usually positive in the West. This is from Seidensticker’s translation of Kawabata’s ‘The Izu Dancer’: ST: 私は非常に素直に言った。泣いているのを見られても平気だった。 … I said very docilely. Even though seen crying, was calm] TT: I saw no need to disguise the truth and I was quite unashamed of my tears. I’m not sure what trip takes one from ‘docilely’ (sunaoni, meekly, tamely, obediently, submissively, pliantly, like a lamb) to ‘not disguising the truth’, but the second phrase makes sense in context and conveys a positive cultural value, as does sunao in Japanese. As a solution type, it is certainly available. As a



Going Japanese

195

concealment of cultural values, it deserves to be questioned. But it is certainly not the most extreme case.

Resegmentation Although the possibility of cutting up or joining sentences is not addressed in Vinay and Darbelnet, it certainly appears among the Chinese theories dating from the work of Zhang et al. (1980). So should Resegmentation, understood in this general way, be pertinent for work into and from Japanese? In the case of Chinese, we saw the claim that English clauses are longer than the Chinese norm, so translators into Chinese may want to cut a long English sentence into two or more Chinese sentences. Japanese, however, is reported as liking clauses and sentence structures that are rather longer than in English (Anzai 1983), so the tendency when going into Japanese might theoretically be to join sentences together. Sakamoto (2014: 184) nevertheless lists ‘dividing a long sentence into two or three sentences’ as one of the solution types (‘strategies’) mentioned in her interviews with Japanese translators (who were working both into and from Japanese) and there is no mention of joining any sentences together. Hasegawa (2011: 188) enlists examples from Steidensticker and Anzai (1983) to illustrate the ‘high tolerance of verbiage in Japanese prose’, which might explain some extra length in Japanese. She offers the general recommendation that when going into English the translator should ‘disregard extra verbiage’ (2011: 189), especially with respect to repetitions and phatic elements. The imposition of English grammatical efficiency could thus result in some joining up of sentences when going in that direction. Translators, however, tend to be averse to creating stylistic complexity, which is usually an inefficient use of their cognitive resources, with minimal conceptual gain for the receiver. I would bet that translators cut up sentences in one direction significantly more than they join them up when going in the other direction, in any language pair, but this is a bet because I really don’t know. So the different sentence lengths and degrees of discursive tolerance could theoretically justify adding Resegmentation to the list of solution types, as has been done for Chinese. The practical lesson is nevertheless that, as Anzai (1983: 19) puts it, there is no reason to retain sentence breaks ‘for their own sake’.

196

Translation Solutions for Many Languages

Politeness: Compensation or Omission? Japanese politeness language (keigo) is rather more problematic. It can be found in many parts of speech, from the need to choose from about a dozen ways of saying ‘I’ according to the relation with the second person (admittedly depending on gender, as well as social relation and formality), to the use of the auxiliary verbs naru ‘to become’ and suru ‘to do’ in order to make a person’s actions appear like a natural process and thereby show respect. There is virtually nothing like this in English, which can at best approximate the choices by fiddling with register. So what is to be done? This is an interesting problem. In Vinay and Darbelnet’s world, all cultures had basically the same things to say: if the French maintain a distinction between formal and informal second persons, the degree of formality can still be expressed in English in various ways – Compensation is possible. The implicit justification is rather like rendering ‘docility’ as ‘telling the truth’: as long as the text functions are about the same, should we really care what linguistic values were in play? In the world that straddles Japanese and English, however, it seems precarious to assume that exactly the same values are in play. There are degrees of Japanese politeness conventions that have no counterpart in English, and the logical temptation, after you have fiddled with register all you can, is just to delete them. So does Omission necessarily become a substantial category? If you look at the websites of a few Japanese–English translation companies, you will probably find this problem explained as a necessary choice between the desu-masu and dearu styles, with the former being very polite and the other more neutral. The companies actually explain this choice to their clients, listing the advantages and risks of each style for the main text types. In Japanese, the formal desu-masu style can be used for technical texts, and instructions may appear to be too forthright or abrupt if it is entirely absent. At the end of the day, for translations into Japanese, this is a marketing decision, apparently to be made by the people paying for the translation. When translating from Japanese, however, Omission is relatively unproblematic in the case of technical texts. In the case of texts that are supposed to tell us something about Japanese culture, the choice remains difficult.



Going Japanese

197

Omission as temptation One of the main categories that Hasegawa (2011) adds to Vinay and Darbelnet is Omission. As we have seen, she particularly emphasizes the virtues of deleting what she terms ‘phatic communion’ and ‘verbiage’ – she is not just talking about Implicitation. The ‘phatic’ linguistic function technically concerns everything that tests whether the communication channel is working, including utterances of the kind ‘Can you hear me?’. As anyone who has witnessed a conversation between Japanese speakers will realize, this mainly concerns the many instances of ‘backchannelling’ (aizuchi) in which listeners confirm that they are still involved in the discussion. Back-channelling also occurs in English–English conversations, but not to the same extent. This could lead to misunderstandings if translated literally, since the heightened frequency of back-channels might not only seem excessively strange but could be misinterpreted as expressing impatience with the speaker or perhaps agreement,7 neither of which is the case in Japanese. In view of such factors, Omission is perhaps the most advisable solution. That said, the presumptions of distaste or misunderstanding may prove a little too hasty. In research on cross-cultural English-language conversations between bilingual Japanese women and monolingual American women, White (1989) found that both sides adjusted their use of back-channels to approach the other’s, and that the frequency of back-channels correlated positively with participant satisfaction rather than with uneasiness or misunderstanding. That is, in non-translational interaction, people accommodate to the other and may appreciate foreignness. And yet translators are recommended to omit the traces of alterity, for fear that receivers might not be able to cope. Here we reach the politics of the deal.

So what is obligatory, what is political? You can probably see where this is heading. Most of the Vinay and Darbelnet categories can indeed be applied to Japanese–English translation, in the way that square pegs can be banged into round holes, if your hammer is heavy enough. Most of the action happens on the level of Transposition and Modulation, to the extent that it is often not worthwhile separating the two, and special emphasis should be placed on solutions able to deal with the different scripts, the intricacies of politeness language and the possible reasons for Resegmentation. The

198

Translation Solutions for Many Languages

harder questions nevertheless concern the temptation to Omission when translating from Japanese – there is a certain desire to eliminate many things that simply sound non-Western. That same desire actually informs most of the options, I suggest. To what extent are the many examples of Transposition, Modulation and Adaptation really obligatory? In the case of grammatical categories that simply do not exist in English, it is easy to say that non-literal translation is necessary, but the various remaining solutions are then subject to considerable choice and creativity. And what of the grammatical categories that are recognizably similar? There, too, real decisions need to be made. So what are the factors impinging on such choices? The translator Edward Seidensticker, cited several times above, left some lengthy commentaries on his translations. As perhaps the key translator of the post-War Japanese novel, he was not an insignificant figure. Eloquent, ironic and self-asserting, he might help to illustrate why the various options are scarcely neutral. In an early essay published in 1958 (the same year as Vinay and Darbelnet, and Loh), Seidensticker admits that Literalism is possible when translating from Japanese, but rarely advisable: Most translators will decide, after the Oh’s and the Ah’s, after the blubbering that never seems to strike the Japanese as sentimental and therefore presumably isn’t, after pages on end in which ellipsis alternates with pleonasm, that something must be done. (1958: 13)

‘Something must be done’ means that the more risk-laden solution types have to be brought into play. But note the tone of the passage. In 1958 it was apparently still possible for a translator to be overtly unsympathetic to the cultural other, and the disparaging tone of ‘blubbering’ and so on is now quietly shocking – ‘something must be done’ connotes a swift assumption of cultural superiority, which was perhaps of its age. Seidensticker worked as a language officer with the US Marines in the Pacific War, going ashore at Iwo Jima ‘loaded down with dictionaries’ and staying on as a military translator in occupied Japan (New York Times 31 August 2007). The disparaging comments referred to a defeated enemy culture. Something similar is found in a few asides by Edwin and Willa Muir (translators of Franz Kafka, Gerhart Hauptmann, Heinrich Mann and others), from the same years, on the anal qualities of the German sentence, likened to a national predilection for sausages (1959: 95). Those were the days! That kind of thing should now probably be written off as transitory excess, were it not for its



Going Japanese

199

underground survival in cases like Hasegawa’s dismissal of Japanese ‘verbiage’. The sound professional recommendation could find a political history. Then again, consider the opinion of the Harvard philosopher Willard Quine, from the same years again, that ‘wanton translation can make natives sound as queer as you please’ – at that time ‘queer’ still meant ‘strange’ – and that ‘better translation imposes our logic upon them’ (1960/2013: 53). What is the moral virtue in making the other sound strange, in a way that invites no more than easy dismissal? And Seidensticker is not actually saying that the ‘blubbering’ is bad: if you read the whole sentence, he is assuming that the verbiage does not have an emotional effect on the Japanese and therefore is more superficial than substantial. Hence a more profound common humanity is better served by getting rid of it, it seems. That was also the age of transformational grammar, when the superficial expression could be undone in search of something deeper and more universal, which was also usually something closer to ‘our logic’. Omission was not just an option; it was an option with ideological underpinnings. One could try the same argument with politeness language: since politeness is all over the place in Japanese political language, for example, and yet Japanese politics is as ruthless and calculated as any, the politeness is superficial, misleading and therefore to be done away with when translated, again in the interests of a deeper common humanity. Make everyone sound like a Westerner and basic understanding will result. The spurious nature of the argument should be obvious. The epistemological weak point is that the Western eye is distinguishing between what is superficial and what is substantial (or what is surface-structure and what is a kernel, in transformational terms), without checking whether anyone else would call the play that way. The strategic error is that Westerners brought up on these illusions will not be able to operate in the cultural intricacies of Japanese interactions (in Japanese or any other language) – hence the proliferation, after the age of domesticating translations, of crash courses in Japanese business culture and the like. And the historical error is to believe that the true essence of the other, translated or not, will always remain that way, without being affected by accumulated cross-cultural interactions and without awareness of what the translations are actually saying. At the end of the day, the initial translation problem is more easily solved by an appeal to purpose: if the aim is to show Japanese ways, then render the Japanese ways. If not, seek things in between. Seidensticker stayed in Japan, becoming part of the inner network of Westerners who interpreted Japanese culture not just for outsiders, but also for

200

Translation Solutions for Many Languages

the Japanese. This was a relatively marginalized translation culture, often gay, frequently reveling in the multiple ironies of their border status, and not likely to maintain the discourse of cultural superiority over an extended historical period. In Seidensticker’s autobiographical Tokyo Central (2002: 122) we read, ‘Tanizaki [one of the major authors Seidensticker translated] wrote clear, rational sentences. I do not, certainly, wish to suggest that I disapprove of such sentences; but translating them is not very interesting.’ At the end of the day, if there were nothing but a common humanity to show, there would be few challenges for translators and no place for their particular prowess. Seidensticker also comments directly on his translated opening of Kawabata’s Snow Country, the one where he inserted the train. He openly enjoys the difficulties of Kawabata’s text: ‘from such difficulties came much of the pleasure of translating the work and in them is the essence of Kawabata’ (2002: 127). In the development from military linguist to renowned literary translator, disparagement turned into a form of professional gratitude. This is in the context of recommending to young translators that they work hard on the first and last sentences of a work, since those are the parts that are most likely to be read. And then Seidensticker ironically regrets having put so much effort into this particular sentence. Here is the passage again: ST: 国境の長いトンネルを抜けると雪国であった。夜の底が白くなった。 [Country border long tunnel when pass snow country. Bottom of night was turned white.] TT: The train came out of the long tunnel into the snow country. The earth lay white under the night sky. I referred to this above as a case of Modulation, since Seidensticker chose to view the scene as someone looking at the train from the outside and this, we have seen, was enlisted in arguments about the specific nature of Japanese. And yet, looking at the passage now in terms of option or obligation, it seems to me that a little free indirect discourse (erlebte Rede) could have done the job: ‘The long border tunnel passed and then it was snow country down there’, or somesuch. Modernist literary English has no real trouble narrating actions from the perspective of a character. Strictly speaking, the train was not necessary, was it? Seidensticker, however, expresses no compunction about the train. Instead he notes that he has been criticized for omitting ‘bottom of night’, about which he does express regret: ‘It is a striking image and the chief reason for the great fame of the passage and it should be there’ (2002: 124–5).8



Going Japanese

201

So why regret the bottom of the night (simple Omission) but not the train (complex Modulation)? One reason could lie in the audience for the translation. Seidensticker notes with reference to these sentences: They have been the most thoroughly scrutinized lines in any of my translations. Most of the scrutinizers have been Japanese. Given the esteem in which the passage is held, this is not, perhaps, so curious. Nowhere else in the world, I suspect, are translations from their literature given the attention they receive in Japan. Americans, at least, care not at all. (2002: 124)

So for whom was Seidensticker translating? Sakai (1997) makes much of the possibility that translation is ‘writing for two different audiences’, addressing speakers of both the start and the target languages. The possibility may indeed be universal; the actual reading and analysis of out-translations could nevertheless be especially Japanese.9 Quite apart from the international marketing of national culture (especially the dependence on translation for something like the Nobel Prize), the Japanese interest in translations from Japanese would seem to go hand-in-hand with a more intimate concern with identifying national cultural specificity, which starts from the identity of language. The implied train enables the Japanese reader to say, look, here is what Japanese does and English does not; this is what we are (as if the Japanese text were not by Kawabata and the English by Seidensticker, both individuals rather than cultures). The translation of individual language use enables the collective self to be defined. In this situation, all sides have a certain interest in highlighting the difficulty of translation and the need for the more creative solutions. The Japanese use it to define their national specificity on the basis of language; the American translator uses it to define and justify his craft; the American readership might accept it, if they cared, as an imposition of their logic. The regretful translator was thus prepared to add the literalist detail (the night with something white under it), at the expense of a minor stylistic foible (white, night) and yet retain the external, objectifying perspective, as if it were a Western necessity. This is not to suggest that mutual interests operate here in any symmetrical way. A culture that is concerned with how it is translated implicitly negotiates from an internally problematized identity; a culture that does not care how it is translated implicitly lives within an assumed superiority.

202

Translation Solutions for Many Languages

Selling out to the other? My sources tend not to like what I am suggesting here. The people who write on Japanese–English translation tend to be Westerners or Westernized Japanese, generally positioned against the tradition of nihonjinron, the ‘theory of the Japanese people’, which Scheinberg (2014: 14n.) glosses as a ‘best-selling genre in Japan, which endlessly expounds on and evaluates what is special and different about the Japanese, [seeing] English and Japanese as somehow irreconcilably different and demonizing one and idealizing the other’. The work of Anzai and Seidensticker, empirically comparing pieces of language and attempting to describe exactly where and how English and Japanese have differing tendencies, would be ideologically opposed to this nihonjinron view (Anzai 1983: 217ff.). That ideological battle, though, does not concern the intricacies of translation as such: it takes place on the other side, amid debates over absolute or relative cultural difference, and over what can or cannot be said easily in a national language. Given the terms of those debates, a typology of translation solutions was not likely to be high on anyone’s agenda in Japan, whereas comparative stylistics certainly has been. That could be one reason why typologies of solution types have no more than an anecdotal presence in this field. A second reason, more perfunctory, might lie in the way translator training has evolved in Japan: it has traditionally been in private institutions and language schools rather than at university level. Although translation has certainly been taught in Foreign Language departments, it has tended to be approached as a way of reinforcing language study. This would again explain why attention has been given to comparing languages but not to actual translating. Even outside Japan, Howard (2011: 26) finds that the graders of the Japanese translation examinations administered by the American Translators Association implicitly condone equivalence at word level and thus ‘have a tendency to view the short passage test as a language test rather than a translation test’. No one really needs elaborate solution types in order to test whether a language has been learned. Indeed, the more adventurous solutions are probably best avoided. The strange thing is that, in the absence of any prolonged attention to the more creative translation solutions, the Japanese language has not stayed locked into a fixed or essential identity. More than fifty years ago, Seidensticker noted one problem that, he said, could occur when translating from Japanese but not into Japanese: ‘How is one to find an equivalent for a style whose most



Going Japanese

203

conspicuous feature is, say, a straining to introduce relative pronouns [which Japanese does not have]?’ (1958: 18). That is, how could one translate from a language that stretches itself to incorporate features of the language you are working into? And later in the same piece, reporting on a phone conversation: ‘The Japanese sentence has taken on a certain twentieth-century efficiency by being dismembered and reassembled in something like the English word order’ (1958: 13). That was more than fifty years ago. Since then, the process has accelerated, in all the major Asian languages. Perhaps the Japanese scholars make little mention of Calque, for example, because it is somehow everywhere in their contemporary language. Yamada (2011) reports that Japanese translation memories give greater productivity when they contain solutions that are more literal renditions of the corresponding English phrases, so it is not hard to see translation technology as one of the reasons behind an accelerated absorption of English structures. A probably more powerful reason, of course, is the cultural prestige of Western culture and the traditional capacity of Japanese to absorb the foreign. That combination of prestige and tolerance produces the nationalist backlash, the proclamation of perpetual linguistic difference and the need to prove it on the basis of comparative stylistics and apparently supportive translations. In this situation, there remains a mutual interest in the more adventurous translation solutions at the level of Modulation and above. Those solutions do serve the causes of cultural and linguistic specificity, in Japan just as in Francophone Canada. At the same time, they are able not only to encourage languages to be creative within their own resources, but might also help to make translations work as communication events.

11

The Proof of the Pudding is in the Classroom

Chapter summary: Classroom experiments with the categories formulated by Vinay and Darbelnet (1958) and Loh (1958) indicate that students prefer the categories that have been developed closest to their cultures. There is nevertheless widespread difficulty with understanding Vinay and Darbelnet’s Literal Translation and distinguishing between Transposition and Modulation. With respect to Loh, students have doubts about Repetition and especially about what is obligatory and what is optional. In general, students’ translations between European languages find a baseline at the level of Literal Translation, whereas the line flitters between Transposition and Modulation when translating between English and East Asian languages. These experiments give rise to several critical reflections and thereby lead to a few suggestions for improving the traditional categories.

I have mentioned several criticisms of the entire tradition of solution types, but I am by no means the first to do so. Jean Delisle (1988: 72–3) raised the following questions about the epistemological status of the categories: Vinay and Darbelnet’s translation procedures do not help the translator to find translation equivalents. A procedure is a method to obtain a result, a way of doing something, of carrying an activity through to its conclusion. But these ‘procedures’ are in fact labels attached to results; the authors describe structural changes that occur in the translation process, or point out what does not change … the categories of comparative stylistics (and particularly the so-called translation procedures) cannot really be applied to the analysis and re-expression of messages, or even the verification of equivalences.

What Delisle is questioning here is the claim, which we have found explicitly made in Vázquez-Ayora and was the ideological underpinning of transformationalism, that the analysis of texts as products can tell us how those texts were produced. Delisle is quite right: the products are not the processes and should

206

Translation Solutions for Many Languages

probably not be mixed up with them. Especially now that we have tools and methods like think-aloud protocols, screen recording, eye-tracking and neuroimaging, we have far better ways of approaching what goes on in the translator’s mind. The one thing all the empirical research shows is that the cognitive processes are far more complex than any interplay of seven or even twenty apparently discrete ‘procedures’. As should be clear, I accept that criticism and I am not too worried about it. The solution types are based on products, not processes, and they need have no psychological status at all. But that does not mean that the typologies are random inventions (the basic categories are in the language systems) or that they are useless (the place of their performance is the translation classroom). A more serious criticism, at least in my opinion, comes from Candace Séguinot (1991) when she observes that students are able to learn the solution types but then they do not actually use the concepts when translating or even when justifying their translations. That is, the metalanguage not only fails to inform the cognitive process, but it adds little to translators’ public discourse about their performance. True enough, studies of what learners say while they are translating (thinkaloud protocols) offer little solace to solution types. Research by Englund Dimitrova (2005), Pavlović (2007) and Hui (2012) on translators’ ‘rationalizations’, ‘justifications’ or ‘arguments’ makes scant mention of anything even remotely similar to these categories. Instead, we find things like personal preferences for certain words and phrases, something just ‘sounds better’, the text is associated with images or ‘sounds as if ’ (evoking a narrative scene), a phrase is simply ‘not said that way’, the ‘reader will not understand’ some reference, the author ‘did not mean’ some other reference, and so on (here I summarize the main ‘arguments’ identified in Pavlović’s think-aloud protocols of students translating collectively). Those are the sorts of things that might lead the translating mind to one kind of solution or another. Nothing in the solution types themselves seems certain to get them there. So why bother to teach the solution types at all? The problem could simply be that the terms have been formulated badly: they are unclear and thus not used. However, it is also possible that we are mixing things up here. Let me try three analogies, in the hope that at least one of them might work: 1. When you discovered how to divide numbers in about Grade 4, you probably learned that the ‘dividend’ over the ‘divisor’ gives a ‘quotient’ plus a ‘remainder’. So you know where those things are and how they are



The Proof of the Pudding is in the Classroom

207

related, and you can divide. But never in the rest of your life will you have to name the four things you are using (I had to ask my 9-year-old son, who has just learned the names). Your head will be full of other estimates and calculations; the conceptual space was opened then so that it can work as a semi-automated skill now. 2. When you learn to drive a car, one of the things you acquire is a set of names for the various instruments that extend your body. There might be seven or so: a steering wheel, three pedals (in Europe we use stick-shifts), a gear-stick, two rear-vision mirrors and perhaps a speedometer. You learn where those things are, how they are interrelated and what they are called. But after a few months, you are able to drive without being aware of any of that.1 Your head will be full of relative urgencies, risks of other drivers, directions to go and things like that. In fact, if you had to think about what your feet are doing and the names of the pedals, you would probably have an accident. 3. Closer to home, any formal language learner probably identifies at least seven metalinguistic terms: noun, verb and so on. But when you speak, you are generally not aware of those terms. And even when you have an argument about language, the chances are you will not use them either. Remarkably, when I look at Pavlović’s think-aloud protocols where translators justify their decisions, there is no significant mention of basic grammatical terms, which have been internalized and automatized, along with an awareness of what tools translators have. So if we complain that the terms for translation solutions are not used, should we not also complain about basic grammar? The analogy with mathematics should suggest there is no tragedy in technical terms being forgotten if the skills remain. The comparison with driving merely underscores that automatized actions do not efface the need for formal training. The similitude with grammar is a little more complex: although no infant needs grammar terms to learn a language, adult learners have found them useful for several millennia. So when we train translators, are we working with infants or with adults? If the latter, then why should we not have our basic grammar terms? Perhaps what is at fault is not so much the idea of teaching a list of categories, a toolbox, a space for translators to operate in. Perhaps the problem is with the categories themselves, which are not all well named, delimited and explained. So I am not particularly worried that the solution types are not cognitive processes and I am not put off by the fact that the terms are learned and then

208

Translation Solutions for Many Languages

mostly forgotten. What concerns me more seriously is that the solution types, such as we find them in history and across languages, may themselves be the cause of their own restricted use. And if they are to be profitable anywhere, it must be in the classroom where people are supposed to learn how to translate. What follows is a report on an extended teaching experiment designed to test what works and does not work in the minds of advanced trainee translators.

Vinay and Darbelnet in the Practicum class From 2009 to 2013 I used two hours each year to teach sets of solution types to the final-year Master’s students in my Translation Practicum at the Middlebury Institute of International Studies at Monterey (a detailed analysis of the experiment is in Pym and Torres-Simón 2014). The class sizes ranged from eighteen to twenty-one students, with a mix of languages that included Chinese, Korean, Japanese, Spanish, French, Russian and German, all used in combination with English. The students were not representative of the global community of translator trainees: they were all brilliant. In the first three of those years I presented the seven main Vinay and Darbelnet categories, illustrating them with examples from the textbooks. In the fourth year, I taught Loh’s main categories, giving examples from his textbook. In all years, I first asked the students to translate a set of phrases especially designed to suit the solution types.2 Then I presented the categories. And finally I asked the students to return to their initial translations and attach labels to the solutions they had found. Categories are so much harder when you have not pre-selected the problems! In fifty minutes, the class was conversant with the seven terms. (It is not a hard lesson to teach.) In the second fifty minutes, I asked the students to use the categories to analyse the solutions they had applied in one of their previous translations into their A language (or L1), attaching comments to the text to label the solutions. They were then invited to answer several questions on how easy the categories were to apply: Is it easy to distinguish between the solution types? If not, why not? Which are the ones used the most? How do you think the categories could be improved? A total of fifty-eight students responded to these questions. Of these, more than half (thirty-eight) were Chinese.



The Proof of the Pudding is in the Classroom

209

One general finding was that the students with European languages as L1 preferred the European typology: Vinay and Darbelnet were ‘difficult to understand’ for 75 per cent of the Japanese and Korean students, 63 per cent of the Chinese students, but only 33 per cent of the Romance-language students (the class on Loh had only fourteen students, too few for the corresponding statistics to be meaningful). Not too much should be read into this, however. Those percentages also say that many students shared the same opinions independently of their languages, and there were many individualistic responses, as we shall soon see. And then, the tendencies may not be due to purely categorical virtues, since there are also identity politics involved: we all prefer categories from our own culture, often simply because they are from our culture. Further, some of the students had been taught these or similar categories in the past, although only the students in the Russian programme had been exposed to them at the Monterey Institute. A more significant kind of result was in the students’ answers to the other questions. I was betting that some solution types would systematically be indistinguishable from others and that a few would be regularly misapplied. Here I present a sample of the responses to the three questions, first for Vinay and Darbelnet, then for Loh. The students are coded according to their non-English language (C=Chinese, F=French, G=German, J=Japanese, K=Korean, R=Russian, S=Spanish). In cases where the indicated language is not the student’s L1, the L1 is indicated in parentheses. In some cases the students answered in pairs. Responding to the question of how easy it is to tell the types apart, C18 comments that ‘by the definitions, yes’, but then states that for a certain example ‘I am really hesitant what type it should belong to’. S1 hints at a similar experience when he argues that ‘[i]t depends on each case. Sometimes, there are cases where there is a fine line: you feel you might be using either one or another technique.’ For F2, some of the solutions were easy while others were difficult. Similar hesitations are expressed by others. Among those who found the categorization difficult to apply, there were several comments regarding its inapplicability to Asian languages. K4 says ‘[b]asically, the Korean translator has to reconstruct the sentence structure all the time in both directions’ and therefore considers the Vinay and Darbelnet categories useless. There are also calls for new categorizations: ‘Since Asian languages are more distant from English than some of the European languages, I suggest that we can come up with some categories for Asian languages specifically.’

210

Translation Solutions for Many Languages

These evaluations, and others like them, make it fairly clear why the most discontent came from students working with Asian languages, which should not be surprising. However, a few other complaints are also fairly widespread: a lack of clear dividing lines, and the fact that several types can co-exist within the one solution (as Vinay and Darbelnet do indeed recognize). Asian students reported difficulties with Calque and Borrowing. Modulation and Transposition were difficult to understand for many students, even for those with Romance languages. C2 summarizes the position of several students who feel they understand the techniques but are unable to distinguish between certain pairs: It is easy for me to distinguish between Borrowing, Transposition, Literary Translation [sic] and Reformulation. But I can’t tell which cases are called Calque, Modulation and Adaptation, for I don’t know the differences between Borrowing and Calque, Modulation and Transposition as well as Reformulation and Adaptation. I think they are too similar. In view of this, we take a closer look at the comments on each of the categories in turn, focusing on the most commonly reported problems.

Borrowing and Calque While many students criticize these two solutions as being unclear (C17, C18, C20) or even useless (K1, J1), two Chinese students considered Borrowing and Calque easy to distinguish. For languages with a different script, transliteration is unavoidable and yet it is not among the Vinay and Darbelnet types. Understandably, C4, C5, C6, C14, C24 and R2 all suggested that ‘Transliteration’ be added as a solution.

Literal Translation In Vinay and Darbelnet, Literal Translation basically means ‘word for word’: it applies to cases where words with the same syntactic function appear in the same order, as generally only happens between highly cognate languages. In the students’ minds, however, it is frequently confused with criteria of accuracy or fidelity: the students have been taught to be ‘faithful’, and this is mixed up with being ‘literal’. This understanding of ‘Literal Translation’ is the key to answers like C11/C12/C13’s (who worked together): ‘Literal Translation was the most used, for we regard faithfulness as our top priority when doing the task.’



The Proof of the Pudding is in the Classroom

211

A more instructive understanding comes from C22/C23: ‘Translators usually do Literal Translation while changing the order of words or phrases according to the features of Chinese or English.’ This could mean that they choose Literal Translation plus a change in word order, or that changes in word order are a part of Literal Translation.

Transposition If Literal Translation can play the role of default solution between cognate languages, Transposition might play the same role when moving between English and Asian languages, perhaps along with Modulation. K1 expresses this clearly: ‘Transposition was used the most in almost every sentence due to structure differences between English and Korean.’ There nevertheless seem to be several different understandings of Transposition. J1(L1E) comments that ‘[s]trictly speaking, the solution used for said sentence was a Transposition since it was just about the word order being different’. He thus understands Transposition to be a change in word order, not in word class, whereas in Vinay and Darbelnet it very definitely concerns word class (1958/72: 50). C17 expresses the same doubt: ‘If everything is word-for-word translation except that the order of one word is switched, is it considered as Literal Translation or Transposition?’ These students are reading ‘trans-position’ in its literal sense, as a change in position (what the Russian and German traditions, as well as Nida, called ‘permutation’). As noted, there seems to be no first-order category for ‘word order change’ in Vinay and Darbelnet. So the students’ comments reflect not so much a misunderstanding as a legitimate quandary – they are pointing to a missing category. On a more positive note, C14 proposes a new category called ‘Segmentation Restructuring’ which would cover moving ‘the orders of meaning units around, or chunks of sentences’.

Modulation The limits of Modulation are unclear for most students. C24 confesses that whenever there was a solution she could not classify, she would put in under Modulation. R1 mentions how it is especially difficult to distinguish between Modulation and Correspondence/Reformulation: ‘Sometimes a semantic difference can appear to be a completely different way of expressing things.’ Two lines later, the same student suggests that ‘Transposition and

212

Translation Solutions for Many Languages

Modulation should be merged since the difference between them seems difficult to discern.’ Merging Modulation with either Reformulation or Transposition is also proposed by C20, C22/C23 and G2.

Reformulation and Adaptation The students’ comments show that they had trouble distinguishing between Correspondence/Reformulation and Adaptation. In part this was due to the lesson being badly taught, since Vinay and Darbelnet restrict their équivalence (which we called Reformulation in this case) to corresponding idioms and proverbs. That difficulty notwithstanding, the students recognized that these were extreme solutions, rarely applied in the technical and referential texts with which they normally work. Hence C8: ‘A translator may use more techniques, such as Reformulation and Adaptation, in literary articles with heavy cultural allusions.’ In fact, when it came to identifying solutions in the students’ previous translations, these categories were only really noted by J3 (L1E) who observed fifteen cases of Adaptation because in her start text ‘reference to Japanese characters [fictional people] had no meaning in English’.

How do you think Vinay and Darbelnet could be improved? As can be seen, the students were not short of creative ideas about how the categories could be improved. Some proposals concerned solutions that are considered necessary for all languages; others are for just one or a few languages. Several students commented on how some new categories are needed across the board. R3 asked for Explicitation and Implicitation to be major categories, as did F1, C18 and C24, among others. (All students had attended a previous lesson on these solution types in the parallel course Research on Translation.) Some of the arguments that support this claim are based on language comparisons: R3 commented on the necessity to ‘explicitate’ articles when translating from Russian into English, since Russian does not use articles in the way English does (although this would not count as Explicitation in all definitions of the term). G1 highlighted how adverbials that express time aspects in German can be expressed implicitly in the verb tense in English. Similarly, K14 proposed that Specification and Generalization be upgraded (as indeed they are in Loh). C18 proposed Simplification, C14 talked about Elaboration, C19 wanted Explanation and C16 (L1E) sought Definition.



The Proof of the Pudding is in the Classroom

213

Different language combinations lead to new possible categories. As mentioned above, Transliteration seems unavoidable between languages with different scripts. Languages with different word orders also require a major category for changes in position. K4 and others also recommended Segmentation, that is, breaking up sentences, while C16 (L1E) felt the need for a category called Omission.

Loh in the Practicum class Much the same procedure was followed in the class on Loh. The main comments on the solution types were as follows.

Transliteration, Semantic Translation, Symbolic Translation, Coinage of New Characters Whereas Vinay and Darbelnet had just two categories for foreign names (Borrowing and Calque), Loh gives four, which can be mixed. This is understandable enough, given the move from Latin script to Chinese characters, and it corresponds to some of the complaints expressed in recent years by students with a background in languages that use non-Latin scripts. Perhaps because that demand was met, or because four types cover all bases (as a highest common multiple rather than a lowest common denominator), there were no proposed changes to Loh on this point. F3 nevertheless missed Calque and Borrowing – and since she worked with Latin script, she had no real need of the rest.

Omission/Amplification Loh uses the terms ‘Omission’ and ‘Amplification’ for what Vinay and Darbelnet call ‘Implicitation’ and ‘Explicitation’, and this usage has been followed by later publications in Chinese. In effect, the categories that are lower-order ‘prosodic effects’ in Vinay and Darbelnet here become first-order categories, in keeping with the idea that syntactic and semantic relations are generally (but not always) more implicit in Chinese than in English. The different nomenclature, however, appears to have created some discontent among both the Chinese and non-Chinese students, perhaps because these students had had a lesson on Explicitation in the parallel course:

214

Translation Solutions for Many Languages

C27: I think Omission sounds like losing something. Perhaps we can change it to ‘reduction’. C28: As for Omission, while it can be an omission of a part of speech without any loss of meaning, it can also be possible that the translator chooses to leave out some information which he/she regards as unnecessary. S5: When Amplification refers to explaining information implicit in the text, it should be called explicitation. F3: Amplification should include explicitation strategies, if it does not already. There was also some doubt about whether the types referred to obligatory or optional changes. This remains a problem in most typologies of solutions, including Vinay and Darbelnet.

Repetition Loh’s inclusion of Repetition as a solution type is somewhat anomalous, although it does respond to the important stylistic role that repetition can play in Chinese. The students were not convinced, though. For C34 ‘the scope of “repetition” Loh described is too limited. Repetition strategies differ and vary in different contexts,’ and K6 suspected a Chinese specificity: ‘Does Repetition solution suit for languages other than Chinese?’

Conversion and Inversion In Loh’s system, Conversion is a word-class change (what Vinay and Darbelnet call Transposition) and Inversion refers to word-order changes that are marked in the target language – like Repetition, they are part of the art of writing in a lively way. The Chinese students had little to say about this. The non-Chinese students were not quite so seduced. For K6, ‘Inversion and Conversion should have more specification since the nature of the sentence structure is quite different or even opposite in some ways to name the translation solution.’ S5, on the other hand, thought these solutions would be better seen as parts of a wider category: ‘Inversion could also be called syntactic rearrangement when it deals with reordering different parts of speech within a clause.’ And F3 returned to the problem of what is obligatory or not: ‘I agree with the explanation of Inversion for emphasis, but necessary Inversion is not a strategy.’



The Proof of the Pudding is in the Classroom

215

Negation Loh elevates ‘negation of negative’ to a solution type (‘I gave an unprepared speech’ becomes ‘I improvised a speech’). The students accepted this, although C32 thought that hairs were being split unnecessarily: ‘I think Paraphrase should be added and Negation will be removed and included into the broader category of Paraphrase.’

How do you think Loh could be improved? Some students, all of them Chinese, suggested that a few solution types should be added to Loh. For C27, ‘[w]e can add Paraphrase or Free Translation to refer to translation not bound by the exact words of the source text but conveys the same meaning on syntactic or paragraph level.’ C34 is similar but more succinct: ‘Loh may include another category called “liberal” strategy in translation.’ C29 suggested that ‘[w]e can add breaking long sentences’, and this is indeed a category that was added in later Chinese publications. C33 was looking for something on the other side of the equation: ‘Loh’s existing categories are fine; however, I think he should have added the category of Word-to-Word Translation.’ In Loh’s day, however, word-to-word interference in Chinese was more likely to be from Japanese or Russian than from English, so it was perhaps not on Loh’s mind. Only two students commented on Loh’s adequacy to languages other than Chinese. We have seen K6 asking if Repetition was important for Chinese only. And F3 (L1E) claimed that, ‘for foreign terms, Borrowing needs to be added; Calques should probably be added as well. I do not think that I would ever use Loh’s categories for dealing with foreign terms, given my language combination. For naturalness, I also feel that Modulation should be added.’ That is, she wanted a return to Vinay and Darbelnet and she did not see that Modulation might be at work in Loh’s Negation. Note that the focus on Chinese was questioned not just from a European perspective, but also from Korean.

Solution types as a pedagogical tool Many of the students’ comments have helped to frame the questions asked in previous chapters – one always learns from students – and some of the above suggestions will be used in the following chapter, where I attempt to formulate

216

Translation Solutions for Many Languages

my own typology. The in-class experiment has also been useful as a way of reflecting on why the solution types should be taught at all, and on the various ways they can be of use in the teaching situation. We have seen that, although a typology might be well suited to a particular language pair, students working with other languages are still able to grasp and assess its categories. There is not some kind of cognitive overload, where non-corresponding systems overheat the brain and students explode. Instead, students are able to produce criticisms and suggest improvements, without constraint by some over-reaching cultural allegiance. The way you teach translation need not depend excessively on the languages you are dealing with, not any more than translation itself should be the reproduction of cultural difference. The class experiment also concerns a second false divide, the one set up between theory and practice. When students try to categorize their translation solutions, they reflect both on their work and on the difficulties of theorization. If there is pedagogical value in these activities, it lies in that double process, rather than in the lists of solution types themselves. To be sure, the solution types should have enough relation to actual practice for the theorization to engage; they should cover the range of what translators can be expected to do (and more, in some versions); and a lack of clarity in the names, definitions or examples can lead to more than a few metalinguistic cul-de-sacs. However, even when these criteria are met, the solution types cannot and should not be held up as gospel truths – that should not be their pedagogical function. Learning is a dynamic process, and the prime function of the types is to stimulate and guide that process, as proposals that are also able to evolve dynamically. This can be encouraged in several ways. First, group discussion and comparison of what solutions students have actually used helps novices to discern values specific to the languages they are working with. In these particular classes, European students learned about the implicit and indirect nature of Asian languages, while Asian students found out that not all European languages are the same. Second, students’ comments bring out hidden assumptions about the acceptability of certain solution types. For example, C8 wanted to restrict some solutions to certain genres: ‘a translator may use more techniques, such as Reformulation and Adaptation, in literary articles with heavy cultural allusions’. Text-genre constraints were also important for F2: ‘the [solution types] are all clearly defined, but I think it was also easy in my case, since the texts that I translate did not require any of the more complex solution types, such as



The Proof of the Pudding is in the Classroom

217

Reformulation/Adaptation.’ In this way, the discussion of solution types can lead into related topics for debate and discovery. Third, students reflect on the relation between translating and theorizing. There is a hidden assumption that theory needs to be prescriptive and should show translators ‘best practices’ (F3). Some of the students proposed changes to make the typologies ‘practical and useful for translators [given that] translators don’t want to spend too much time learning those categories yet find them difficult to apply’ (C22/C23). The drive to applicability is an understandable ambition. The important point is that the students had to propose the improvements, rather than accept bad categories or dismiss theorization altogether. That is, the students started theorizing, no matter what their initial rejection of theory might have been.

12

A Typology of Translation Solutions for Many Languages

Chapter summary: Comparisons between previous categories, coupled with classroom testing, allow a new seven-term typology to be proposed. The solution types are assumed to be optional, applicable to many language pairs and can be broken down into an unlimited number of sub-categories to suit specific pedagogical or analytical purposes. It is proposed that the selection of one solution type or another depends, on the one hand, on the relative degrees of effort to be invested by the translator and the text-user, and on the other, the relative risk to the translator’s credibility and to the success of the communication event. In general, the effort invested must be proportional to the risk run. This approach to translation solutions thus cuts across the grand binarisms that traditionally pitted one pole against another, as if there were only two options. Many solutions are always available; they can be applied creatively to countless problems.

Here I gather up various good things found in the above excursions and try to fit them into a tentative typology of translation solutions. Given the confusion created by the many previous efforts in this field, some clarity is required – then again, every theorist has presumed to be clarifying all previous theories, so I am probably succumbing to the same self-delusion. More honestly, I will attempt to be as explicit as possible about what I am trying to do. This is not particularly a question of who wins. If it were that easy, the answers would also be easy: the most beautifully systemic theorist is Torop, I suggest; the most integrative is Schreiber; the most complete might be Chesterman. So why not just stay with them? Basically because the previous typologies tend to flounder in the classroom, and little international consensus has gathered around them anyway.

220

Translation Solutions for Many Languages

A proposed map of solution types After considerable trial-and-error,1 my proposal (Table 12.1) has seven central categories, which turn out to be depressingly traditional. Hopefully the names are just a little more transparent than the traditions on which they have been based. Note, first, that this is a map of translation solutions, not of everything that goes on in a translator’s head. When the translator is moving through the text applying their standard mapping operations, those routines constitute what we might call ‘cruise’ mode, in which no translator needs any special help from a typology: there is usually one clear solution to each problem. But then a problem arises, several viable solutions seem available, there is no rule for choosing between them and the translator thus enters ‘bump’ mode (the analogy is with air turbulence when piloting planes). A choice then has to be made, and that is where some assistance might indeed be found in a pedagogy of solution types. So the following is a map of what can happen in ‘bump’ mode; cruising translators have no particular need of our solutions (or more worryingly, they tend not to be aware of the needs they might have). Table 12.1  A typology of translation solution types for many languages Cruise mode (normal use of language skills, reference resources, parallel texts, intuition – anything prior to bump mode – so no special solutions are needed) Copying

Expression Change

Content Change

Copying Words

Copying Sounds Copying Morphology Copying Script …

Copying Structure

Copying Prosodic Features Copying Fixed Phrases Copying Text Structure …

Perspective Change

Changing Sentence Focus Changing Semantic Focus Changing Voice …

Density Change

Generalization/Specification Explicitation/Implicitation Multiple Translation Resegmentation …

Compensation

New Level of Expression New Place in Text (notes, paratexts) …

Cultural Correspondence

Corresponding Idioms Corresponding Culture-Specific Items …

Text Tailoring

Correction/Censorship/Updating Omission of Content Addition of Content …



A Typology of Translation Solutions for Many Languages

221

As can be seen, the first column has three simple categories: you can copy something that is there, you can change the way it is expressed or you can change what is there. That pretty much covers the ground of Schreiber’s three ‘translation methods’, although I make no assumption that the categories should apply to the whole of a text. The second column has the seven categories that are ripe for learning activities. And the third column is an open-ended list that can be extended to include numerous tricks of the trade, probably without limit. In the remainder of this chapter I will explain the solution types and outline the principles on which they are based. A few examples are given, but since the typology is drawn from tradition, the illustrations are mainly in the preceding chapters. I will then offer a few pointers as to how a translator might decide which solution type is appropriate for a given translation problem.

The proposed solution types The typology is supposed to be pedagogical, and for any teaching purpose you select the degree of specificity appropriate to the people you are working with and why. The reduction to three terms is usually too abstract to stimulate curiosity, and twenty or so quickly become confusing. So I offer explanations and comments in terms of the seven central categories, which is where lessons might be anchored.

1. Copying Words The name ‘Copying Words’ is as clear as I can manage but it is not exact (some languages have ideograms). One could say that the copying is based on the minimal linguistic units that are meaningful when used alone, but even that is inexact and there should be little need to insist. There are three main linguistic levels involved: phonetics (sounds), the meaning-expressing units (morphemes) and the written form (scripts). A translation solution may involve one, two or all of these. The main sub-types are thus as follows: 1.1 Copying Sounds The norm in some translation cultures is to introduce terms by imitating the foreign sounds, applying what Newmark (1988) calls ‘naturalization’. That solution is common enough in translations from European into Asian languages, although it also occurs between European languages: ‘football’ in Spanish is ‘fútbol’, a ‘team’ in Brazilian Portuguese is a ‘time’ and the French

222

Translation Solutions for Many Languages

politician Jacques Chirac becomes ‘Žaks Širaks’ in Latvian. The translation solution can also be extended to text level, as in the French volume Mots d’heures: gosses, rimes (rendering Mother Goose Rhymes) or the quasiphonetic translations of Catullus by the Zukovskys (1969).   When translators copy sounds in this way, their actions may be the first steps in a process by which a new term or phrase enters a language. For example, the sounds of the English name ‘Mary’ were copied into Chinese as ‘玛丽’ (ma li) in translated English novels in the 1950s, but the name has since been assimilated into Chinese culture and is no longer seen as a translation solution (Tian 2013: 67). 1.2 Copying Morphology Copying at the level of morphology would be what happens when ‘football’ becomes ‘balompié’ (literally ‘ball-foot’) in Spanish, or ‘skyscraper’ is a Spanish ‘rascacielos’ (‘scratch-sky’). Examples are easy enough to find in Fedorov (1953), Vinay and Darbelnet (1958) and Loh (1958), which might suggest they are of a bygone era.   Many translation cultures these days prefer to import foreign expressions without modifying the local script. This gives cases like McDonald’s and other brand names, which mostly remain invariable and are presented in English, even as they are pronounced differently in different parts of the world. One might say the same of cases like the English words ‘kindergarten’ (from German) and ‘ombudsman’ (from Swedish), although such straight imports into the globally central language might also be reminiscent of a bygone era. 1.3 Copying Script Of course, the copying of script can be combined with the copying of sounds. McDonald’s in Russian is Макдоналдс, in Arabic it is ‫ماكدونالد‬‎, but both can be found alongside reproduction of the version in the Latin alphabet. For that matter, a different script can have the same phonetic values, as when English is transposed into uʍopǝpısdn English (which exists as an option on Facebook).   When moving between different scripts (e.g. Cyrillic to Latin to Arabic), there are various transcription conventions to be respected. So you respect them. But that is scarcely a translation solution type – it is just obeying rules. Copying becomes a very useful solution in cases of extreme doubt. When translators in twelfth-century Hispania found Arabic medical texts referring to a ‘night plant’ that caused skin infections, they had no idea what the plant was.



A Typology of Translation Solutions for Many Languages

223

They nevertheless rendered it literally as ‘planta noctis’ (Sudhoff 1909: 352). Some centuries later, as European medicine confronted the problem of syphilis, it became clear that the origins of the plant are to be found in the Arabic nabat, a defective transcription of the term banat, meaning girl – the causes of the infections were not plants but ‘women of the night’. The literalist solution may not have been a highly successful translation in its day, but it did succeed in passing knowledge on to later generations. These days, Copying Words is more likely to be a solution limited to the occasional textual item, either because the alternatives are worse (too much work or too high-risk, as we shall see) or because the purpose is only to give a minor textual effect. For example, if I say ‘translation problems are indécidables’, it is not because I can’t put it in English; it is more because the French refers to a theoretical concern that has been elaborated in that language and it might sound cool along the way. Similarly, when a newspaper article says ‘Madrileños are surprisingly unworldly’, the Spanish term for the inhabitants of Madrid operates as an inviting touch of local colour. Note that such textual effects are relatively frequent among non-translations in some text genres but tend not to be used by translators, who are too hung up on having to change the language of everything. The take-away: in cases of doubt, Copying Words is low-effort, might work in the long run and can add colour and panache.

2. Copying Structure ‘Structure’ here refers to the relations between the expressions in a text, mostly with respect to order of presentation. Alternative terms could be ‘calque’ (as in Vinay and Darbelnet), ‘word order’, ‘syntax’ or ‘composition’, but I am looking for something to cover all of that (in much the same way as the young Fedorov decided to use ‘syntax’ to refer to element order on all levels). Since this solution type sails close to word-for-word literalism, the point to stress is that the solutions here are not obligatory: if the syntax in the start text can easily be reproduced in the target language, then there is no special solution type involved (since there is no substantial problem to be solved). We can nevertheless talk about a solution type when an unusual syntactic pattern is brought into the target language. This is the case of what Vinay and Darbelnet called Calque, as in the English expression ‘Governor General’, which has English words but French syntax. Element order can also be copied in translations that are designed to explain what is going on in the foreign language, as I have been

224

Translation Solutions for Many Languages

doing to explain examples in this book (in some cases I have been copying morphological structure as well as syntax) or as might be found in Hölderlin’s Sophocles or Felberbaum’s Herodotus. An extension of this could be the rightto-left reading of frames that is used when manga are translated from Japanese into English. And a more playful example might be the antipodean television series Spartacus (from 2010), where characters use Latin syntax in English, speaking pseudotranslations of the style, ‘Open mouth and lend voice to tongue’ (instead of ‘Tell me what you have to say’). The main sub-types here would be: 2.1 Copying Prosodic Features (rhyme, rhythm, alliteration) Example: ‘The Price is Right’ rendered as ‘Der Preis ist heiß’ [The price is hot]. 2.2 Copying Fixed Phrases (syntax, idioms, common expressions) Example: ‘Like a dog in a game of skittles’ (as a literal rendition of the French ‘Comme un chien dans un jeu de quilles’) (cf. Vinay and Darbelnet 1958/72: 55). 2.3 Copying Text Structure (paragraph and major units of text organization) Example: Translations of the acquis communautaire (the body of accumulated EU law) are required to obey the same sentence divisions of the French or English texts, even when languages like Finnish or Turkish would normally require different divisions. Despite its relatively infrequent use, Copying Structure is a major way in which fixed expressions and text formats (especially literary forms) travel from culture to culture.

3. Perspective Change Perspective Change is another name for classical Modulation, in its simplest and widest possible sense. The basic message is that you can see the same object from a different angle, and this has proved extremely valuable for students who are blocked by literalism. The tricks to be included here are the classical transformations of active into passive, positive into negated negative, half-full into half-empty, verbal structures into nominal structures, a directionality being reversed and indeed all those things being reversed. The use of a particular script or marked register might also function as a Perspective Change in some situations. The main sub-types would be:



A Typology of Translation Solutions for Many Languages

225

3.1 Changing Sentence Focus (all the non-obligatory sentence-level syntactic transformations) Examples: ‘His failure to feel excitement about …’ rendered as ‘Er war gar nicht scharf darauf …’ [‘He was not at all keen on …’] (Neubert 1991: 37). Or ‘I see Mt. Fuji’ rendered as ‘富士が見える’ [Fuji is visible] (Kanaya 2004). Or ‘Yesterday I gave an unprepared speech before a big audience’ as ‘昨天我在广大听众之前即席致词’ [yesterday I big audience before improvise speech] (Loh 1958). 3.2 Changing Semantic Focus (change due to the choice of different values) Example: ‘No vacancies’ rendered as ‘Complet’ [Full] (Vinay and Darbelnet 1958/72: 55) 3.3 Changing Voice (changes in register and sometimes in pronoun structure) Examples: Publicity that uses the formal second person in French and German will often use the informal second person in Spain but the formal second person in Latin America. Or again, phrases like ‘nuestra economía’ (our economy) regularly lose the first-person reference when translated, perhaps becoming ‘the Spanish economy’ (this might also count as Specification). More tellingly, the Tokyo-based translation company Gengo currently offers an online service where it asks the customer, among much else, ‘Is it more “Hey dude” or “Dear Sir”?’ (https://gengo.com/order/optional). It seems clear that translation between English and Japanese or Korean, for example, involves quite massive Perspective Change and that many of those changes will be automatized to some extent. However, whenever there is a difficult conscious choice to be made, Perspective Change should count as a solution type. That claim might find some backing in theory. The linguist Michael Halliday (1985) argues that there are ‘grammatical metaphors’, which in effect means that the distribution of word classes can be marked as the result of a clear choice: ‘meanings may be cross-coded, phenomena represented by categories other than those that evolved to represent them’ (Halliday 1985: xviii). If pushed, this would mean that all changes in grammar could bring about changes in perspective. To return to a classical Transposition from Vinay and Darbelnet: ST: As timber becomes more valuable TT: Depuis la revalorisation du bois [Since the increase in the value of timber]

226

Translation Solutions for Many Languages

Here the verb phrase ‘becomes more valuable’ has become a noun phrase ‘la revalorisation’, which is a change in grammar. But the change also alters the focus of the entire sentence: the subject in English is ‘timber’; in French it is ‘the increase in value’ – the same thing is seen from a different perspective. The distinction between Transposition and Modulation then becomes theoretically untenable (and this can be shown on the basis of an example from the main proponents of that distinction). I thus offer just one solution type: Perspective Change.

4. Density Change To pursue the visual metaphor, there is a kind of change where the angles are not altered but the proximity of the object is. The examples that come to mind are changes in degrees of generalization and specification, implicitation and explicitation, which express the same thing by using different grammatical resources but from the same point of view. The translator is zooming in to show greater detail, or zooming out so as to grasp the basic outline – Kussmaul (2000) actually uses the metaphors ‘zooming’ and ‘focusing’ as names for creative translation techniques. Andrew Chesterman suggests that the variable should be the degree of ‘granularity’, which makes an analogy with photography.2 If we are going to take the metaphors seriously, these should form a distinct type of translation solution. The notion of ‘density’ here minimally concerns how much text is used in order to convey a given set of information. It has played a very minor role in the history of solution types, although it has been a major category not only for Kussmaul but also for Malone (1988) and Klaudy (2010; cf. Klaudy and Károly 2005), both of whom start by asking how much text is added and how much is taken away. As soon as you look at the various solutions in terms of Density Change, quite a few things fall into place. Generalization and Specification, of course, can be seen as changes in how much text is used to express a set of information, and this change in density is not necessarily involved in anything I have placed under Perspective Change. Associated with these are the categories of Explicitation and Implicitation, which similarly affect density but are restricted by what can be considered implicit information. But there is more. When you use what Newmark called a Translation Couplet (e.g. ‘Gemeinde, German unit of local government’), you are copying an item and giving an explanation, reducing density by using more text. You could always argue, of course, that the giving of two translations is indicating a change in perspective



A Typology of Translation Solutions for Many Languages

227

(this is how they said it and here is how we understand it), so this could be Perspective Change. There is nothing wrong with two solution types occurring at the same time. In this case I would suggest, though, that the opacity of the German term Gemeinde is not saying a lot about any cultural point of view, and that the most useful lesson to be learned here is that the translator can indeed use six words just to translate one – density is the prime thing being changed here. The same can be said, by extension, for cases of double translation (rubi characters in Japanese, for example), in-text expansion or contraction of any kind. All these things can be lumped together as Multiple Translation, if you like, meaning that you give more than one solution to the problem, probably involving the use of more than one solution type. Perhaps more contentiously, I would also like to place the various modes of Resegmentation under this rubric. When you cut a complex sentence into two or three shorter, simpler sentences, or when you join sentences together, you are necessarily working on degrees of syntactic density. Independently of how many words and punctuation marks you might actually have to add, several short sentences are cognitively less dense than one long sentence, and that difference should justify placing those transformations here. (Then again, if this justification is a hard sell, Resegmentation could easily be recognized as a solution type in itself.) The main sub-types might thus be: 4.1 Generalization/Specification (changing the degree of specificity) Example: ‘Canada’s acclaimed speech recognition system’ might become ‘the internationally acclaimed speech recognition system’ (i.e. ‘Canada’ is generalized as ‘international’) in situations where the system is for languages other than the official ones in Canada. That is, the generalization removes possible misunderstandings. (Many other cases of Generalization and Specification are difficult to distinguish from Implication and Explicitation, to the point where the distinctions are pointless.) 4.2 Explicitation/Implicitation (showing or hiding implicit information) Example: The explicit naming of ‘legs’ and ‘paralyzed’ in ‘The long hours of unrelieved kneeling had so paralyzed his legs that he could pick himself up only with a special effort’, as a rendition of ‘長い間端座の形を崩さずにいたので、努力しなければ立ちあがれ なかった。’ [‘For a long interval, had not changed his kneeling position, so without straining would not have been able to stand up.’] (Seidensticker

228

Translation Solutions for Many Languages

and Anzai 1983: 54; Scheinberg 2014: 50). Implicitation would then be what happens when you go the other way. 4.3 Multiple Translation This means visibly applying more than one translation solution, as in: ‘Gemeinde (German unit of local government)’ (Newmark 1981: 31). 4.4 Resegmentation (changing sentence or clause breaks) Example: The current Spanish passport has one sentence in Spanish that is accompanied by an English translation that has three sentences. A literal version of the Spanish is as follows: ‘The Spanish Government retains the sole property rights to this document without prejudice to rights of the holder and therefore recommends that the holder take the utmost care regarding the custody and use of the passport and requests that whatsoever authority or other person return the passport to the Spanish authorities should it be lost or used in an unjustified way [breathe now].’ (My thanks to Kevin Costello for the example, and to the anonymous translator into English who had the courage to introduce a few short sentences into Spanish bureaucracy.) Density Change is used all the time in modes of translation that have generally not been considered by the theorists of translation solutions: subtitling (obviously), conference interpreting (both simultaneous and consecutive) and pre-editing in localization workflows, for instance. Density can moreover become a dominant criterion in situations where translations address different social levels, often for purposes of dissemination, teaching or the introduction of new concepts into a language.3 To take a banal example, most bureaucratic prose can be simplified, as has been sought by various campaigns in English: ST: The public service is carrying out ongoing adjustments to its tax review policy. ST: We are changing the way we review your tax payments. That sort of reduction in density (here with a little Explicitation thrown in), laudable in English, becomes seriously advisable when translating into languages that have no corresponding registers, or for users like recent immigrants who require accessible language. Density is not a banal consideration.



A Typology of Translation Solutions for Many Languages

229

5. Compensation Understood as placing the solution in a new textual location and/or at a new linguistic level, Compensation has long been present in the main typologies, albeit mostly playing a minor role. You could argue that it is not properly a solution type in itself, since it merely concerns the place or level where other types are used. Or again, since it is often used in order to decongest the location of the initial problem, one might be tempted to place it under Density Change, especially since its occurrences are relatively rare. However, Compensation has a real value as the topic of a lesson, as an eye-opener for novices who, focusing only on the segment at hand, tend not to consider that a value can be expressed in other ways and/or in other parts of the text. This is especially true in the age of translation memories and post-editing, where our technologies keep our eyes focused on just one segment at a time. The message of Compensation, and more generally of solution types as activities that can humanize the technologies, is that human translators can and should think at the level of the text and beyond: here Compensation includes thinking about what can be done in footnotes, translators’ prefaces, front covers and the like. The use of such resources can, of course, affect textual density and might in such cases easily be listed under Density Change as well as Compensation. There is no reason why our categories should pretend to be watertight on this point. The main sub-types are distinguished on the basis of where the new resource is placed (these two sub-types can obviously be mixed): 5.1 New Level of Expression Example: ‘On se tutoie …’ [We can use the intimate second-person pronoun … ] rendered as ‘My friends call me Bill …’ (Vinay and Darbelnet 1958/72: 190). 5.2 New Place in Text Example: The use of footnotes, endnotes, other paratexts (including images, covers and promotional material). According to this way of cutting the cake, a translator’s footnote could be Explicitation + Compensation (if it is elaborating information that is implicit in the start text) or Addition + Compensation (if it is adding new information).

230

Translation Solutions for Many Languages

6. Cultural Correspondence As indicated in feedback from students, it makes some sense to bring together the use of corresponding idioms (Vinay and Darbelnet’s équivalence) and functionally corresponding referents (Adaptation of the kind that replaces ‘cricket’ with ‘baseball’). There is no need, however, to assume that the items thus paired occupy exactly the same position in two huge systems called ‘cultures’. Two items might share some functional or semantic values but be wildly different on many other levels. ‘Correspondence’ is to be understood in that loose sense. For example, consider the following alternatives (the point is easy enough to illustrate in English): Before you could say Jack Robinson. In two shakes of a lamb’s tail. On the double. In a flash. In a jiffy. Lickety-split. Very quickly. Such variants exist for a reason. As Vinay and Darbelnet fantasize, when the translator hits upon a happily corresponding idiom, the reader will exclaim, ‘Yes, that’s exactly what you would say!’ (1958/95: 37). But what also happens is that the idiom, with its regional and social specificity, defines the ‘you’ in question: ‘two shakes of a lamb’s tail’ is what my mother would say (I thought it was Australian, but it seems to come from Yorkshire); ‘on the double’ connotes a military ‘you’; ‘lickety-split’ is informal, fun and apparently American; ‘Jack Robinson’ is a mythical character who circulates as culture-specific knowledge (he seems to have been British in origin) – some knowledge of him is necessary in order to understand quips such as Joan Rivers’ ‘She’ll be back in the closet quicker than you can say Jodie Foster’ (the humour stems precisely from the cultural restriction of both the names and the structure). The use of one or the other expression conveys information about who is saying this and who is being addressed, on levels that are usually much more specific than whole cultures or language systems. The pity is that translators tend to opt for the easiest alternatives, equivalents of the unmarked ‘very quickly’, creating texts that eschew perspective, becoming flat and unengaging. The same argument can be applied to the kinds of things that Vinay and Darbelnet called Adaptation. When a translator has to decide whether to



A Typology of Translation Solutions for Many Languages

231

replace one summer sport with another, the decision depends very much on which cultural perspective is to be adopted: ‘baseball’ may be ‘our’ sport, whereas ‘cricket’ is ‘theirs’. There is no neutral correspondence; these are the ways cultural scenes are set and peopled with specific identities. The main sub-types would be: 6.1 Corresponding Idioms Example: ‘Give him an inch and he will take a mile’ rendered as ‘得寸进尺’ [get inch, take foot] (Loh 1958: 2.107). 6.2 Corresponding Culture-Specific Items Examples could be units of measurement, currency, forms of address, etc.

7. Text Tailoring Whereas all the above solutions allow the translator to try to say what is said in the text, Text Tailoring (or Content Change)4 recognizes that there are situations where translators legitimately alter what is actually said in the text. This may involve simple corrections of mistakes, the deletion of significant stretches of material that is not relevant to the purpose of the new text (thus going beyond anything that can be justified as Implicitation), and the addition of new material that may enhance that purpose, as in the case of extensive explanations of historically dated or otherwise little-known terms and customs (thus going beyond Explicitation). I am clearly adopting this category from the work of Michael Schreiber and the messages of Skopos theory, and I am very aware that it will be criticized by those who believe translators have no mandate to make such changes to texts. Without entering into futile polemics, I offer two considerations. First, there are many situations where the translator should clearly not attempt to tailor texts, for example in most legal environments. Second, now that machine translation has reached the stage where it is frequently more cost-effective to have translators post-edit rather than translate from scratch, the added value of our work must increasingly be found at higher levels of cross-cultural communication, including text selection and alteration. The term ‘translation’ is not anchored to some eternal essence where such things are unthinkable. The main sub-types here, more or less as in Schreiber (Table 8.1), are: 7.1 Correction/Censorship/Updating I was once asked to translate the following sentence for a children’s encyclopaedia: ‘El hombre blanco ha marcado el paso del progreso humano durante

232

Translation Solutions for Many Languages

los últimos siglos’ [‘The white man has marked the pace of human progress for the last few centuries’]. My refusal to translate this could be called correction, censorship or an updating of ideologies. 7.2 Omission of Content Whole paragraphs or chapters may be omitted where they are not pertinent to the translation purpose. You might ask yourself why translations of the One Thousand and One Nights almost never have exactly 1001 stories. 7.3 Addition of Content Translators may participate in the addition of notes, glossaries, commentaries, interviews with the author, illustrations and so on. In Spivak’s translation of Mahasweta Devi’s Breast Stories (1997), only 75 of the 176 pages are actually authored by Devi, the rest being Spivak’s introduction and essays on the author.

Principles for a typology The proposed typology has several features that have gained in currency as the history of solution types has progressed: It concerns translators’ transformations of text: It does not deal with things like finding information, correcting mistakes, using machine translation, revising what you have written, taking a break when you are tired, drinking lots of coffee, asking the client, and anything else that might help you create a good translation but is not actually producing translation solutions. And it concerns the kinds of transformations that tend to be made by translators and interpreters more than by anyone else – it is not just about using the target language in a laudable way. It concerns situations where the translator has a choice to make: It does not deal with the application of absolute grammatical rules; it does not involve the outright ‘servitude’ that Vinay and Darbelnet included within their purview; here there are only ‘options’. In other words, I am interested in how mediators solve problems, not in how they obey rules. (If the translation can be completed with grammar and dictionary, then there is no real problem to solve anyway.) It concerns more than one language pair: The typology is not derived from any comparative stylistics in the traditional sense, although it should certainly be able to provide conceptual tools that might be of use when comparing the tendencies of specific languages.



A Typology of Translation Solutions for Many Languages

233

It need not classify complete linguistic data: Since the one expression can solve several problems simultaneously, we must allow that the same segment of a translation can involve several solution types. Its purpose is pedagogical: The typology does not purport to describe the actual cognitive processes of translating translators. It should be judged successful only when trainee translators and interpreters are able to grasp the terms and use them to extend their initial conceptions of the translator’s task. Learners should thus enlarge the range of solutions they feel able to select from. It is mostly derived bottom-up: The tradition I have been trying to piece together has been based on close attention to actual translations, to the things that translators have done. This means that the divisions between categories have not been envisaged in terms of abstract binary polarities (perhaps from literal to free, or foreignizing to domesticating). In practice, most of the dividing criteria tend to come from linguistics, which should not be surprising if we consider that a natural language is first and foremost a set of categories that divide up the world, and linguistics is partly a set of techniques for identifying those divisions. This could account for the predominance of non-continuous values (actual divisions are assumed, rather than clines, blends or fuzzy things). It should also explain the open-ended lists at the lower levels and the relative absence of large-scale notions of the way translation establishes relations between cultures. Although fairly loose, these features provide some clarity of purpose. They also leave many issues quite open. Note that there is no formal definition of how long a ‘piece of text’ has to be; there is no assumption of the translator’s competence or professionalism (I think translation should be taught to all adult language learners); there is no attempt to gauge the ways in which social norms structure and constrain the alternatives available; the typology does not say at what particular level the learning process is supposed to be operating (but let’s hope the students know enough about the languages to be interested in translation as more than a check on language acquisition); it does not say which solutions might be condoned or enhanced by translation memories and machine translation (although I will return to this); it does not assume that translation is based on some kind of perfect understanding of the start text (I would more generally claim that only ‘understandings’ are possible). All those things belong to the situations in which the typology might be applied. This absence of abstract principles need not be a drawback. Indeed, it may allow students to use the typology as a tool for discovery. For example, if they compare fully human translations with their post-editing of machine

234

Translation Solutions for Many Languages

translations, they are likely to find that the machine-translation output avoids solutions at both the top and the bottom of the grid: Copying tends to be what happens when the algorithms give up, rather than a stylistic effect, and things like Compensation and Cultural Correspondence tend not to be frequent enough to surface through the statistics (although there are always some entertaining surprises). To play the full range of our instrument could mean fighting against the machine. Another learning activity is for students to analyse one of their previous translations and to map their own solutions onto the grid, then compare their preferences with others. They should soon become aware that they have a certain translation style, and that their style could always be altered, if they so desire. The typology can also be made to speak to wider intercultural relations in several interesting ways. Traditionally, this is done in terms of polarities, with the start culture at one end and the target at the other, and each solution type is then positioned somewhere along the line between the two (as for example in Hervey and Higgins 1992: 33, reproduced in Figure 9.1 above). So each solution would in some way represent or affect the perceived distance of the foreign language and culture. Those simple abstractions not only invite a collapse into binarism but also seem particularly unable to position solutions such as Omission, Addition or Density Change, or indeed the more technical pirouettes of Perspective Change. One might nevertheless imagine two Cartesian axes, one for the perceived location of the item in the eyes of the receiver (from start to target, foreignizing to domesticating, exotic to transplanting, or whatever) and the other for the perceived amount of accessible information given on the foreign language and culture (from less to more information) (cf. Veselica Majhut 2012: 96–100). This second axis would account for the active pedagogical function of Density Change, for instance, where changes in accessibility or opacity become ways of intervening in the relation with the other, rather than merely representing the other’s location. A very tentative distribution of the main solution types across the two axes is shown in Figure 12.1. Such attempts allow for considerable speculation and a few interesting questions. An engaging activity in the Translation Theory class might be to generate actual solutions for a few problems and then try to position them in relation to the axes – much depends on the values in play in each case. More generally, you might ask which sectors of this space of actions best allow one culture to invade another, or for a culture to defend itself from another, or for



A Typology of Translation Solutions for Many Languages

235

+Start location [Stereotyping?]

[Some multiple solutions?] copying

– Information

omission

less density

greater density

addition

+ Information

perspective change correction updating cultural correspondence [Re-creation?]

[Meta-text?] +Target location

Figure 12.1  Tentative positioning of main solution types in terms of accessible information on start culture (horizontal axis) and perceived location of item with respect to start culture (vertical axis)

selective adaptation. Would all kinds of Perspective Change necessarily be in the same quadrant? (Or are the virtues of Perspective Change merely pedagogical, without a fixed place along these axes?) Where would Compensation go? (As a secondary form, it seems to play no fixed role here.) And then, for creative theorists, why do the main solution types seem to leave the top-left and bottom-right quarters relatively empty? What happens there? Part of the answer might be in the way Holmes (1970/2005) positioned poetry translation as the overlap of ‘critical essay on poem’ and ‘poem inspired by poem’ (cf. Bakker and Naaijkens 1991), which suggests one way of filling in the bottom corners: a ‘text inspired by the start text’ (‘re-creation’ for short) would give minimal information on the foreign culture and be positioned in the target culture, without wholly being a translation, while a ‘text about the foreign text’ (‘metatext’) would give added information while also remaining in the target culture. As Holmes suggested, the space of translation starts in the overlap of these two forms. As for what happens in the corresponding corners at the top, I can only conjecture that a highly reductive, stereotyping discourse on the foreign culture would be one low-information form, without being wholly translational, and that combined translation solutions such as Copying plus Addition might fulfil

236

Translation Solutions for Many Languages

the other logical possibility of maximum information with perceived location in the start culture. At the end of the day, our bottom-up typology could suggest that the Western translation form is best suited to only certain areas in this space of possible actions.

How to select solutions As it stands, the typology cannot really say what each solution type is good for, or when and how it should be used. This is partly because the types themselves do not take account of the specific cultural situation in which the translation is to operate – they have no sociology – so we can only guess as to the range of possible effects. At the same time, the categories have no psychological status either; indeed, they overly simplify complex cognitive processes. On this second shortcoming, one might nevertheless justify the reductive categories with the following narrative: When the translator is confronted with several different possible renditions, when a dozen minor hypotheses and scenarios are running through your mind, the reason that counts is the last one, the one that justifies your final selection. That last idea is the solution type that becomes ‘dominant’, and it need not account for all the intricacies of language, including the many ways in which form creates value. It certainly does not exclude any of the other ways of producing solutions; it just trumps them, enacting a decision. On the basis of that scene, I formulate three further principles that concern how the dominant solution might be reached rationally: In principle, all solution types can be applied to all problems. I steer clear of any suggestion that ‘when confronted by problem A, use solution type X’. If you work very hard and are stupid enough to risk losing your audience and clients (followed by your spouse and your dog), then you could randomly apply any solution type to any problem at all, in theory. From this it follows that, in principle, all the solution types can be used in all possible combinations. The selection of a solution type partly depends on degrees of translator effort: This follows from the above: although all solution types are available, it is rational to opt for the one or two that require the least effort for the desired effect. Further, since it is conceivable to calculate an abstract number of transformations involved in each solution type for each problem, it is possible to present the types in terms of degrees of abstract effort, in much the same way as Vinay and Darbelnet arranged their solutions in terms of increasing presumed difficulty. My list thus goes from minor to major transformations, with abstract effort increasing as



A Typology of Translation Solutions for Many Languages

237

you move down. This need not correspond to actual effort in the translation process, of course. Experienced translators use automatized or semi-automatized solutions that require minimal effort, partly because their ‘translator effort’ has been expended in their training or early experience. Further, if a translator opts for a very low-effort solution like Copying a foreign word (assuming not too much effort was put into selecting this solution type), the ‘translator effort’ might be passed on to the receivers who then have to make sense of the foreign word. The degree of translator effort should be proportional to the degree of credibility risk: For a given problem in a given situation, each possible solution will involve a risk of not contributing to the success of the translation.5 When that risk is high, then high translator effort is warranted. When it is low, then there is no reason to work very hard – use an easy solution. The selection of an expected solution type reduces the receiver’s effort: We know that different cultures are used to different kinds of translation (they have different ‘translation norms’) and this also plays a role. If receivers expect a certain kind of solution, then their processing effort will be less and the risk of losing their trust will also be less (cf. the role of ‘expectancy norms’ in Chesterman 1997). On some occasions, though, translators might want their receivers to be surprised, or to be forced to think, with all the added effort and heightened risks involved. That is, translators can choose to adhere to, challenge or break the prevailing norms (Toury 1992). To quickly illustrate how effort and risk work together, consider Omission and Copying, which can theoretically always be used to solve any problem at all, since they are both often low-effort solutions. There are nevertheless many problems for which you would not want to use them too much, since they incur the risk of your text looking incomplete, opaque or otherwise not helping to establish trust. And the first consequence of you not being trusted is that you might not be asked to translate again. As an illustration, let me take the title from Marlene Van Niekerk’s Afrikaans novel Agaat (2004). (The names of books, films and television shows are susceptible to numerous extreme translations and for this reason make good examples to use in class.) In this case, the title could conceivably be rendered as follows, among much else: Copying: Agaat, A-g-g-g-g-g-gaa Expression Change: Agate, Agatha, The Black Maidservant/Nanny (illustration of young black girl on front cover) Content Change: Ash Bottom, The Agony of Apartheid, A New Matriarchy

238

Translation Solutions for Many Languages

Straight copying (‘Agaat’) has the disadvantage of being opaque for many readers: it is not immediately obvious that ‘Agaat’ is the name of a person. And how should the letters be pronounced? There is a real risk of alienating readerships. The main translator into English, Michiel Heyns, nevertheless notes a certain strategic use of this opacity: ‘Marlene [the author] felt strongly and I agreed, that the sound of the name is such an important part of the meaning that one did not want to lose it, even where its presence could only signal, to a foreign audience, that this was indeed a foreign sound’ (2009: 124). In the text of the English novel we thus find the phonetic copying ‘A-g-g-g-g-g-gaa’, which runs the risk not just of being pronounced incorrectly (the guttural has no letter in English) but perhaps also of creating gratuitous alienation. An Expression Change could then be justified in order to reduce precisely that opacity, minimizing the risk. You could Anglicize the name, or make it the name of a stone (which is also done in the text), or insist it is the Dutch form of ‘Agatha’, derived from the Greek agathos, meaning ‘good’, and the woman named Agaat is indeed called Good at the end of the novel. You could also play with Perspective Change: in the novel, Agaat is the maidservant of the mother and the nanny of the son, and these twin points of view are played upon. In practice, though, the published translations merely copy the name on the front cover and then change both Perspective and Density (the amount of information in a stretch of text) by adding an image of a black girl or woman (at least in the translations into English, Swedish, Dutch and French). That iconic addition should be seen as Compensation, since it offsets the risk of the opaque name. As for Tailoring the name of the novel (not the novel as a whole), anything could happen – just imagine a Hollywood version and invent your own a box-office title (the film rights have been sold; I await something catchy and uplifting). The British edition of the book changed the title to ‘The Way of the Women’, conceivably appealing to a feminist readership. Yet much more is possible. To stay just with the name, the novel notes that the character Agaat was first called ‘Asgat’, literally ‘ash arse’, ‘because she sits with her arse in the ash in the fireplace all the time’ (2004: 569). The central sound of that name is also semanticized in the novel, since ‘G-g-g-g-g-g-g-g-g’ represents the sounds the young girl first makes and perhaps also the death rattle of Apartheid as a way of life (‘agony’ indeed). The novel itself explores so many different readings of the deceptively simple name that in this case it would be more appropriate to talk about the translator selecting content rather than changing it. In principle, all solutions are possible. But some involve less risk and less effort, for both translator and reader. And some risks are worth the effort.



A Typology of Translation Solutions for Many Languages

239

For an example where all seven basic solutions can be envisaged, I turn to the point in Agaat where the narrator is wondering how to translate some Afrikaans words: Vertalings vir wolfneusgewel, rûens, droëland, drif. Dink dit uit. Altyd ‘n gelag by die werk. (Van Niekerk 2004: 8) [Translations for wolfneusgewel, rûens, droëland, drif. Think it out. Always a laugh at work.] So what should the translation into English be? Here are suggestions for the seven solution types: 1. Copying Words: Yes, leave the four problematic Afrikaans words in Afrikaans and let the readers sort them out (if they know some German, they will get halfway there). Or create morphological English versions of the Afrikaans words. 2. Copying Structure: Render the whole passage in the precise order of grammatical elements, so it sounds like the interior monologue of an Afrikaner. 3. Perspective Change: Since the narrator is bilingual, have him ponder how to translate four corresponding English words into Afrikaans: ‘jerkin-raised gables, ridges, dry farming-land, crossing’. 4. Density Change: For the four problematic words, do all of the above, one after the other, and add an explanation of what the words mean. Or generalize the four words as ‘Afrikaans for things in this land’. 5. Compensation: Place the explanation in a translator’s note. 6. Cultural Correspondence: Try something like ‘wrought-iron verandahs, gullies, stations, the wet’, at the risk of making the character an honorary Australian (and at the extreme risk of no one wanting the privilege). 7. Text Tailoring: Delete the passage, since it involves too much work for too little gain; or give further additional information on the nature of the translation problem. Here is what we find in the version by Michiel Heyns and Marlene van Niekerk: Translations for wolfneusgewels, rûens, droëland, drif: jerkin-raised gables, ridges, dry farming-land, crossing. Prosaic. Devise something: wolfnosed gables, humpbacked hills, dryland, drift. Always the laughter at the office …. (Van Niekerk 2006: 6)

240

Translation Solutions for Many Languages

This is actually a cocktail of solutions: Copying Words, straight dictionary translation, morphologically translated words, Perspective Change (the start text makes no comment on the ‘prosaic’ nature of the dictionary translations), Density Change (fifteen words have become twenty-seven) and Text Tailoring (the difference between the translations has been added, as has the comment on their nature).6 The effort is substantial; the risk is minimal (the translator had the agreement of the author); the gain in cross-cultural understanding is considerable, and fun. No list of solution types will tell you which one to use in all cases. That decision can only be made when other factors are taken into account, probably in the split second in which translators select one of the alternatives. The prime factors, I have proposed, are translator effort and the risk of communicative failure. Those general principles are easy enough to teach and they are more important than having anyone learn by rote a list of fixed categories. Yet the list has its virtues. At one end of the learning process, it can alert novices to the range of what translators can do. And at the other end, when students are asked to comment on the solutions they chose in a translation, this kind of grid gives them an initial metalanguage and obliges them to reflect on the reasons why they have necessarily excluded many alternatives.

Responses to a few extreme objections Not everyone will agree with my typology, of course, and that is fine – this is an ongoing, imperfect exploration. To close, let me nevertheless indicate how the logics of effort and risk might respond to a few hypothetical objections, admittedly of the extreme variety and certainly not limited to questions of literary translation. For example, why not reproduce, in all cases, the syntactic order of the foreign text? That would generally be low-effort for the translator (a machine could do it for us), but then it would probably be high-effort for the person receiving the text. In fact, the effort required in reception would in most cases be so high as to exceed the benefits of the text itself – the reception process will stop and communicative failure will be the result. That is why Copying Structure is restricted to situations where it is worth investing high effort in the reception: for the understanding of foreign syntax (as in the examples in this book), for an aesthetics of estrangement (as in literary experiments) and in situations where



A Typology of Translation Solutions for Many Languages

241

the start text is not well understood in the first place but is considered to be of great importance (as in the translations of the physician Galen from Arabic to Latin in the twelfth century, for example, when Europe was rediscovering medicine). Why not just go around Copying Words all the time? It would once again be low-effort for the translator, high-effort for the receiver and probably high-risk for the receivers who suffer conceptual overload. Copying Words is thus a solution usually limited to just a few problems per text. Why not reproduce foreign idioms exactly as they are? Consider the following classical pair (Vinay and Darbelnet 1958/72: 55): ST: Comme un chien dans un jeu de quilles [Like a dog in a game of skittles] TT: Like a bull in a china shop When translating from French we could indeed say, ‘Like a dog in a game of skittles’, why not? That would be novel, understandable, possibly entertaining and it might tell us a little about French culture (surely it should be a game of pétanque, though?). It would be relatively low-effort for the translator, but then once again high-effort in reception and high-risk in situations where the receiver might not figure it out. If you opt for ‘Like a bull in a china shop’, the effort for the translator is a little higher, but the risk of communication failure becomes a little lower. On the other hand, to render it as ‘in a brusque and violent manner’ would be relatively low-effort and low-risk all round, resulting in yet another boring translation, which should constitute a general social risk in itself. As soon as translations are associated with tedium, the translator might be trusted but the translations will fail to communicate, incurring the grave danger that cultures will cease to interact with the foreign. The lesson: use idioms, use the full resources of your languages, if and when the effort can be justified. And why not use Perspective Change all the time? When the French novelist Michel Tournier rewrote Defoe’s Robinson Crusoe (1719), he gave the narrative voice to the native Friday (Vendredi, ou les Limbes du Pacifique, 1967), which is certainly a text-length change in perspective. Although it is intellectually instructive to view the French novel as a translation, that kind of solution is certainly not common among translations. Why? Basically because it would require too much effort from the translator, probably incur a high-risk relation with any publisher who commissioned the translation and potentially mislead any reader who bought a book with ‘Daniel Defoe’ named as the author. There is no need to invent essentialist definitions of translation here; it is enough to

242

Translation Solutions for Many Languages

point out the efforts and risks involved. If you are going to rewrite a whole text, make sure your efforts are recognized and paid for as such – become something more like an author. Finally, why not just be ‘close to the text’? This is what many of my students say when I first present the typology and I try to convince them of its virtues: ‘The translator has to be literal/accurate/faithful/close to the text.’ So apparently there is no reason at all to copy words (that is for terminologists), no reason to correct logical errors (that is for copy-editors), no reason to simplify opaque bureaucratic prose (that is a problem for the author of the start text); in fact there is probably no reason for any translation solution beyond the odd elegant Perspective Change, at a push. Agreed, in many instances it is advisable to stay close to the start text. But there is a reason for this and that reason need not contradict the need for the more complex solution types. In most cases, relative literalism is low-effort for the translator, mid-effort but probably acceptable in terms of effort for the receiver (who may benefit from a degree of foreignness), low-risk for the client (who generally likes to see a translation that resembles the start text and can thus give a sense of control) and low-risk for the translator as well (in cases of error or misunderstanding, you can always point to the start text and blame the problem on the author – literalism is often an instance of risk-transfer). There are many good pragmatic reasons for rendering a text as closely as possible, and there is no need to overlook those logics. However, as soon as the efforts and risks run out of kilter, as soon as ‘fidelity’ is to a communicative function or a purpose more than to start text expression, then other solution types are needed. And that will require more work, higher risks and hopefully greater gains. Hence the interest of solution types.

Postscript

The Flaw in the Dream

We have seen that the lists of solution types were originally formulated within an ideology of languages as separate systems, in a world where the prime aim of the translator was supposedly to maintain the separations by reducing interference. That is why there was so much attention to the levels of Modulation, Adaptation and the like. That was also a world where, in 1963, a Francophone linguist could still render the German name Wilhelm von Humboldt as a Gallicized ‘G[uillaume] de Humboldt’, effectively naturalizing the morphology to make the foreigner an honorary Frenchman, just as the American Edgar Allan Poe had long been rendered as ‘E. A. Poë’. That was a world where Vinay and Darbelnet could half-seriously contemplate ‘cricket’ being translated as ‘cycling’, where idiomatic expressions dutifully lined up in parallel (since all national cultures were presumed to have basically the same things to say) and where there was a plenitude of similar methods to ensure that foreignness would be abolished and translations would above all not read as translations. Each language had its génie, its personality, its apparently systematic preferences, and the role of translation was to keep those features in place. The overwhelming effect of those linguistic studies was to promote what other theories would call ‘domesticating’ translation (dating from Schleiermacher’s 1813 notion of ‘verdeutschende Übersetzung’, ‘Germanizing translation’), with highly conservative linguistic effects. The assumptions beneath such an approach were that the languages were not significantly dependent on each other: they were large and independent enough to be considered on an equal footing. At the time, French, English, German, Russian and Chinese were indeed strong and independent enough for such an assumption to be made, at least in the eyes of the reigning nationalist ideologies. Those linguistic approaches, and the translation approaches that they advocated, were designed to maintain that illusion of equal power, like the illusory equal votes in the Security Council of the United Nations, which was perhaps ultimately where the theorists dreamed their ideas might be applied.

244

Translation Solutions for Many Languages

Those features constituted the nucleus of a discourse on translation that was modernist and Western, which was then adapted in many different parts of the world and was made to speak to many more languages. One might say the same of Translation Studies itself as a set of Western discourses. The world is no longer there, however. Thanks to enhanced mobility on all levels (the international movements of professionals and their knowledge, as well as the mass migrations of the needy), there are no monolingual nations to be protected, and protectionist assumptions have become precarious. The demands placed on translation these days, in all countries, tend to concern languages with asymmetric power relations: we can no longer assume the kind of equal status that once underwrote the nationalist illusions of natural equivalence. Humboldt is no longer ‘Guillaume’ in French – he is ‘Wilhelm’, and happily so. Languages now develop through their dynamic relations with each other rather than from within internal systems. The status of the foreign is thus all the more important, acceptable, constantly available for integration, a part of normal development. And when the foreign has more to do with content than with the symbolics of naming, translators are called upon to bridge asymmetries by adapting texts, providing explanations, adding content and enhancing focus. The typology of solution types that I am proposing here hopefully reflects ways in which this new set of problems can be faced. It is still in the Western tradition, of course, but not unreflexively so. I have given weight to the copying of words, which means drawing on and presenting the foreign; I have highlighted the role of textual density as a factor that can be worked on by the translator when appropriate; I have allowed that translation can involve changes in content, moving beyond the conceptual limits of narrow equivalence. On all these fronts, a new typology should invite translators to widen their range of possible actions, pointing towards the higher-risk options further down the list, proposing more intervention and more invention. And beyond the professional coterie, the typology should be able to show the wider community that translators do more than replace words: we are on the frontlines of your terminology, we help shape the styles of your languages, we do what the language machines cannot do, we make texts work, we communicate across cultures. In this, paradoxically, the high-risk solution types that were once formulated in order to protect large national languages are now able to legitimize new roles for the translator. The ideas of the past can be used beyond the contexts of their original formulation. If there is a flaw in the project, it is not there. So will the new typology work for all languages; is it universal? That is not a claim I would want to make. There are several reasons for baulking at



The Flaw in the Dream

245

universality. The first is sophistic: I have taken my categories from language, but natural languages are systems with different categories (if not, there would be no need for translation), so the categories cannot be extended with any confidence beyond the languages I have dealt with (this can be assimilated to Chomsky’s claim that deep-seated universals will not help translation). A second potential flaw, more realist, is that there are so many variables in each translation situation that their sheer complexity overrides any attempt to isolate simple categories: at best we would need the fuzzy logics by which the types are complicated by the perspectives of effort and risk. The more fundamental flaw in the universalist dream is nevertheless staring us in the face as soon as we realize that the nature of basic translation preferences has changed significantly since the 1950s: we are now able to call Wilhelm by his name and add information about him if appropriate. This means that the fundamental concept we are elaborating, called ‘translation’, is itself a dynamic response to a moving world. Since the meaning of the term ‘translation’ is not universal, the solutions to its problems cannot be universal. That, however, is not a fatal flaw; it need not stop us in our tracks. It simply means that there is much waiting to be discovered and invented as we go along, in our translations and in our classrooms, in an endless quest of improvement.

Notes Preface 1

I prefer the term ‘start text’ rather than ‘source text’ for technical reasons that have little to do with the current topic. These days translators work not just from a single text but also from glossaries, translation memories and machine translation output, and any one of those resources could provide the ‘source’ for a solution. The role of the initial text is thus relativized; it is no more than the starting point for the translation process. Hence the use of ‘start text’, which is, after all, what is regularly said in neighbouring languages: Ausgangstext, texte de départ, texto de partida, testo di partenza (see also Pym 2011: 92).

Chapter 1: Charles Bally and the Missing Equivalents 1

2

3

Although not translated into English, to my knowledge, Bally occupies a major place in the modern history of French linguistics and is known in Germany. He was translated into Russian from 1923 and had a particular influence on Ukrainian translation theories of the 1920s (as we shall see). A form of reformulation is the basis of the ‘variational approach’ of Hewson and Martin (1991), where we find series of syntagms that are superficially not unlike Bally’s (e.g. 1991: 48). The concept of a ‘homologon’ that unites and underlies variants does not, however, seem to be useful or necessary. It is perhaps a pity that Hewson and Martin dismiss a misspelt ‘Bailly’ at the beginning of their approach (1991: 6), visibly without having read him. Similarly unfortunate is the dismissal of Vinay and Darbelnet because their ‘techniques’ are not all obligatory (1991: 7), when they were never meant to be. I have spent years looking for the way the term ‘equivalence’ entered translation theory. For a while I thought the word was used by the Valencian scholar Joan Lluís Vives in his De ratione dicende of 1533, but then it turned out that an over-enthusiastic Spanish translator from the Latin had put it there. Vives had a concept of equivalence, I suspect, when he outlined a ‘third genre’ of translation where ‘the matter and the words are [both] weighed up’ (‘Tertium genus est, ubi & res & verba ponderantur’, 1533: 168v). Yet he used no word for ‘equivalence’. Later I found the word in forgotten Russian theory, as we shall see in a later chapter. As

248 Notes

4

5

6

far as I can tell, though, no one really knows how ‘equivalence’ entered Translation Studies – Wilss (1982: 134–5) presumed it came from mathematics, but he was guessing, like the rest of us. Now, perhaps, Bally provides a new lead. ‘Il faudrait pouvoir insister sur la précision des résultats qu’on peut obtenir par l’emploi de ces procédés, mais cette digression nous entraînerait trop loin de notre sujet; il s’agit en effet de toute une méthode de traduction; je la pratique depuis plusieurs années au Séminaire de français moderne de l’Université de Genève; elle est pour moi une réalité absolument concrète; mais elle demanderait à être démontrée systématiquement, textes en mains; cela ne peut se faire que dans un ouvrage spécial, qui verra peut-être le jour plus tard’ (Bally 1909/51: 1.138). The repeated term ‘mechanical’ has a resonance in terms of Bergson’s distinction between mechanical (clock) time and interior time or durée (1889/1904) and indeed with his association of laughter as being caused by the human body acting like a machine (1900/40). For Bally, as for many reacting against the industrial revolution, the machine was the opposite of life and its affectivité. Bally goes on to talk about translation into the learner’s L2, in which ‘nothing is learned’ since it can be no more than the application of previous lessons (1905: 165). Here Bally applies a reasoning curiously similar to Ladmiral’s much later but similarly sophistic assertion that there is no such thing as L2 translation: ‘Le thème n’existe pas’ (1979: 50).

Chapter 2: Vinay and Darbelnet Hit the Road 1 2

3

4

http://www.dailymail.co.uk/news/article-2382237/California-motorists-puzzledPlease-Slow-Drively-road-sign.html (accessed November 2013). Malblanc described his work as ‘une simple confrontation des structures psychologiques de deux langues s’après leurs représentations personnelles du monde où, pour parler selon G. de Humboldt, gemäss der Verschiedenheit ihrer Weltansichten’ (1963: Avant-propos). The reference is from Wilhelm von Humboldt, Üeber das vergleichende Sprachstudium (1820/1945). Muñoz Martín (2000: 131) suggests that the term probably comes from Latin grammar, where the nominative case is the casus rectus and all other cases are casus obliqui. ‘Les quatre Grands’ may be a calque of ‘the four great powers’, but surely not a complete one? The omission of ‘power’ from the French is intriguing – four great what? I note that France did not test its first nuclear weapon until 1960, prior to which (and indeed following it), greatness was somehow more cultural than anything else.

Notes

249

5

Goffman actually describes the principal in terms of responsibility to content: ‘someone whose position is established by the words that are spoken, someone whose beliefs have been told, someone who is committed to what the words say’ (1981: 146). The specific responsibility of translators is investigated in Pym (2012). 6 ‘on peut craindre de voir les quatre-cinquièmes du globe se nourrir exclusivement de traductions et périr intellectuellement de ce régime de bouillie pour les chats’ (1958/72: 54). 7 ‘un galimatias qui n’a de nom dans aucune langue.’ Chesterman renders ‘galimatias’ as ‘balderdash’. 8 Okay, just one more: ‘Transposition frees translation from its empirical corset, renews and elevates it, rejuvenated, to the higher realm of the adult language, more subtle and more fluent’ (1945: 25). 9 I cite from memory my own translation of Laugier into English, done when I was an undergraduate, in 1978. My much belated thanks go to Professor Didier Coste for suggesting I translate the text. 10 In the later study, Folkart relates Laugier to Venuti’s toying with Italianized prepositions and notes that ‘it is considered good form to tout the virtues of translating close to the grain’ (2007: 291). She suggests that this may be due to a ‘colossal misunderstanding of Berman’s central concept, “la lettre”’ (292), which is quite possible, although one doubts that Berman was the necessary authority for timid literary/philosophical literalism into English.

Chapter 3: A Tradition in Russian and Environs 1

2

3 4

This chapter has been completed in constant dialogue with Nune Ayvazyan and with help from many colleagues, to all of whom I give my thanks (see Pym and Ayvazyan 2014). It has been revised with significant contributions from Oleksandr Kalnychenko. Azov (2013) gives the reference to F. D. Batyushkov, ‘Задачи художественных переводов’ [Problems of literary translation], in Принципы художественного перевода. Всемирная литература [Principles of literary translation. World literature], 1920: 7–15, which is apparently a two-volume collection including articles by Chukovskiy and Gumilyov (or Gumilev?), whom we will soon see arguing about whether there can be rules for translators. I cite from the online version: http://feb-web.ru/feb/litenc/encyclop/le8/le8-5121. htm, accessed January 2014. ‘Адекватным мы должны признать такой П., в к-ром переданы все намерения автора (как продуманные им, так и бессознательные) в смысле определенного идейно-эмоционального художественного воздействия на

250 Notes читателя, с соблюдением по мере возможности [путем точных эквивалентов или удовлетворительных субститутов (подстановок)] всех применяемых автором ресурсов образности, колорита, ритма и т. п.; последние должны рассматриваться однако не как самоцель, а только как средство для достижения общего эффекта.’ 5 The list of references includes M. Morozov, Technical translation from English into Russian. State Central Institute of Distance Learning Languages​​ (Moscow, 1934), which suggests that translation was being taught in distance mode in the 1930s. 6 This is not a trivial ambiguity. In part, it returns us to the difficulties of the term ‘règle’ as found in Bally, just as it might also be read into similar problems with the term ‘norm’ in Toury (1995/2012). One could certainly argue that frequency of occurrence underlies compliance with a rule, but there are also many rule-breaking actions that are statistically frequent (in moments of system change, for example, or in the criminal actions that do not change the nature of the law) and there are many non-statistical methods for the discovery of rules (sanctions for non-compliance, judgements by competent speakers, etc.). 7 ‘Ступінь трудности і в той же час ступінь точности перекладу для різних прозаїчних жанрів неоднакова і залежить вона від співвідношення конструктивних та естетичних елементів в мові жанрів, з одного боку, і від специфічної фразеології — з другого.’ 8 ‘Однако, как только мы обратимся к практике перевода, мы убедимся, что мера точности меняется в зависимости от цели перевода, характера переводимого текста и читателя, которому перевод предназначается.’ 9 In principle, this is not to be confused with the term ‘adequacy’ as used in Toury (1995/2012), which operates in opposition to ‘appropriateness’ to the target situation, as yet another binary opposition. At the same time, Fedorov’s focus in his examples is almost exclusively on the start text, with scant attention being paid to the target situation. 10 Tynyanov wrote a 1921 paper on translations of Heine (included in his Arxaisty i novatory in 1929 and in the French translation Formalisme et histoire littéraire of 1991); Fedorov wrote on the Russian Heine in 1929, in a volume edited by Eykhenbaum and Tynyanov; Tynyanov’s work on Heine was later explicitly commented on by Fedorov (1965/83). 11 My thanks to Oleksandr Kalnychenko and Itamar Even-Zohar for this information, which coincides with the names listed in the introduction to the fifth edition of Fedorov (1953/2002: 3); Schippel (2014) adds S. I. Bernstein to the list. 12 My thanks to Oleksandr Kalnychenko for sending Chukovskiy’s later explanation of this apparent collaboration. The connections with the Formalist period are found in the personal relations, when not in the publications. Chukovskiy’s diary

Notes

13

14

15 16 17

251

provides a fascinating view of the period. Chukovskiy was a friend of Tynyanov and knew the Formalists well. In his diary for 1927 we read, among much else, an account of a meeting between the two and Viktor Shklovskiy: ‘Then we had one of those wonderful talks about literature that flourished during the golden hungry days of Formalism – all quips and aphorisms’ (1991/2005: 212). Fedorov is mentioned peripherally in the diary: he was 21 at the time of the meeting, of a different generation (Chukovskiy was 55, Tynyanov was 33, Shklovskiy was 34). This is reported in Retsker (1974/2007: 6). The Gorky Literature Institute was founded in 1933. The important point is that Retsker was aware of what Fedorov (whom he calls ‘my teacher’) was doing in the 1930s and 1940s. In 1928 Bally coincided with the Russians Roman Jakobson and Nikolay Trubetskoy at the First International Conference of Linguists in The Hague, the Netherlands. That meeting of the Geneva School with traces of Russian Formalism nevertheless had more to do with the development of the Cercle linguistique de Prague, which had been meeting since 1926, than it seems to have had with systemic studies in the Soviet Union. The general direction of Soviet linguistics, under the influence of Marr, was way from the study of synchronic systems. Bally was certainly translated into Russian from 1923 (Velmezova 2006: 41); he is noted as having influenced the phraseology of Vinogradov; we have seen him cited by Ukrainian theorists in the late 1920s; but he was definitely out of favour during the period of class-based historical linguistics (as indeed were the Ukrainian theorists). Voloshinov commented negatively on Bally in Marxism and the Philosophy of Language (1929/86: 58–59), regarding him as a representative of Saussurean ‘abstract objectivism’, which was considered incompatible with the social concerns of a properly Marxist linguistics. In fact, the likes of Bally would appear to have remained out of favour until Stalin’s intervention in linguistics in 1950 (when Vinogradov was definitively brought back from banishment). After that intervention, Sobolev cites Bally’s stylistics in his textbook for French– Russian translation (1952: 91) and actually recommends that translation students do the exercises in the second volume of Bally’s Traité de stylistique française (1952: 396). So Bally was at least citable when Fedorov was working on his 1953 book. I note that Bally’s main work Linguistique générale et linguistique française (1932) was not published in Russian translation until 1955, although virtually all the translation theorists and linguists, including Fedorov, could have read Bally in French. There is a reference to Bally in the fifth edition (1953/2002: 180–1), but not in the first. Oleksander Kalnychenko, personal communication, September 26, 2014. The ideological justification for this goes back to Lenin’s 1905 ‘principle of Party literature’: ‘literature cannot be a means of enriching individuals or groups: it

252 Notes

18

19

20 21 22

cannot, in fact, be an individual undertaking, independent of the common cause of the proletariat. Down with non-partisan writers! Down with literary supermen! Literature must become part of the common cause of the proletariat’ (1905: 44). Such is the basic totalitarian argument. We will see this principle return in the translation theorists of Central Europe. ‘The first edition attempted to put the problem of translation (with all its variants) on a large scale as a problem of linguistics. This direction and categorical tone aroused many […] objections and comments about the fact that not everything in the book had been perceived and interpreted objectively and the author was also attributed views that went far beyond the direct sense of his judgments and accusations were made that he rejected the possibility of any other consideration of translation problems but the linguistic one and that he had ignored multilateral relations between the theory of (artistic) translation with literary and other humanitarian sciences. Some critics were also dissatisfied with the fact that different types of translation were compared within the book – news reports and information, research and others – alongside literary works. Literary translation was seen to be in some kind of danger and the critics failed to see that the author had made the comparisons in order to establish the specificities of each of the types.’ (1953/2002: 5–6) The term ‘formalist’ used here as an accusation has little to do with Formalism as an intellectual tradition. It belongs to the period between the 1948 decree (following the 1946 Zhdanov Doctrine) that denounced composers of the order of Shostakovich, Prokof ’yev and Khachaturyan for the presence of foreign elements in their music and the official ‘rehabilitation’ of these composers in 1958. ‘Formalist’ effectively meant ‘foreignizing’, ‘non-Soviet’, ‘not realistically portraying [idealized] life’. Lev Gumilyov, who one might presume is the poet in the recounted argument, was arrested by the Cheka and executed in 1921. These are noted in the fifth edition of Fedorov (1953/2002): Zeitschrift für slawische Philologie 24 (2) (1956) and the Italian Rassegna sovietica (1958). Kade himself is considerably more dialectical but nevertheless ultimately sees form as working ‘in the service’ of content: ‘Die Funktion der Form erschöpft sich im Dienst am Inhalt’ (1968: 47). Social determinism wins out in the last analysis. He laments that Fedorov only saw the unity of form and content, rather than the dynamic historical relations between the two categories. The issue is important in Marxist theory because, in principle, form belongs to the superstructure and content is the social life at the base, and the base is supposed to determine the superstructure. The theoretical problem with this, quite apart from the survival of language systems as in Stalin, is that it leaves the superstructure without any real work to do. Why write books or give speeches (why use forms?) if the social

Notes

23

24

25

26

253

system itself will naturally bring us to socialist paradise? The problem was resolved, for me, by Althusser’s insistence that the base determines the superstructure only in the very long run, in the ‘last instance’ (Althusser 1965). Awaiting that instance, the two levels are dynamically interdependent, such that work on forms can do much to improve the way people think, drawing on and refining the practical theorization that, for Althusser, comes from the base. To take an example close at hand, the initial models of translation solution types were very certainly framed by their social moments, yet their forms could be transferred from society to society, extending forms that were then not wholly determined by the new social contexts. ‘The Chair for German at the School of Advanced Studies in Foreign Trade in Budapest has graciously provided me with a typed copy of an almost complete German translation of the second edition (1958) of Fedorov’s book’ (Wilss 1977: 60n). Despite this precedent, no official German translation has yet been published. Retsker actually takes issue with the broader usages of the term, as found in Catford (1965) among others: ‘Equivalents are kinds of catalysts in the process of translation. Their role can hardly be overestimated’ (1974/2007: 13). Osimo’s presentation (2008) suggests that Torop’s mode of thought here follows the Danish semiologist or semiotician Louis Hjelmslev (1943/61), who divided language into expression and content and then pointed out that both these planes have form and substance, thus giving four levels at which things can happen. As productive as this could be for the organization of solution types, Torop seems not to use the second level of division; his distinction between ‘recoding’ and ‘transposition’ relies simply on the notions of form and content: recoding is where form is translated by form and transposition is where content is translated by content (1995/2000: 101). The relative non-operability of the typology seems to be manifested in the fact that, in May 2014, the online Italian, Brazilian Portuguese and German versions of Osimo (2008) correspond with Torop’s table (1995/2000: 104), but the English, Dutch, French, Spanish and Persian translations systematically reverse the order of elements in the last line, so that ‘Analysis/Autonomous’ gives ‘Macrostyle’, and so on. And no one seems to have noticed!

Chapter 4: A Loh Road to China 1 2

Here I retain the form ‘Loh Dian-yang’ because it is what appears on his books. The Pinyin transcription system used currently gives ‘Lu Dianyan’. Information taken from the Suzhou Biographical Records: http://www.dfzb. suzhou.gov.cn/zsbl/239527.htm (accessed October 2015). The biographies also

254 Notes

3

4

5

6

7

8

include notes on the fate of his daughter Lu Lanxiu: http://www.360doc.com/ content/07/1222/12/15797915716.shtml (accessed November 2014). My sincere thanks go to Professor Zhang Ling for the translation from Chinese. 而在1933年就加入国民党的陆殿扬，虽于胜利后脱离国民党，建国后曾 任解放军外语学院英语系主任、教授，立过三等功，此时也遭上海’ 造反派’威逼与诱供，打成’中统特务’ From Suzhou Biographical Records: http://www.dfzb.suzhou.gov.cn/zsbl/239527.htm (accessed October 2015). Translation by Ye Zinan, to whom I give my thanks. I owe these details to Ye Zinan of the Middlebury Institute for International Studies at Monterey. Zhang and Pan (2009: 369) are less sure that the two scholars were in the same faculty at the same time, but they do note that Zhang et al. (1980) repeat some of the examples given in Loh (1958). Later in the same talk: ‘Lu Xun translated Dead Souls, “Destruction,” and so on; however, his brilliance didn’t lie primarily in translation but in his own creative works’ (1956/91: 103). Chan (2001: 199) points out that the three principles were far from new: they were mentioned by Zhi Qian in his ‘Preface to the Faju jing’, published in 224 (yes, in the third century). Much as Loh might have picked up something similar from Tytler’s Essay on the Principles of Translation (1791/1978), he mentions Tytler only in passing, and negatively: ‘Before liberation, we got from the West almost nothing about the theories of translation, except the principles of A. F. Tytler, which were, however, of little help to us’ (1958: 1.16). The edition identifies the author as ‘Professor of English, Southeastern [Dongnan] University and Government Teachers College Nanking; Principle of Kiangsu First Middle School, Nanking’. This concurs with the biographical notes given in Zhang and Li (2009). Born in 1891, Loh would have been 33 at the time of this edition. Brian Baer, to whom I give my uncountable thanks, suggests that a further influence of Fedorov on Loh might be in the semi-technical term ‘way’. Where Vinay and Darbelnet refer to ‘procédés’, Loh talks about ‘translation principles’ (in his general taxonomy), occasionally ‘rules’ (cf. 1958: 2.115), ostensibly ‘techniques’ (in the title of his work) and then ‘ways’, in a looser sense that may not be highly technical. For example, when he says ‘[t]his is probably the way in which most of the people will translate the sentence’ (1958: 2.166), there seems to be nothing technical at stake. But then, in the chapter on ‘Repetition as a translation principle’, we find the heading ‘Repetition in Many Ways: Words are repeated in translation for the sake of clearness in many ways’ (1958: 2.151), which suggests that a ‘way’ is a specific operationalization of a principle. Later we find uses like the following: ‘Here the principle of conversion points out the way leading to adequate translation. It shows you the technique essential in improving your version’ (1958: 2.207, italics mine), where ‘way’ and ‘technique’ operate as

Notes

255

synonyms, subordinated by ‘principle’. This would perhaps not be of great interest if we did not find the following in Fedorov, published in Chinese just a few years before Loh: …vykhod is polozheniia dostigaetsia ili putem zameny slova, […] ili putem grammaticheskoi perestroika. (1953: 156) [An escape from this situation can be achieved either by means of [putem] substituting words or by means of [putem] grammatical restructuring.] Razumeetsia, […] put’ k takim zamenam – ne vsegda legkii put’ (1953: 156). [Of course, the path [put’] to such substitutions is not always an easy path [put.’]] No smyslovaia rol’ neopredelennogo artiklia mozhet byt’ otobrazhena po-russki putem postanovki podlezhashchego v kontse predlozheniia (1953: 167–8). [But the conceptual role played by the indefinite article can be reflected in Russian by means of [putem] placing the subject at the end of the sentence.] Here the word put’ (path, way) is functioning in a semi-technical way to describe how to tackle problems posed by non-corresponding grammatical structures. This appears to be echoed in the use of ‘way’ in Loh. The Chinese translation of Fedorov uses ‘手法’ (1953/5: 152, 154) or ‘方法’ (1953/5: 152), both of which can be translated as ‘ways’ or ‘methods’. In his References section (1958: 2.254), Loh refers to ‘Fedorov, A. B., Principles of Translation, Moscow’, giving the English title for a text never translated into English (unfortunately). This, plus the transcription of the name, would suggest Loh did indeed learn from the Chinese translation of Fedorov. 9 All examples from Loh have been back-translated and explained with the invaluable help of Maggie Ting Ting Hui. I have also kept one eye on the examples explained in Zhang and Pan (2009). 10 There are some precedents in the history of Chinese translation theory. For example, the seventh century Buddhist monk Xuanzang has a theory of ‘Five Untranslatables’ (五种不翻), which were situations where Buddhist terms had to be transliterated from Sanskrit. Yue (in Cheung 2006: 157) explains these as: (1) if a term carries a magic spell; (2) if a term bears multiple meanings; (3) if a term does not exist in the target culture; (4) if a term bears an established translation; and (5) if a term generates positive associations. Strictly speaking, these are reasons for choosing this solution, rather than actual solution types. By all accounts, though, Xuanzang was using a ‘toolbox’ approach to translation, breaking with primitive literalism when necessary yet importing terms when advisable. The category of ‘magic spell’ might apply to the irrational appeal of many English terms that are transcribed into other languages today.

256 Notes 11 This has remarkably little to do with the translation ‘method’ described as ‘inversion’ in Chan’s Dictionary of Translation Technology (2004: 117): ‘…used in rendering a series of long sentences [sic]. The translation begins from the last sentence and goes all the way back to the first.’ The reference is to Lin Jihai (1983: 17–19), whose most prominent term is nevertheless ‘reverse translation’. Why would any translator do such a thing? Of all of Loh’s terms, only Repetition appears in Chan’s dictionary. 12 Zhang and Pan strangely state that these ‘have no equivalents in Vinay & Darbelnet’s model’ (Zhang and Pan 2009: 367). 13 Nida (1964: 231) does mention repetition as a possible ‘adjustment’, specifically in the case of languages where repetition is used for emphasis or high epic style (‘I will kill him with my weapon, I will slay him with my sword’). Nida appreciates the rhetorical effects of repetition, but his final comment is faithful to the economy of contemporary English: ‘if it proves to be nothing more than a misleading tautology, it should be eliminated.’

Chapter 5: Spontaneous Combustion in Central Europe? 1

2

3

The name ‘Cercle linguistique de Prague’ (the Prague Linguistic Circle), significantly named in French, refers to scholars who met from 1926. These included the Russian Roman Jakobson, who had taken a position in Brno (and whose escape from Nazi-occupied Prague then took him to Copenhagen, Stockholm, New York and Harvard) and Nikolai Trubetzkoi, who actually held a chair in Vienna. A further member of the group was Henrik Becker, who attended the first meeting but lived in Leipzig (see Dušková 1999). The circle clearly extended beyond the city of Prague. In 1928, Jakobson, Trubetzkoi and other members of the group attended the First International Conference of Linguists in The Hague, the Netherlands, where they signed a resolution calling for synchronic linguistic analysis. One of the co-signers was Charles Bally. Popovič compiled lists that were more elaborate and perhaps more coherent in his later works (1975/2006, 1976), so my use of the 1968 version is probably unfair for anyone looking for the truth of solution types. The 1968 version is nevertheless of historical interest here, since I am concerned with where the ideas came from and what they were doing in their initial political context. The Slovak ‘vplyv kultúrnej politiky (princíp straníckosti) na prekladateľský program a prekladateľské podnikanie’ (Popovič 1975: 239) becomes the Italian ‘influenza della politica culturale sulla poetica e sull’impresa traduttiva’ (Popovič 1975/2006: 139), where the ‘Party principle’ is clearly missing. My thanks go to Igor Tyšš for the Slovak text.

Notes 4

5

6

257

This is the title suggested by Christina Schäffner, who was kind enough to lend me her copy of the book. Mary Snell-Hornby (2006) has also proposed ‘Chance and Regularity’. The work’s original title can be translated as ‘Subjective and objective factors in the translation process’ (Salevsky 2007: 367). My translation of the Italian is as follows: ‘The lack of competition between translators produces texts marked by conformity and routine. Examples can be found in languages that have few translators of quality’ (1975/2006: 141). The corresponding Slovak text is: ‘Nedostatok prekladateľskej konkurencie plodí vznik prekladov, ktoré sa vyznačujú rutinou a šablónovitosťou’ (again, my thanks go to Igor Tyšš). My thanks to Jana Králová for her translations from Levý’s work (sent as a personal communication, 27 November 2009).

Chapter 6: Cold War Dalliance with Transformational Grammar 1

2

3 4

5

6

‘This work was supported in part by the U.S.A. Army (Signal Corps), the Air Force (Office of Scientific Research, Air Research and Development Command) and the Navy (Office of Naval Research); and in part by the National Science Foundation and the Eastman Kodak Corporation’ (Chomsky 1957/2002: Preface). ‘Эквивалентом следует считать постоянное равнозначное соответствие, как правило, не зависящее от контекста.’ ‘“Equivalent” should be taken to mean a constant, regular, context-free correspondence’ (Retsker 1974/2007: 13; trans. Zlateva 1993: 22) and on the very next page Retsker is arguing against Catford, who may have been responsible for the newer, pluralist concept of equivalence. Kade (1968: 99–100) posits that the number of translation acts is infinite but the number of equivalence relations between two languages is finite. Chomsky may have thought that, since Quine (1960/2013) associated indeterminism with translation, the problem could be solved by simply pushing translation away. However, Chomsky later recognized the general nature of Quine’s theory of indeterminism, which he described as ‘true and uninteresting’ (1980: 14, 16). Harris’s retort here would be that the syntactic elision is also dependent on semantics: ‘One can say He plays violin and she piano, but can one say except jokingly He plays violin and she chess?’ (1991: 179). Well, yes, one can say that, and the area of doubt opened by Harris’s interrogative indicates the active requirement for individual interpretation. We say many things jokingly, and much of the wit is based on usage, not rules. This is not to suggest that Barkhudarov was a theorist of translation tendencies or universals. When he argues that the item ‘world conference’ calls up the relevant

258 Notes

7

verb ‘to convene’ in English but does not do so in the same way in Russian (1975: 221), he is conscripting Harris into straight comparative stylistics. I hope this is what Kade was trying to say, although I am not sure. He further states that translational transformations ‘may be grasped as complex interlingual substitution (when 1:1 correspondence obtains) or interlingual transformations (when there is no 1:1 correspondence)’ (1968: 100). But no example is offered (and a few would really have helped!). I can similarly only hope that Kade’s criticism of the productive intralingual transformations does not imply that the translational transformations are any less productive: Kade requires that aspect for his proposition that a finite number of solution types (okay, ‘equivalence types’) can produce an infinite number of translation solutions.

Chapter 7: Forays into Romance 1

In this text Vázquez-Ayora actually distinguishes between four levels of ‘stylistic performance in translation’: 1. Zero-degree (from Barthes): ‘a translation in whose performance no semiostylistic procedures have been applied’ (1979: 206). This may be because the prototext does not require procedures, or because the translator has failed to apply the procedures available. 2. First-stage translation: Application of transposition, omission, amplification, explicitation. ‘Native quality and text building increase considerably.’ Translation has ‘initiated the process of decentred translation’. 3. Second-stage translation: ‘Modulation and equivalence (Vinay and Darbelnet’s terminology) determine, at a more sophisticated level, the real metamorphosis of signifiers.’ 4. Third-stage translation: Use of macrostylistic procedures like ‘the cultural viability or epistemological adaptation and the compensation of both the textual invariant and the effects’. This concerns the confrontation of ‘the thinking patterns of the two confronted languages’ (the reference is to Van Dijk 1972).

Chapter 8: Meanwhile Back in German 1

This comment does not appear in the fourth edition (1992), but then Vinay and Darbelnet’s categories do not appear there either.

Notes 2

3

4

5

259

Despite this potentially radical break with previous approaches, Schreiber (2006a: 41ff.) reproduces Vinay and Darbelnet’s categories, runs through the usual criticisms and then recognizes that most of the subsequent typologies, including his own, have indeed been based on Vinay and Darbelnet. ‘Invariants’ are presumably values that are considered to be the same in the start text and the translation. Since the Skopos theorists had made equivalence-based theories seem unacceptable, they consciously avoided the term ‘equivalence’ in cases like this. For the rest of the world, ‘invariance’ might still mean ‘equivalence’. There is no easy equivalent for Umfeld here. Etymologically, it is the field of things that are immediately around the text. One could play with terms like ‘context translation’, ‘scene translation’, ‘setting translation’ and so on, but the translations are not really of the contexts, scenes or settings as such. Schreiber (2006b) compares this tripartite model with some classical theories, noting that his third category would ‘not really be translation’ for Perrot d’Ablancourt and involves whole genres that are excluded from consideration in Schleiermacher.

Chapter 9: Disciplinary Corrections 1

2

Since Sándor Hervey died in 1997, I assume that his appearance as co-author of books after that date refers to the fact that his methodological framework and text was being recycled for new languages. Hervey was a Hungarian-born Oxfordtrained Sinologist who taught for almost thirty years at St Andrews University, Scotland. The first book is presented as a ‘tried and tested course in translation methodology for third-year undergraduates in Modern Languages at the University of St Andrews’ (1992: 1). It was first designed for French, then used for the other languages. The coursebook for Chinese, however, was first published in 2010 (Pellat and Liu 2010) and does not include the solution types; it seems much more oriented to considerations on the level of cultural theory, for ‘final-year students’. The one for Russian, also published in 2010 (Andrews and Maksimova 2010), does not have ‘thinking’ in the title and completely replaces the analysis of solution types with the authors’ pragmatic Communicative Act Model. These two volumes might indicate the extent to which the discussion of solution types has been replaced by cultural theory and pragmatics, at least in the United Kingdom. They could also point to the wilful abandoning of a common metalanguage: the students who use the Chinese book will not be able to discuss translation with those who use the Russian book, and the disconnect will not be due to anything in the Chinese and Russian languages.

260 Notes 3

4

5 6

Elsewhere Molina and Hurtado-Albir describe a technique as ‘the result of a choice made by a translator’ (2002: 508, italics mine). Happy as I am with this lucid focus on product, there is little to be said for the term. In every English dictionary I have been able to consult, a ‘technique’ is a way of doing something, not the thing done. This is not to mention the inevitable reductive inaccuracies: Ramadān is a lunar month involving far more than fasting (see Thawabteh 2011: 114) and Panettone is, I believe, eaten at Christmas as well, and in many countries in addition to Italy. Professor Chesterman (personal communication October 2014) assures me that the footnotes would certainly be instances of Visibility Change. Phrase Structure Change, for example, includes the example ‘Dies[e] Ausgabe […] enthält’ (‘This issue […] contains’) rendered as ‘In the present issue […] you will find’ (1997: 96). This looks like classical Modulation (sorry, ‘Converses’), surely as much semantic as syntactic? The same example is also used for ‘clause structure change’, which begs the question of why two different solution types should be necessary.

Chapter 10: Going Japanese 1

2 3

4

5 6

Hasegawa (2005) says she designed her translation course at Berkeley starting out from five (fairly arbitrary) ‘techniques’: Generalization, Cultural Substitution, Loan Word Plus Explanation, Culture-Specific Collocations and Idioms. These were ostensibly taken over from Baker (1992), which is more like an introduction to the levels of text analysis (watch out for collocations, watch out for idioms, etc.) than a reasoned typology of solutions. That early list is replaced by Vinay and Darbelnet in Hasegawa (2011). My thanks to Masaru Yamada for this example. This differs from the ‘chunks’ that are used in translation memory tools, which are basically at sentence level. Here the term is closer to what other theories call ‘translation units’, albeit without any testing of their psychological pertinence. The way sentence elements can be aligned in comparative analysis should not be misconstrued as a portrayal of any ‘translation units’, which I understand to be the stretch of text the translator processes prior to solving a translation problem. Translation units concern the cognitive processes of working translators; they do not concern the comparative analysis of translation products. 状況全体の中から名詞に相当するものを取り出し、そうして抽出した名詞 と名詞との関係を、前置詞や動詞によって示す. The comparative linguists make much of these structures in Japanese, claiming that they represent a view of actions as being somehow open to the force of

Notes

7

8

9

261

nature, in a way that is apparently impossible in English. In Spanish, however, we have agentless actions all the time, as in the reflexive ‘Se abrió la puerta con el viento’ (‘Opened itself the door with the wind’) (Vázquez-Ayora 1977: 282) and no one tries to base a cultural identity on them. Maltz and Borker (1982, cit. Wardhaugh 1986/2006: 327–8) suggest there is also misunderstanding within English: ‘The mhmm a woman uses quite frequently means only “I’m listening,” whereas the mhmm a man uses, but much less frequently, tends to mean “I’m agreeing”.’ Consequently, men often believe that “women are always agreeing with them and then conclude that it’s impossible to tell what a woman really thinks,” whereas women “get upset with men who never seem to be listening”.’ He continues: ‘My reason for omitting it now seems wholly inadequate. I did not like to have “night” and “white” in such intimate juxtaposition’ (2002: 125). This was repaired in a later version, as cited above, although ‘bottom of night’ still seems absent without leave. Seidensticker was also criticized for omitting the fact that the tunnel marks the border between two provinces, which he still considers not ‘worth worrying about’ (2002: 124). Sakai is quite right to point out that translations address not only target-language readers but also members of the start culture, whose cultural identity is being signified in some way in the translation. This ‘heterolingual address’ certainly concerns most Western cultures in cases where, for example, a work is ranked according to the number of foreign translations it has, or subtitles appear on the screen and convey information to those who do not need to read them, or again when language services provided to immigrant communities are visible to non-immigrants, speaking not only of alterity but also of public spending. Yes, instances of heterolingual address can certainly be found. Yet there are very few cases, I suggest, where members of a culture critically read, analyse and debate the ways their texts are translated abroad. In this, the Japanese might approach exceptional status.

Chapter 11: The Proof of the Pudding is in the Classroom 1

2

The analogy with driving is remarkably common in the education literature. For example: ‘A simple analysis might break this activity into (say) 80 discrete components, e.g. start engine, release hand brake, [etc.]. However, not much thought is required to see that someone might be capable of demonstrating each of these discrete attributes yet still be an incompetent driver’ (Hager et al. 2002: 7). The sample problems were the work of Mike Holt, a professional translator who taught in my Master’s in Tarragona in the early 2000s.

262 Notes

Chapter 12: A Typology of Translation Solutions for Many Languages 1

2 3

4

5

6

A version of the typology was taught to a class at Stellenbosch University in March 2014, involving speakers of Afrikaans, French, German and Chinese. Later in 2014 several revised versions were then tried on some seventy-eight students at the Middlebury Institute for International Studies at Monterey, where the languages were Chinese, Japanese, Korean, Spanish, Portuguese, German, French and Russian. I have benefited greatly from the critical comments by all those students. With each successive revision, the typology has become increasingly traditional. Personal communication, 1 November 2014. The importance of density changes was driven home to me when teaching solution types in South Africa (my special thanks to Professor Ilse Feinauer), where official translation into many of the eleven official languages is required to invent terms and explain concepts. The relative absence of the concept from the traditional typologies might in turn be attributed to their use of languages of similar status and historical development (English and French do not present the same problems as translation from English to isiNdebele, for example). I have tried several terms for this, but none is immediately understood by all. Here I borrow from localization theory, where products and texts are regularly described as being ‘tailored’ to suit new markets. A tailor cuts cloth to the measure of the client, sewing things together in a one-off way, and students tend to grasp this. So when presenting the types in class, I am currently going with Text Tailoring, as a metaphor that is misunderstood less frequently than the others. A more complete definition of ‘communicative success’ is the maintenance or enhancement of long-term mutual benefits for all partners in the communication act (as in Pym 2012). In his later commentary, Heyns (2009) recognizes the insertion of his own voice here: ‘taking up [the narrator’s] injunction, I thought it out, but allowing something of my own dilemma to enter into his efforts and using the opportunity to provide a foreign reader with both the literal meaning and the more imaginative version’.

References Althusser, L. (1965), Pour Marx, Paris: François Maspero. Alves, F. (2005), ‘Bridging the Gap between Declarative and Procedural Knowledge in the Training of Translators: Meta-Reflection under Scrutiny’, Meta 50 (4). Available at: http://goo.gl/jMpVSC [accessed February 2014]. Andrews, E. and Maksimova, E. A. (2010), Russian Translation: Theory and Practice, London and New York: Routledge. Anzai, T. (1982a), 翻訳英文法:訳し方のルール [English Grammar for Translating: Rules of Translation], Tokyo: Babel Press. Anzai, T. (1982b), 英文翻訳術 [Techniques for Translating English Prose], Tokyo: Chikuma Gakugei Bunko. Anzai, T. (1983), 英語の発想 [Conceptualization of English], Tokyo: Chikuma Gakugei Bunko. As-Safi, A. (2011), Translation Theories, Strategies and Basic Theoretical Issues, Amman: Dar Amwaj. Azov, A. (2013), Поверженные буквалисты: из истории художественного перевода в СССР в 1920–1960-е годы, [Defeated Literalists: From the History of Literary Translation in the USSR in 1920–1960s], Moscow: Izd dom visshey shkoly ekonomiki. Badiou, A. (1988), L’Être et l’évènement, Paris: Seuil. Baker, M. (1992), In Other Words: A Coursebook on Translation, London and New York: Routledge. Bakker, M. and Naaijkens, T. (1991), ‘A Postscript: Fans of Holmes’, in K. van LeuvenZwart and T. Naaijkens (eds), Translation Studies: The State of the Art, Amsterdam and Atlanta: Rodopi, 193–208. Ballard, M. (1985), ‘Introduction’, in E. Cary, Comment faut-il traduire?, Lille: Presses Universitaires de Lille, 9–23. Bally, C. (1905), Précis de stylistique: Esquisse d’une méthode fondée sur l’étude du français moderne, Geneva: A. Eggiman. Bally, C. (1910), L’étude systématique des moyens d’expression, Geneva: A. Eggiman. Bally, C. (1951 [1909]), Traité de stylistique française, 2nd edn, Paris: Klincksieck. Bally, C. (1952 [1913]), Le langage et la vie, 2nd edn, Geneva: Droz. Bally, C. (1965 [1932]), Linguistique générale et linguistique française, 4th edn, Bern: Francke. Barkhudarov, L. S. (1975), Язык и перевод (Вопросы общей и частной теории перевода) [Language and Translation: Questions in the General and Specialized Theory of Translation], Moscow: International Relations.

264 References Barthes, R. (1953), Le degré zéro de l’écriture, Paris: Seuil. Bausch, K. R. (1981 [1968]), ‘Die Transposition’, in W. Wilss (ed.), Übersetzungswissenschaft, Darmstadt: Wissenschaftliche Buchgesellschaft, 277–9. Bedson, T. and Schulz, M. (2015), Sowjetische Übersetzungskultur in den 1920er und 1930er Jahren. Die Verlage Vsemirnaja literatura und Academia, Berlin: Frank & Timme. Benjamin, W. (2004 [1923]), ‘The Task of the Translator’, translated by H. Zohn, in L. Venuti (ed.), The Translation Studies Reader, 2nd edn, London and New York: Routledge, 75–85. Benveniste, E. (1966), Problèmes de linguistique générale, vol. 1, Paris: Gallimard. Bergson, H. (1904 [1889]), Essai sur les données immédiates de la conscience, 4th edn, Paris: F. Alcan. Bergson, H. (1940 [1900]), Le Rire: Essai sur la signification du comique, Paris: PUF. Berman, A. (1992 [1984]), The Experience of the Foreign: Culture and Translation in Romantic Germany, translated by S. Heyvaert, Albany: State University of New York Press. Blum-Kulka, S. (2004 [1986]), ‘Shifts of Cohesion and Coherence in Translation’, in L. Venuti (ed.), The Translation Studies Reader, London and New York: Routledge, 290–305. Bode, J. (2009), ‘Translation Strategies for Japanese: Reconsidering Chesterman’s Theory on Translation Strategies’, OTB Forum 2 (1): 15–21. Bourdieu, P. (1984 [1979]), Distinction: A Social Critique of the Judgement of Taste, London: Routledge. Bourdieu, P. (1993), The Field of Cultural Production, Cambridge: Polity Press. Brang, P. (1963 [1955]), ‘Das Problem der Übersetzung in sowjetischer Sicht’, Sprachforum 2: 124–34, reprinted in H. J. Störig (ed.), Das Problem des Übersetzens, Darmstadt: Wissenschaftliche Buchgesellschaft, 384–401. Bühler, K. (1982 [1934]), Sprachtheorie: Die Darstellungsfunktion der Sprache, Stuttgart and New York: Gustav Fischer. Callon, M. (1986), ‘Some Elements for a Sociology of Translation: Domestication of the Scallops and the Fishermen of St-Brieuc Bay’, in J. Law (ed.), Power, Action and Belief: A New Sociology of Knowledge?, London: Routledge, 196–223. Cary, E. (1957), ‘Théories soviétiques de la traduction’, Babel 3: 179–90. Cary, E. (1959), ‘Andréi Fédorov: Introduction à la théorie de la traduction’, Babel 5: 19–20. Catford, J. C. (1965), A Linguistic Theory of Translation: An Essay in Applied Linguistics, London: Oxford University Press. Chan, L. T. H. (2001), ‘What’s Modern in Chinese Translation Theory? Lu Xun and the Debates on Literalism and Foreignization in the May Fourth Period’, TTR: traduction, terminologie, rédaction 14 (2): 195–223. Chan, L. T. H. (2004), Twentieth-Century Chinese Translation Theory: Modes, Issues and Debates, Amsterdam and Philadelphia: John Benjamins.

References

265

Chan, S. W. (2004), Dictionary of Translation Technology, Hong Kong: The Chinese University Press. Chen, D. A. (2000), 英漢比較與翻譯 [English-Chinese Comparison and Translation], Beijing: China Translation and Publishing Corporation. Chernetsky, V. (2011), ‘Nation and Translation: Literary Translation and the Shaping of Modern Ukrainian Culture’, in B. J. Baer (ed.), Contexts, Subtexts and Pretexts: Literary Translation in Eastern Europe and Russia, Amsterdam and Philadelphia: John Benjamins, 33–54. Chesterman, A. (1997), Memes of Translation: The Spread of Ideas in Translation Theory, Amsterdam and Philadelphia: John Benjamins. Chesterman, A. (2005), ‘Problems with Strategies’, in K. Károly and Á. Fóris (eds), New Trends in Translation Studies: In Honour of Kinga Klaudy, Budapest: Akadémiai Kiadó, 17–28. Cheung, M. (2006), An Anthology of Chinese Discourse on Translation, vol. 1: From Earliest Times to the Buddhist Project, Manchester and Kinderhook: St Jerome. Chomsky, N. (1965), Aspects of the Theory of Syntax, Cambridge, MA: MIT Press. Chomsky, N. (1966), Cartesian Linguistics, New York and London: Harper & Row. Chomsky, N. (1980), Rules and Representations, New York: Columbia University Press. Chomsky, N. (2002 [1957]), Syntactic Structures, 2nd edn, Berlin and New York: Mouton de Gruyter. Chukovskiy, K. (2005 [1991]), Diary 1901–1969, translated by M. H. Heim, New Haven: Yale University Press. Chukovskiy, K. and Fedorov, A. V. (1930), Искусство перевода [The Art of Translation], Leningrad: Akademia. Chuquet, H. and Paillard, M. (1989), Approche linguistique des problèmes de traduction, Paris: Ophrys. Cragie, S., Gambarotta, P., Higgins, I. and Sándor, H. (2000), Thinking Italian Translation, A Course in Translation Method: Italian to English, London and New York: Routledge. Dai, C. (2010), ‘A Chinese Translation of Finnegans Wake: The Work in Progress’, James Joyce Quarterly 47 (4): 579–88. Delisle, J. (1988), Translation: An Interpretive Approach, translated by P. Logan and M. Creery, Ottawa: Ottawa University Press. Delisle, J., Lee-Jahnke, H., Cormier, M. C. and Albrecht, J. (1999), Terminologie de la traduction, Amsterdam and Philadelphia: John Benjamins. Derrida, J. (1985), ‘Des tours de Babel’, in J. F. Graham (ed.), Difference in Translation, Ithaca, NY: Cornell University Press, 165–207, 209–48. Derzhavin, V. (1927), ‘Проблема віршованого перекладу’ [The Problem of Poetic Translation], Pluzhanyn 9/10: 44–51. Elena García, P. (1990), Aspectos teóricos y prácticos de la traducción, Salamanca: Publicaciones de la Universidad de Salamanca.

266 References Englund Dimitrova, B. (2005), Expertise and Explicitation in the Translation Process, Amsterdam and Philadelphia: John Benjamins. Fawcett, P. (1997), Translation and Language: Linguistic Theories Explained, Manchester: St. Jerome. Faye, J. P. (1972), Langages totalitaires: la raison, critique de la raison narrative, l’économie, Paris: Hermann. Fedorov, A. V. (1929), ‘Русский Гейне’ [Russian Heine], in B. Eykhenbaum and Y. Tynyanov (eds), Русская поэзия 19 века [Russian Poetry of the Nineteenth Century], Leningrad: Gosudarstvennyi Institut Istorii Iskusstv, 248–98. Fedorov, A. V. (1941), О художественном переводе [On Literary Translation], Moscow: Izd-vo OGIZ. Fedorov, A. V. (1953), Введение в теорию перевода [Introduction to the Theory of Translation], Moscow: Literatury na inostrannykh yazykakh. Fedorov, A. V. (1955 [1953]), Введение в теорию перевода, Chinese translation: 翻译理论概要 [Introduction to the Theory of Translation], Beijing: Zhonghua. Fedorov, A. V. (1983 [1965]), ‘Юрий Тынянов и его переводы из Гейне’ [Yuriy Tynyanov and his Translations of Heine], in Искусство перевода и жизнь литературы: очерки [The Art of Translation and the Life of Literature: Essays], Leningrad: Otd-niye, 262–85. Fedorov, A. V. (1974 [1927]), ‘Проблема стихотворного перевода’ [Problems in the Translation of Poetry], Poetika 2: 104–18. Translated as ‘The Problem of Verse Translation’, Linguistics: An International Review 137: 13–29. Fedorov, A. V. (2002 [1953]), Основы общей теории перевода (лингвистические проблемы) [The General Theory of Translation (Linguistic Problems)], 5th edn of Fedorov (1953), Saint Petersburg: St. Petersburg State University Faculty of Philology; Moscow: Filologiya tri. Available at: http://goo.gl/wnWPvI [accessed 13 January 2015]. Finkel, O. (2007 [1929]), Теорія і працтика перекладу [The Theory and Practice of Translation], Kharkiv: Derzhvydav Ukrayiny. Folkart, B. (1991), Le Conflit des énonciations: traduction et discours rapporté, Montreal: Balzac. Folkart, B. (2007), Second Finding: A Poetics of Translation, Ottawa: University of Ottawa Press. Gambier, Y. (2008), ‘Stratégies et tactiques en traduction et interprétation’, in G. Hansen, A. Chesterman and H. Gerzymisch-Arbogast (eds), Efforts and Models in Interpreting and Translation Research, Amsterdam and Philadelphia: John Benjamins, 63–82. Ganshina, K. A. and Karpov, I. (1950), Вопросы теории и методики учебного перевода [Questions of the Theory and Methodology of Translation Teaching], Moscow: Akademiya pedagogicheskikh nauk RSFSR. Gentzler, E. (2001), Contemporary Translation Theories, 2nd edn, Clevedon: Multilingual Matters.

References

267

Gibová, K. (2012), Translation Procedures in the Non-Literary and Literary Text Compared, Norderstedt: Books on Demand. Gil-Bajardí, A. (2003), Procedimientos, técnicas, estrategias: operadores del proceso traductor, Master’s thesis, Universitat Autònoma de Barcelona. Goffman, E. (1981), Forms of Talk, Philadelphia: University of Pennsylvania Press. Gorbachov, S. (2013), ‘Grammatical Transformations in Ukrainian–English Translation of Official Texts’, Verges: Germanic & Slavic Studies in Review 2 (2): 26–37. Grice, H. P. (1975), ‘Logic and Conversation’, in P. Cole and J. L. Morgan (eds), Syntax and Semantics, vol. 3, New York: Academic Press, 41–58. Gyogi, E. (2015), ‘Critical Cultural Awareness in Language Classrooms through Translation: A Reflection on the Use of katakana’, in E. Torres-Simón and A. Fuertes Puerta (eds), And Translation Changed the World, Newcastle: Cambridge Scholars, 61–73. Hager, P., Holland, S. and Beckett, D. (2002), ‘Enhancing the learning and employability of graduates: The role of generic skills’, Business/Higher Education Round Table Position Paper 9. Available at: http://goo.gl/QvjmQn [accessed February 2014]. Halliday, M. A. K. (1977), ‘Text as Semantic Choice in Social Contexts’, in T. A. Van Dijk and J. S. Petöfi (eds), Grammars and Descriptions (Studies in text theory and text analysis), New York: Walter de Gruyter, 176–225. Halliday, M. A. K. (1985), An Introduction to Functional Grammar, London: Edward Arnold. Hardin, G. and Picot, C. (1990), Translate: Initiation à la pratique de la traduction, Paris: Dunod. Harris, Z. S. (1951), Methods in Structural Linguistics, Chicago: University of Chicago Press. Harris, Z. S. (1972 [1965]), ‘Transformational Theory’, Language 41 (3): 363–401; in S. Plötz (ed.), Transformationelle Analyse: Die Transformationstheorie von Zellig Harris und ihre Entwicklung, Frankfurt am Main: Athenäum, 103–54. Harris, Z. S. (1991), A Theory of Language and Information: A Mathematical Approach, Cambridge: Clarendon Press. Hasegawa, Y. (2005), ‘Designing a translation course for Japanese majors’, Proceedings of the 16th International Japanese/English Translation Conference, 80–4, Chicago, 3–6 June 2005. Available at: http://goo.gl/pMcvvA [accessed August 2014]. Hasegawa, Y. (2011), The Routledge Course in Japanese Translation, London and New York: Routledge. Haywood, L., Thompson, M. and Hervey, S. (2008 [1995]), Thinking Spanish Translation: A Course in Translation Method: Spanish to English, 2nd edn, London and New York: Routledge. Henschelmann, K. (1993), Zur Beschreibung und Klassifizierung von Übersetzungsverfahren, Lausanne: Centre de Traduction Littéraire. Hervey, S. and Higgins, I. (1992), Thinking Translation: A Course in Translation Method: French to English, London and New York: Routledge.

268 References Hervey, S. and Higgins, I. (2002 [1992]), Thinking French Translation: A Course in Translation Method: French to English, 2nd edn, London and New York: Routledge. Hervey, S., Higgins, I. and Loughridge, M. (2006 [2003]), Thinking German Translation: A Course in Translation Method: German to English, 2nd edn, London and New York: Routledge. Hewson, L. and Martin, J. (1991), Redefining Translation: The Variational Approach, London and New York: Routledge. Heyns, M. (2009), ‘Irreparable Loss and Exorbitant Gain: On Translating Agaat’, Journal of Literary Studies 25 (3): 124–35. Hjelmslev, L. (1961 [1943]), Omkring Sprogteoriens Grundlæggelse, translated by F. J. Whitfield as Prolegomena to a Theory of Language, Madison, WI: University of Wisconsin. Holmes, J. S (2005 [1970]), ‘Forms of Verse Translation and the Translation of Verse Form’, in Translated! Papers on Literary Translation and Translation Studies, 3rd edn, Amsterdam and Atlanta, GA: Rodopi, 23–34. Holmes, J. S (2005 [1972]), ‘The Name and Nature of Translation Studies’, Paper presented to the Third International Congress of Applied Linguistics, Copenhagen, 21–6 August 1972; revised version in J. S Holmes, Translated! Papers on Literary Translation and Translation Studies, 3rd edn, Amsterdam and Atlanta: Rodopi, 66–80. Hönig, H. G. and Kussmaul, P. (1996 [1982]), Strategie der Übersetzung, Ein Lehr- und Arbeitsbuch, Tübingen: Narr. Howard, D. (2011), ‘Failure Analysis in a Professional Translation Setting’, in A. Pym (ed.), Translation Research Projects 3, Tarragona: Intercultural Studies Group, 15–29. Howatt, A. P. (1985), A History of English Language Teaching, Oxford: Oxford University Press. Hui, M. T. T. (2012), ‘Risk management by trainee translators: A study of translation procedures and justifications in peer-group interaction’, PhD thesis, Universitat Rovira i Virgili. Available at: http://www.tdx.cat/handle/10803/83497 [accessed May 2015]. Humboldt, W. von (1945 [1820]), Über das vergleichende Sprachstudium in Beziehung auf die verschiedenen Epochen der Sprachentwicklung, Leipzig: F. Meiner. Humboldt, W. von (1963 [1816]), ‘Einleitung zu Aeschylos Agamemnon metrisch übersetzt von Wilhelm von Humboldt’, reprinted in H. J. Störig (ed.), Das Problem des Übersetzens, Darmstadt: Wissenschaftliche Buchgesellschaft, 71–96. Jakobson, R. (1981 [1935]), ‘The Dominant’, in S. Rudy (ed.), Selected Writings, vol. 3, The Hague: Mouton, 751–6. Jerome (Eusebius Hieronymus) (395/1962), ‘Ad pammachivm de optimo genere interpretandi’, in D. Ruiz Bueno (ed.), Cartas de San Jerónimo, Madrid: Biblioteca de Autores Cristianos, 486–504. Jettmarová, Z. (2011), ‘Editor’s Introduction to the English Edition’, in J. Levý, The Art of Translation, Amsterdam and Philadelphia: John Benjamins, xv–xxvi.

References

269

Jumpelt, R. (1961), Die Übersetzung naturwissenschaftlicher und technischer Literatur, Berlin: Langenscheidt. Kade, O. (1968), Zufall und Gesetzmäßigkeit in der Übersetzung, Leipzig: VEB Verlag Enzyklopädie. Kal’nychenko, O. (2011), ‘A Sketch of the Ukrainian History of Translation of the 1920s’, in T. Hermans, A. Chalvin, A. Lange and D. Monticelli (eds), Between Cultures and Texts: Itineraries in Translation History, Frankfurt am Main, Berlin, Bern, Brussels, New York, Oxford, Vienna: Peter Lang, 255–67. Kanaya, T. (2004), 英語にも主語はなかった [English Used to Have No Grammatical Subject], Tokyo: Kodansha. Karaban, V. (2004), Переклад англійської наукової і технічної літератури [Translation of English Scientific and Technical Texts], Vinnytsa: Nova Knyga. Kashkin, I. (1951), ‘О языке перевода’ [On the Language of Translation], Литературная газета [Literary Gazette] 142 (1/ December). Klaudy, K. (2010), ‘Specification and Generalization of Meaning in Translation’, in B. Lewandowska-Tomaszczyk and M. Thelen (eds), Meaning in Translation, Frankfurt am Main: Peter Lang, 81–104. Klaudy, K. and Károly, K. (2005), ‘Implicitation in Translation: An Empirical Justification of Operational Asymmetry in Translation’, Across Languages and Cultures 6 (1): 13–28. Koller, W. (1992 [1979]), Einführung in die Übersetzungswissenschaft, Heidelberg and Wiesbaden: Quelle & Meyer. Komissarov, V. (1972), ‘Лингвистические модели процесса перевода’ [Linguistic Models of the Process of Translation], Tetradi Perevodchika 9: 3–14, unpublished translation by D. Robinson and S. Ilinskaya. Kristeva, J. (1969), Semiotikè: Éléments pour une sémanalyse, Paris: Seuil. Kussmaul, P. (2000), ‘Types of Creative Translating’, in A. Chesterman, N. Gallardo San Salvador and Y. Gambier (eds), Translation in Context, Amsterdam and Philadelphia: John Benjamins, 117–26. Ladmiral, J. R. (1979), Théorèmes pour la traduction, Paris: Payot. Laugier, J. L. (1973), ‘Finalité sociale de la traduction: le même et l’autre’, in B. Malmberg, J. L. Laugier and M. Wandruszka (eds), La Traduzione: saggi e studi, Trieste: Lint, 23–32. Leighton, L. (1984), ‘Translator’s Introduction: Kornei Chukovsky and A High Art’, The Art of Translation: Kornei Chukovsky’s A High Art, Knoxville: University of Tennessee Press, ix–xxxii. Lenin, V. I. (1965 [1905]), ‘Party Organization and Party Literature’, Collected Works, Moscow: Progress Publishers, vol. 10, 44–9. Leuven-Zwart, K. van (1989), ‘Translation and Original: Similarities and Dissimilarities, I’, Target 1 (2): 151–82. Leuven-Zwart, K. van (1990), ‘Translation and Original: Similarities and Dissimilarities, II’, Target 2 (1): 69–96.

270 References Levý, J. (1969 [1963]), Die literarische Übersetzung, Bonn: Athenäum. Levý, J. (2011 [1963]), Umění překladu, translated by P. Corness, ed. Z. Jettmarová, The Art of Translation, Amsterdam and Philadelphia: John Benjamins. Lin, J. (林基海) (1983), 英譯漢「反譯法」探討 [The Method of ‘Reverse Translation’ in English-Chinese Translation], 翻譯通訊 [Translators’ Notes] 5: 17–19. Lin, Z. (2004 [1998]), ‘On Theories in Translation Studies’, translated by L. T. H. Chan, in L. T. H. Chan (ed.), Twentieth-Century Chinese Translation Theory: Modes, Issues and Debates, Amsterdam and Philadelphia: John Benjamins, 244–5. Liu, M. (2004 [1990]), ‘The Basic Paradigm of Chinese Translation Theory’, translated by H. Yang, in L. T. H. Chan (ed.), Twentieth-Century Chinese Translation Theory: Modes, Issues and Debates, Amsterdam and Philadelphia: John Benjamins, 236–9. Loh, D. Y. (1924), Thirty-two Questions on Practical Rhetoric, 2nd edn, Nanking: Kiangsu First Middle School. Loh, D. Y. (陆殿扬). (1958), 英汉翻译的理论与技巧 / Translation: Its Principles and Techniques, 2 vols, Beijing: Times Publishing. López Guix, J. G. and Minett Wilkinson, J. (1997), Manual de traducción inglés/ castellano: Teoría y práctica, Barcelona: Gedisa. Lu, X. (2004 [1931]), ‘A Reply to Qu Qiubai’, translated by L. T. H. Chan, in L. T. H. Chan (ed.), Twentieth-Century Chinese Translation Theory: Modes, Issues and Debates, Amsterdam and Philadelphia: John Benjamins, 158–61. Luo, X. (2003), ‘Elegance in Translation: A Postcolonial Mark’, Perspectives: Studies in Translatology 10: 207–14. Malblanc, A. (1944), Pour une stylistique comparée du français et de l’allemand: essai de représentation linguistique comparée, Paris: Didier. Malblanc, A. (1963), Stylistique comparée du français et de l’allemand: essai de représentation linguistique et étude de traduction, Paris: Didier. Mallafrè, J. (1991), Llengua de tribu i llengua de polis: Bases d’una traducció literària, Barcelona: Quaderns Crema. Malone, J. L. (1988), The Science of Linguistics and the Art of Translation: Some Tools from Linguistics for the Analysis and Practice of Translation, Albany: State University of New York Press. Maltz, D. N. and Borker, R. A. (1982), ‘A Cultural Approach to Male–Female Miscommunication’, in J. Gumperz (ed.), Language and Social Identity, Cambridge: Cambridge University Press, 196–215. Mao, Z. (1991 [1956]), ‘Talk with Music Workers, 24 August, 1956’, in M. Y. M. Kau and J. K. Leung (eds), The Writings of Mao Zedong 1949–1976, vol. 2, London: M. E. Sharpe, 94–107. Marx, K. (1959 [1867]), Capital, Moscow: Foreign Languages Publishing House. Maynard, S. K. (2007), Linguistic Creativity in Japanese Discourse: Exploring the Multiplicity of Self, Perspective and Voice, Amsterdam and Philadelphia: John Benjamins.

References

271

Mayoral Asensio, R., Kelly, D. and Gallardo, N. (1988), ‘Concept of Constrained Translation, Non-Linguistic Perspectives of Translation’, Meta 33 (3): 356–67. Mayoral Asensio, R. and Muñoz Martín, R. (1997), ‘Estrategias comunicativas en la traducción intercultural’, in P. Fernández Nistal and J. M. Bravo Gozalo (eds), Aproximaciones a los Estudios de Traducción, Valladolid: SAE, 143–92. Melby, A. and Warner, C. T. (1995), The Possibility of Language: A Discussion of the Nature of Language, with Implications for Human and Machine Translation, Amsterdam and Philadelphia: John Benjamins. Meschonnic, H. (1999), Poétique du traduire, Paris: Verdier. Miram, G. E., Dayneko, V. V., Taranukha, L. A., Grishchenko, M. V. and Gon, O. (2005), Основи Перекладу [Fundamentals of Translation], Kyiv: Nika-Center. Molina, L. and Hurtado Albir, A. (2002), ‘Translation Techniques Revisited: A Dynamic and Functionalist Approach’, Meta 4: 498–512. Mossop, B. (2013), ‘Andrei Fedorov and the origins of linguistic translation theory’, available at: http://goo.gl/KVmIgY [accessed November 2013]. Mounin, G. (1963), Les problèmes théoriques de la traduction, Paris: Gallimard. Muir, E. and Muir, W. (1959), ‘Translating from the German’, in R. A. Brower (ed.), On Translation, Cambridge, MA: Harvard University Press, 93–6. Muñoz Martín, R. (1998), ‘Translation strategies: Somewhere over the rainbow’, Paper presented to the 4th International Congress on Translation, Universitat Autònoma de Barcelona (includes diagrams that are not in the published version). Muñoz Martín, R. (2000), ‘Translation Strategies: Somewhere over the Rainbow’, in A. Beeby, D. Ensinger and M. Presas (eds), Investigating Translation, Amsterdam: John Benjamins, 129–38. Murakami, H. (2013), 恋しくて / Ten Selected Love Stories, Tokyo: ChuokoronShinsha. Nagal, S. (2008), ‘Lokale Grammatiken zur Beschreibung von lokativen Sätzen und ihre Anwendung im Information Retrieval’, PhD thesis, Ludwig-MaximiliansUniversität, Munich. Newmark, P. (1981), Approaches to Translation, Oxford: Pergamon Press. Newmark, P. (1988), A Textbook of Translation, Hemel Hempstead: Prentice Hall. Neubert, A. (1991), Die Wörter in der Übersetzung, Berlin: Akademie. Nida, E. (1964), Toward a Science of Translating: With Special Reference to Principles and Procedures Involved in Bible Translating, Leiden: E. J. Brill. Nida, E. and Taber, C. (1969), The Theory and Practice of Translation, Leiden: Brill. Nord, C. (1997), Translating as a Purposeful Activity: Functionalist Approaches Explained, Manchester: St. Jerome. Ortega y Gasset, J. (1983 [1937]), ‘Miseria y esplendor de la traducción’, in Obras completas, vol. 5, Madrid: Alianza, Revista de Occidente, 431–52. Osimo, B. (2008), Translation Course, translated into ten languages, available at: http:// courses.logos.it/ [accessed April 2014]. Panneton, G. (1945), ‘La Transposition, principe de la traduction: Son rôle essentiel

272 References dans l’interprétation de la pensée, Sa valeur de base technique’, unpublished thesis, Faculté des lettres de l’Université de Montréal. Pascua Febles, I. and Peñate Soares, A. L. (1991), Introducción a los Estudios de Traducción, Las Palmas de Gran Canaria: Corona. Pavlović, N. (2007) ‘Directionality in collaborative translation processes: A study of novice translators’, PhD thesis, Universitat Rovira i Virgili, available at: http://www. tdx.cat/handle/10803/8770 [accessed August 2014]. Pellat, V. and Liu, E. Y. (2010), Thinking Chinese Translation, A Course in Translation Method: Chinese to English, London and New York: Routledge. Pokorn, N. K. (2012), Post-Socialist Translation Practices: Ideological Struggle in Children’s Literature, Amsterdam and Philadelphia: John Benjamins. Popovič, A. (1968), Preklad a výraz [Translation and Expression], Bratislava: VSAV. Popovič, A. (1971), Poetika umeleckého prekladu: Proces a text [Poetics of Artistic Translation: Process and Text], Bratislava: Tatran. Popovič, A. (1975), Teória umeleckého prekladu: Aspekty textu a literárnej metakomunikácie [Theory of Artistic Translation: Aspects of Text and Literary Metacommunication], Bratislava: Tatran. Popovič, A. (1976), Dictionary for the Analysis of Literary Translation, Edmonton: University of Alberta, Department of Comparative Literature. Popovič, A. (2006 [1975]), Teória umeleckého prekladu: Aspekty textu a literárnej metakomunikácie, translated by D. Laudani as La scienza della traduzione, Aspetti metodologici: La comunicazione traduttiva, Milan: Hoepli. Pym, A. (2000), Negotiating the Frontier: Translators and Intercultures in Hispanic History, Manchester: St Jerome. Pym, A. (2005), ‘Explaining Explicitation’, in K. Karoly and Á. Fóris (eds), New Trends in Translation Studies: In Honour of Kinga Klaudy, Budapest: Akadémia Kiadó, 29–34. Pym, A. (2007), ‘On Shlesinger’s Proposed Equalizing Universal for Interpreting’, in F. Pöchhacker, A. L. Jakobsen and I. M. Mees (eds), Interpreting Studies and Beyond: A Tribute to Miriam Shlesinger, Copenhagen: Samfundslitteratur Press, 175–90. Pym, A. (2008), ‘On Toury’s Laws of How Translators Translate’, in A. Pym, M. Shlesinger and D. Simeoni (eds), Beyond Descriptive Translation Studies: Investigations in Homage to Gideon Toury, Amsterdam and Philadelphia: John Benjamins, 311–28. Pym, A. (2009), ‘Using Process Studies in Translator Training: Self-Discovery through Lousy Experiments’, in S. Göpferich, F. Alves and I. M. Mees (eds), Methodology, Technology and Innovation in Translation Process Research, Copenhagen: Samfundslitteratur, 135–56. Pym, A. (2010 [1992]), Translation and Text Transfer, Tarragona: Intercultural Studies Group. Pym, A. (2011), ‘Translation Research Terms – A Tentative Glossary for Moments of Perplexity and Dispute’, in A. Pym (ed.), Translation Research Projects 3, Tarragona: Intercultural Studies Group, 75–99.

References

273

Pym, A. (2012), On Translator Ethics: Principles for Mediation between Cultures, Amsterdam and Philadelphia: John Benjamins. Pym, A. (2014 [2010]), Exploring Translation Theories, 2nd edn, London and New York: Routledge. Pym, A., Malmkjær, K. and Gutiérrez-Colon, M. (2013), Translation and Language Learning, Luxembourg: European Commission. Pym, A. and Ayvazyan, N. (2014), ‘The case of the missing Russian translation theories’, Translation Studies. Available at: http://goo.gl/Q10pww [accessed November 2014]. Pym, A. and Torres-Simón, E. (2014), ‘The Pedagogical Value of Translation Solution Types’, Perspectives: Studies in Translatology 23 (1): 89–106. Available at: http://goo. gl/9XMCmq [accessed August 2014]. Pym, A. (2015a), ‘A response to the response to “The case of the missing Russian translation theories”’, Translation Studies. Available at: http://goo.gl/eCEL4j [accessed July 2015]. Pym, A, (2015b), ‘Translating as Risk Management’, Journal of Pragmatics, 85: 67–80, available at: http://dx.doi.org/10.1016/j.pragma.2015.06.010 [accessed September 2015]. Quine, W. (2013 [1960]), Word and Object, Cambridge, MA: MIT Press. Rajaud, V. and Brunetti, M. (1992), Traducir: Initiation à la pratique de la traduction, Paris: Dunod. Real Academia Española (2005), ‘Diccionario panhispánico de dudas’, available at: http://www.rae.es/recursos/diccionarios/dpd [accessed January 2014]. Reiss, K. (2000 [1971]), Translation Criticism: Potential and Limitations, Categories and Criteria for Translation Quality Assessment, translated by E. F. Rhodes, Manchester: St Jerome. Reiss, K. and Vermeer, H. J. (1984), Grundlegung einer allgemeinen Translationstheorie, Tübingen: Niemeyer; translated by C. Nord as Towards a General Theory of Translational Action: Skopos Theory Explained, Manchester: St. Jerome, 2013; page numbers herein are to the German version. Retsker, Y. I. (1950), ‘О Закономерных соответствиях при переводе на родной язык’ [On Regular Correspondences in Translation into the Vernacular], in K. A. Ganshina and I. Karpov (eds), Вопросы теории и методики учебного перевода [Questions of the Theory and Methodology of Translation Teaching], Moscow: Akademiya pedagogicheskikh nauk RSFSR, 156–83. Retsker, Y. I. (2007 [1974]), Теория перевода и переводческая практика [Theory of Translation and Translation Practice], Moscow: R. Valent. A version of the first chapter is edited and translated in Zlateva 1993, 19–31. Revzin, I. I. and Rozentsveyg, V. I. (1964), Основы Общего И Машинного Перевода [Fundamentals of General and Machine Translation], Moscow: Vysshaia Shkola. Robinson, D. (2003), Performative Linguistics: Speaking and Translating as Doing Things with Words, London and New York: Routledge.

274 References Robinson, D. (2013a), Feeling Extended: Sociality as Extended Body-Becoming-Mind, Cambridge, MA: MIT Press. Robinson, D. (2013b), Schleiermacher’s Icoses: Social Ecologies of the Different Methods of Translating, Bucharest: Zeta Books. Sakai, N. (1997), Translation and Subjectivity, on ‘Japan’ and Cultural Nationalism, Minneapolis and London: University of Minnesota Press. Sakamoto, A. (2014), ‘Translators theorising translation: A study of Japanese/English translators’ accounts of dispute situations and its implications for translation pedagogy’, PhD thesis, University of Leicester. Salevsky, H. (2007), ‘Über die Sprache hinaus (In memoriam Otto Kade)’, in G. Wotjak (ed.), Quo vadis Translatologie? Ein halbes Jahrhundert universitäre Ausbildung von Dolmetschern und Übersetzern in Leipzig, Berlin: Frank & Timme, 367–86. Saussure, F. de (1974 [1916]), Course in General Linguistics, translated by W. Baskin, Glasgow: Fontana Collins. Saussure, F. de (2002), Écrits de linguistique générale, S. Bouquet and R. Engler (eds), Paris: Gallimard. Scavée, P. and Intravaia, P. (1979), Traité de stylistique comparée: analyse comparative de l’italien et du français, Paris, Brussels, Mons: Didier, Centre International de Phonétique Appliquée. Scheinberg, R. (2014), ‘Vinay and Darbelnet’s methods of translation applied to translation between Japanese and English’, Master’s thesis, Middlebury Institute of International Studies at Monterey. Schippel, L. (2014), ‘Travelling Concepts? oder von “eisernen” und “gläsernen” Vorhängen und anderen Hindernissen’, Paper presented to the conference Going East: Discovering New and Alternative Traditions in Translation (studies), University of Vienna, 12–13 December 2014. Schleiermacher, F. (1963 [1813]), ‘Ueber die verschiedenen Methoden des Uebersezens’, in H. J. Störig (ed.), Das Problem des Übersetzens, Darmstadt: Wissenschaftliche Buchgesellschaft, 38–70. Schmitt, P. A. (1987), ‘Fachtextübersetzungen und “Texttreue”’, Lebende Sprachen 32: 1–7. Schreiber, M. (1993), Übersetzung und Bearbeitung, Zur Differenzierung und Abgrenzung des Übersetzungsbegriffs, Tübingen: Narr. Schreiber, M. (1997), ‘Übersetzungsverfahren: Klassifikation und didaktische Anwendung’, in E. Fleischmann, P. A. Schmitt and W. Kutz (eds), Translationsdidaktik, Tübingen: Narr, 219–26. Schreiber, M. (1998), ‘Übersetzungstypen und Übersetzungsverfahren’, in M. SnellHornby, H. G. Hönig, P. Kussmaul and P. A. Schmitt (eds), Handbuch Translation, Tübingen: Stauffenburg, 151–4. Schreiber, M. (2006a), Grundlagen der Übersetzungswissenschaft, Französisch, Italienisch, Spanisch, Tübingen: Niemeyer.

References

275

Schreiber, M. (2006b), ‘A Prototypical Model of Translation Types’, in C. Heine, K. Schubert and H. Gerzymisch-Arbogast (eds), Text and Translation: Theory and Methodology of Translation, Tübingen: Narr, 382–90. Séguinot, C. (1991), ‘A Study of Student Translation Strategies’, in S. Tirkkonen-Condit (ed.), Empirical Research in Translation and Intercultural Studies, Tübingen: Narr, 79–88. Seidensticker, E. (1958), ‘On Trying to Translate Japanese’, Encounter 59 (August 1958): 12–19. Seidensticker, E. (2002), Tokyo Central, Seattle: University of Washington Press. Seidensticker, E. and Anzai, T. (1983), 日本文の翻訳 [The Translation of the Japanese Sentence], Tokyo: Daishukan Shoten. Shadrin, V. (2011), ‘Как обучать переводчиков XXI века?’ [How to Train 21st Century Translators?], Zhurnal Sankt-Peterburgskiy Universitet [Journal of the University of Saint-Petersburg] 8 (3833): unpaginated, available at: http://journal. spbu.ru/?p=405 [accessed September 2014]. Shveytser, A. D. (1987 [1973]), Перевод и лингвистика [Translation and Linguistics], Moscow: Voyennoye Izdatelstvo; translated as Übersetzung und Linguistik, Berlin: Akademie. Siever, H. (2010), Übersetzen und Interpretation: die Herausbildung der Übersetzungswissenschaft als eigenständige wissenschaftliche Disziplin im deutschen Sprachraum von 1960 bis 2000, Frankfurt am Main: Peter Lang. Smirnov, A. A. (1934), ‘Перевод’ [Translation], in Литературная энциклопедия [Encyclopaedia of Literature and Folklore], available at: http://feb-web.ru/feb/litenc/ encyclop/le8/le8-5121.htm [accessed January 2014]. Snell-Hornby, M. (2006), The Turns of Translation Studies: New Paradigms or Shifting Viewpoints?, Amsterdam and Philadelphia: John Benjamins. Sobolev, L. N. (1950), ‘О мере точности в Переводе в Теория и Методика Учебного Перевода’ [On the Measure of Exactitude in Translation], in K. A. Ganshina and I. Karpov (eds), Вопросы теории и методики учебного перевода [Questions of the Theory and Methodology of Translation Teaching], Moscow: Akademiya pedagogicheskikh nauk RSFSR, 141–55. Sobolev, L. N. (1952), Пособие по переводу с русского языка на французский [Manual for Translating from Russian into French], Moscow: Izd-vo lit-ry na inostrannykh yazykakh. Špirk, J. (2009), ‘Anton Popovič’s Contribution to Translation Studies’, Target 21 (1): 3–29. Špirk, J. (2014), Indirect Translations and Non-Translation: The (Fateful) Adventures of Czech Literature in 20th-Century Portugal, Newcastle: Cambridge Scholars Publishers. Spivak, G. C. (trans.) (1997), Breast Stories, translation ‘with introductory essays’ of three stories by Mahasweta Devi, Kolkata: Seagull. Stalin, J. V. (1954 [1950]), ‘Concerning Marxism and Linguistics’, Pravda 20 June 1950; in Marxism and Problems of Linguistics, Moscow: Foreign Languages Publishing House. Available at: http://goo.gl/H7s5PF [accessed January 2015].

276 References Steiner, G. (1998 [1975]), After Babel: Aspects of Language and Translation, Oxford: Oxford University Press. Störig, H. J. (ed.) (1963), Das Problem des Übersetzens, Darmstadt: Wissenschaftliche Buchgesellschaft. Strohmeyer, F. (1921), Französische Grammatik auf sprachhistorisch-psychologischer Grundlage, Berlin and Leipzig: Teubner. Strohmeyer, F. (1924), Der Stil der französische Sprache, Berlin: Weidmann. Sudhoff, K. (1909), ‘Planta noctis’, Archiv für Geschichte der Medizin 3 (4–5): 352. Tao, Y. (2005), ‘Translation Studies and Textbooks’, Perspectives: Studies in Translatology 13 (3): 188–204. Tao, Y. (2011), ‘The Development of Translation Textbooks in China (1949–2009): A Historical View’, Translation Quarterly 60: 36–65. Thawabteh, M. A. (2011), ‘The Other Side of the Coin of Lexical Borrowing from Arabic into English’, TranscUlturAl 1 (4): 103–22. Available at: http://goo.gl/prSl4n [accessed December 2013]. Thome, G. (1981), ‘Die wörtliche Übersetzung’, in W. Wilss (ed.), Übersetzungswissenschaft, Darmstadt: Wissenschaftliche Buchgesellschaft, 302–22. Tian C. (2013), ‘A sociocultural analysis of retranslations of classic English novels in Mainland China 1949–2009’, PhD thesis, Universitat Rovira i Virgili. Available at: http://hdl.handle.net/10803/135007 [accessed May 2015]. Torop, P. (2000 [1995]), Тотальный перевод, Tartu: Tartu University Press, translated by B. Osimo as La traduzione totale, Modena: Guaraldi Logos. Toury, G. (1992), ‘“Everything has its Price”: An Alternative to Normative Conditioning in Translator Training’, Interface 6 (2): 60–72. Toury, G. (2012 [1995]), Descriptive Translation Studies and beyond, Amsterdam and Philadelphia: John Benjamins. Tynyanov, Y. (1929), Архаисты и новаторы [Archaists and Innovators], Moscow: Akademia. Tynyanov, Y. (1991), Formalisme et histoire littéraire, translated by C. Depretto-Genty, Lausanne: L’Âge d’Homme. Tytler, A. F. (1978 [1791]), Essay on the Principles of Translation, Amsterdam: John Benjamins. Vallejo, J. de (1913), Memorial de la vida de fray Francisco Jiménez de Cisneros, A. de la Torre y del Cerro (ed.), Madrid: Centro de Estudios Históricos Bailly-Bailliere. Van Dijk, T. (1972), Some Aspects of Text Grammars, The Hague: Mouton. Van Niekerk, M. (2004), Agaat, Cape Town: Tafelberg. Van Niekerk, M. (2006), Agaat, translated by M. Heyns and M. van Niekerk, Cape Town: Jonathan Ball. Vázquez-Ayora, G. (1977), Introducción a la traductología: Curso básico de traducción, Washington, DC: Georgetown University Press. Vázquez-Ayora, G. (1979), ‘Semiostylistics: Stylistics plus Interpretation’, Babel 25 (4): 204–6.

References

277

Velmezova, E. (2006), ‘Charles Bally au filtre de sa réception en Russie’, in J. L. Chiss (ed.), Charles Bally (1865–1947): Historicité des débats linguistiques et didactiques: stylistique, énonciation, crise du français, Leuven: Peeters, 41–51. Venuti, L. (2013), Translation Changes Everything, London and New York: Routledge. Vermeer, H. J. (1987), ‘What Does it Mean to Translate?’, Indian Journal of Applied Linguistics 13 (2): 25–33. Veselica Majhut, S. (2012), ‘Cultural specificity in the translation of popular fiction from English into Croatian during the Communist and transition periods (1960– 2010)’, PhD thesis, Universitat Rovira i Virgili, available at: http://goo.gl/Rp6h8P [accessed December 2014]. Vinay, J. P. and Darbelnet, J. (1972 [1958]), Stylistique comparée du français et de l’anglais: méthode de traduction, Nouvelle édition revue et corrigée, Paris: Didier. Vinay, J. P. and Darbelnet, J. (1989 [1958]), ‘Translation procedures’, translated by A. Chesterman, in A. Chesterman (ed.), Readings in Translation Theory, Helsinki: Oy Finn Lectura, 61–9. Vinay, J. P. and Darbelnet, J. (1995 [1958]), Comparative Stylistics of French and English: A Methodology for translation, translated by J. C. Sager and M. J. Hamel, Amsterdam and Philadelphia: John Benjamins. Vives, J. L. (1533), De ratione dicendi, Lovanij: ex officina Rutgeri Rescija. Voloshinov, V. N. (1986 [1929]), Marxism and the Philosophy of Language, Cambridge, MA: Harvard University Press. Von Polenz, P. (1979 [1967]), ‘Fremdwort und Lehnwort sprachwissenschaftlich betrachtet’, Muttersprache 77: 65–80; reprinted in P. Braun (ed.), FremdwortDiskussion, Munich: UTB Fink, 9–31. Vossler, K. (1929), Frankreichs Kultur und Sprache: Geschichte der französischen Schriftsprache von den Anfängen bis zur Gegenwart, Heidelberg: Carl Winter. Wakabayashi, J. (2006), ‘Translating in a Forked Tongue: Interlinear Glosses as a Creative Device in Japanese Translations’, Translation and Interpreting Studies 1 (2): 3–41. Wardhaugh, R. (2006 [1986]), An Introduction to Sociolinguistics, 5th edn, Malden, MA: Blackwell. Waugh, L., Álvarez Valencia, J., A Do, T. H., Michelson, K. and Thomas, M. B. (2013), ‘Meaning in Texts and Contexts’, in K. Allan (ed.), The Oxford Handbook of the History of Linguistics, Oxford: Oxford University Press, 613–34. White, S. (1989), ‘Backchannels across Cultures: A Study of Americans and Japanese’, Language and Society 18: 59–76. Wilss, W. (1977), Übersetzungswissenschaft, Probleme und Methoden, Stuttgart: Ernst Klett. Witt, S. (2016), ‘Byron’s Don Juan in Russian and the “Soviet School of Translation”’, Translation and Interpreting Studies 11 (1). Wotjak, G. (1985), ‘Techniken der Übersetzung’, Fremdsprachen 29: 24–34.

278 References Yamada, M. (2011), ‘The Effect of Translation Memory Databases on Productivity’, in Anthony Pym (ed.), Translation Research Projects 3, Tarragona: Intercultural Studies Group, 63–73. Yan, F. (1901), ‘Preface to Tianyanlun’, translated by C. Y. Hsu, in L. T. H. Chan (ed.), Twentieth-Century Chinese Translation Theory: Modes, Issues and Debates, Amsterdam and Philadelphia: John Benjamins, 69–71. Ye, Z. and Shi, L. X. (2009), Introduction to Chinese–English Translation, New York: Hippocrene Books. Zabalbeascoa, P. (2000), ‘From Techniques of Translation to Types of Solutions’, in A. Beeby, D. Ensinger and M. Presas (eds), Investigating Translation, Amsterdam and Philadelphia: John Benjamins, 117–27. Zhang, M. (2001), English–Chinese Translation Textbooks in China (1949–1998), Shanghai: Shanghai Foreign Languages Education Publishing House. Zhang, M. and Li, P. (2009), ‘Introducing a Chinese Perspective on Translation Shifts: A Comparative Study of Shift Models by Loh and Vinay & Darbelnet’, The Translator 15 (2): 351–74. Zhang, P., Yu, Y., Li, Z. and Moyu, P. (1980), 英汉翻译教程 [Textbook of Translation from English into Chinese], Shanghai: Shanghai Foreign Language Education Press. Zimmer, R. (1990), Äquivalenzen zwischen Französisch und Deutsch, Tèbingen: Niemeyer. Zlateva, P. (trans. and ed.) (1993), Translation as Social Action: Russian and Bulgarian Perspectives, London and New York: Routledge. Zukovsky, C. and Zukovsky, L. (1969), Catullus, London: Cape Goliard Press.

Index accuracy 108 Adaptation 27, 47, 144–5, 160, 179, 194, 212, 230–1 Addition 39, 66, 71, 125, 127, 160, 232 adequacy (адекватность) 40–1, 47–9, 54, 62, 68, 82, 86–7 Althusser, L. 253 n.22 Amplification 28, 93, 145, 175–5, 213–14 Analogue translation 42, 105 Antonomy (negation) 91, 143, 215 Anzai, T. 186, 191–5 Aury, D. 33 authority 20, 167 back-channels 197 Badiou, A. xii Baer, B. 254 n.8 Bally, C. 1–13, 44, 50, 78, 124, 153, 247 nn.1, 2, 251 n.14 Barkhudarov, L. S. 48, 64–7, 257 n.6 Bearbeitung 157–8, 161, 166, 231 Benjamin, W. 39, 153 Bergson, H. 248 n.5 Berman, A. 249 n.10 bilinguals 30 binarism xiv, 49 borrowing (emprunt) 47, 189–90, 210 Brang, P. 61 Bühler, K. 43 Calque 23, 189, 190, 203, 210, 223 Calquing (калькирование) 46 Cary, E. 57, 60–4 Catford, J. C. 65, 109 Cercle Linguistique de Prague 102–3, 256 n.1 Chassé-croisé (Cross-transposition) 26, 131, 141–2 Chesterman, A. 175–9, 226 Chomsky, N. 121–6, 257 nn.1, 4 Chukovskiy, K. 50, 58–9, 250 n.12 Cisneros, F. 77

Compensation 29, 67, 110, 146–7, 166, 169–70, 177, 196, 229 Componential analysis 165–6 Conversion 87, 214 copying 105, 221–3 correspondence (соответствие) 39, 47 correspondence (Entsprechung) 155 Correspondence (équivalence) 27, 30, 67, 173, 212, 230 cultural transposition 168 culture-specific items 46, 105, 180, 231 Dai, C. 85 Darbelnet, J. 17 Delisle, J. 172, 205 Density change 226 Derrida, J. 105 Derzhavin, V. 40, 43–4 Description 174 Descriptive Translation Studies 119–20 dialectic unity 57, 63, 252 n.22 effort/work 179, 182, 236, 240–1 equivalence xii, 6, 8, 41, 62, 65, 122, 124, 173, 247 n.3 Équivalence 19, 27, 30, 67, 144, 173–4, 230 equivalence types (Kade) 114–15, 155 Étoffement 28 exactitude 44 exoticism 168–9 Explicitation 28, 92, 145–6, 160, 162, 227–8 Fawcett, P. 37, 42 Fedorov, A. 38, 43, 45–61, 81–2 Finkel, O. 44 Folkart, B. 33–4, 249 n.10 footnotes 177 formalism 57–9, 252 n.19 functions 6–8 Generalization 28–9, 175, 227

280 Index Goffman, E. 249 n.5 Gorbachov, S. 71 Gorkiy, M. 54, 61 Gumilyov, L. 59, 249 n.2, 252 n.20 Halliday, M. A. K. 45 Harris, Z. 121, 125, 127–8, 257 n.5 Hasegawa, Y. 186, 192, 195, 197, 260 n.1 hermeneutics 149 Hervey, S. 168–70, 257 n.1 Hewson, L. 247 n.2 Heyns, M. 238–9, 262 n.6 Higgins, I. 168–70 Hitler, A. 76–7 Holmes, J. S 102, 119, 235 Hönig, H. G. 44 Humboldt, W. von 152 Hurtado-Albir, A. 173 idioms 27, 144, 194, 231 Implicitation 28 interlanguage 131 Intravaia, P. 149 Inversion 87–90, 214 Jakobson, R. 70, 103, 105, 114, 251 n.14, 256 n.1 Jettmarovà, Z. 103 Jumpelt, R. 63 Kade, O. 63–4, 113–16, 129–32, 153, 155, 252 n.22, 257 n.3, 258 n.7 Kashkin, I. A. 57–8 kernels 125 Klaudy, K. 226 Koller, W. 154–5 Komissarov, V. 132 Králová, J. 257 n.6 Kussmaul, P. 44, 226 Ladmiral, J. R. 248 n.6 Laugier, J. L. 33 laws 4 levelling 110 Levý, J. 63, 102–9 Literal translation 25, 86, 139–40, 148–9, 172, 178 literalism 55, 198, 242 literary translation 46, 54–5

Liu, M. 98 Loan (emprunt) 23, 47, 192, 189–92, 210 Loh Dian-yang 73–5, 81, 213–15, 253 n.1, 254 n.7 Lu Xun 79–80, 94, 254 n.5 machine translation 133 Malblanc, A. 1, 21–3, 144 Mallafrè, J. 95 Malone, J. L. 226 Mao Zedong 77–8, 80 Martin, J. 247 n.2 Mayoral, R. 180 McDonald’s 84 mediation xi Meschonnic, H. 34 Miko, F. 111 modernism 76 Modulation 26–7, 142–4, 160, 173, 175, 193, 211–12, 225 Molina, L. 173 Mossop B. 38, 52 Multiple translation 228 Muñoz Martín, R. 171, 180, 248 n.3 Naturalization 166 Negation (Antonomy) 90–1, 143, 215 Newmark, P. 165–8 Nida, E. A. 92, 124–6, 256 n.13 Nivelization (Levelling) 107, 110 norms xiv, 59, 120, 179, 233, 237, 250 n.6 oblique translation 139–40, 149, 248 n.3 Omission 39, 66, 72, 127, 146, 160, 196–8, 213–14, 232 Ortega y Gasset, J. 149 Osimo, B. 68, 253 nn.25, 26 Panneton, G. 22–3, 32 Paraphrase 92, 165, 167, 173 Particularization 28–9 performativity 9 Permutation 47, 65, 125, 159, 161 Perspective change 224, 241–2 politeness 196 political ideologies 35, 49–56, 75–7, 112–13, 117, 198–201, 243–4, 251 n.17 Popovič, A. 103, 109–12, 115, 256 n.2 post-structuralism 149

Index

281

procedures (procédés) 20–1, 157, 172, 205–6 purposes (цели) 44, 56 Pym, A. 9, 34

Störig, H. J. 60 stylistics 2–6, 140, 148–9 Substitution 39, 41, 65, 105, 110, 154 syntactics (Fedorov) 39

Quine, W. V. O. 123, 199, 257 n.4

Tao Youlan 97 tendencies (Levý) 106, 117 text types 43, 35, 54 Text-tailoring 231, 238 Torop, P. 68–70 totalitarian discourse 77–8 Toury, G. 59, 102, 120, 164, 179–80, 250 nn.6, 9 Transcription 84 transformational grammar 121–4, 148 transformations 123–31, 135, 148 translatability 56, 64, 114 translation 245 translation couplet 165–6, 189, 226–8 translation methods 157–8, 162 Translation Studies 60, 163–4, 182–3 Transliteration 46, 84, 105, 213 Transposition 8, 23, 25–6, 71, 141, 148, 159, 165, 172–5, 192, 211, 225–6 Tynyanov, Y. 38, 50, 250 n.10

realia 46, 105, 180, 231 Recategorization 172–3, 183 Reduction 28 Reformulation (Équivalence) 27, 30, 67, 173, 212, 230 regularity (rules) 43, 46, 59, 131, 134, 250 n.6 Reiss, K. 43 relevance 180 Repetition 81, 94–5, 214 Replacement 71, 125 Resegmentation 96, 166, 195, 227–8 Restructuring 47, 48, 65, 159, 211 Retsker, Y. 41, 42, 64, 155, 253 n.24 Revzin, I. I. 122, 127, 130 risk 237, 241–2 Robinson, D. xiv, 9 Rozentsveyg, V. I. 122, 127, 130 rubi 188–9, 227 rules/règles 3, 43, 46, 59, 129, 131, 134, 250 n.6 Sakai, N. 201, 261 n.9 Sakamoto, A. 186, 195 Saussure, F. de 2, 44 Scavée, P. 149 Scheinberg, R. 186, 191, 202 Schreiber, M. 157–62, 258 n.2, 259 n.5 script 187–9, 222 Séguinot, C. 206 Seidensticker, E. 192–4, 189–201, 261 n.8 Shveytser, A. D. 127 signs on roads 15–20 Skopos 44–5, 82, 155–7, 162, 167 Smirnov, A. A. 40–1, 47, 62 Sobolev, L. N. 44, 62, 156 solution types (as term) x Špirk, J. 112 Stalin, J. V. 52, 53 start text 247 n.1

universals of translation 106, 128 Van Niekerk, M. 237, 239 Vázquez-Ayora, G. 92, 126, 138, 258 n.1 Vermeer, H. J. 44, 156 Vinay, J. P. 17–8 Vinay, J. P. & Darbelnet, J. xii–xiv, 15–35, 154–5, 172–3, 25, 209, 212, 230 Vinogradov, V. 50–3, 251 n.14 violations (нарушения) 39 Vives, L. 247 n.3 Voloshinov, V. N. 251 n.14 Wilss, W. 64, 135, 154 work/effort 179, 182, 236, 240–1 Xuanzang 254 n.4 Yamada, M. 203 Yan Fu 78–9 Ye Zinan 96, 255 n.10