Language across Languages : New Perspectives on Translation [1 ed.] 9781443883115, 9781443877114

Citation preview

Language across Languages

Language across Languages: New Perspectives on Translation Edited by

Emanuele Miola and Paolo Ramat

Language across Languages: New Perspectives on Translation Edited by Emanuele Miola and Paolo Ramat This book first published 2015 Cambridge Scholars Publishing Lady Stephenson Library, Newcastle upon Tyne, NE6 2PA, UK British Library Cataloguing in Publication Data A catalogue record for this book is available from the British Library Copyright © 2015 by Emanuele Miola, Paolo Ramat and contributors All rights for this book reserved. No part of this book may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without the prior permission of the copyright owner. ISBN (10): 1-4438-7711-5 ISBN (13): 978-1-4438-7711-4

TABLE OF CONTENTS

Introduction ................................................................................................. 1 Emanuele Miola and Paolo Ramat Part I: Language across Space, Time and Culture When Rome Met Greece and When Canton, Beijing and Shanghai Met Western Cultures: Translations and… Lost in Translation ................ 11 Emanuele Banfi “Lost in Translation” between Typologically Different Grammars .......... 35 Alessio Muro Metalinguistic References in Cross-modal Translations: Sign Language Interpreting and its Issues ................................................. 59 Giulia Petitta Part II: Theories of Translation Translating Metaphors ............................................................................... 83 Michele Prandi Translation Strategy and the Constructed Reader: Italian Translations of Contemporary Irish Poetry .................................................................. 105 Debora Biancheri Mood and Modal Verbs in the English, German, and Greek Official Versions of the Charter of Fundamental Rights of the European Union ... 123 Ursula Stephany The Translation of Conversation and Film Dubbing as a Discovery Procedure: Evidence from Demonstratives ............................................. 143 Maria Pavesi

vi

Table of Contents

Interpretari, transferre, tra(ns)ducere, (con)vertere, imitari: Different Viewpoints on Translating ....................................................... 173 Paolo Ramat Contributors ............................................................................................. 187

INTRODUCTION EMANUELE MIOLA UNIVERSITY OF MILANO-BICOCCA

AND PAOLO RAMAT IUSS PAVIA AND UNIVERSITY OF PAVIA

1. Bringing Language across Languages: New challenges for translators and translation studies This volume collects most of the papers presented at a workshop on 'Translation' held in Pavia on October 3-4, 2013 and organized by the LETiSS (on which see below). The challenging theme yielded valuable contributions pointing up multiple aspects of this basic linguistic activity. Understandably, the two half-days of the international workshop were insufficient to provide an exhaustive overview of the issues connected with 'translation'. However, the editors are confident that this collection of papers will be of interest and of use to all those focussing on one or more aspects of the topic. For many cultures of the past, but also for contemporary Weltanschauungen, history is characterized by cyclicity. The same is true of language. From the internal-linguistic and typological perspective, scholars have identified many different kinds of cycle, notably the negation cycle, or negation spiral (Jespersen 1917, Bernini and Ramat 1996), the synthesis-analysis cycle (Schlegel 1818, Schwegler 1990, cp. Ledgeway 2012: 10-29), and the cycle of verbal functional load in the Germanic languages (see De Angelis and Di Giovine 2002, Ramat 1988: 191). Cycles or spiral movements may be viewed as characteristic not only of language but also of language studies. A kind of cyclicity is also to be observed in the sub-branches of linguistics—and, to come closer to our own specific interest in the present context, in sociolinguistics, that is to say, the linguistic approach to the very lives of people as speakers of one or more languages. For sure, we are currently experiencing dwindling times and, with regard to language use, a period of rapid change. Up to the

2

Introduction

mid-20th century, active knowledge of one language was predominant—at least in Western countries and multilingualism was even viewed as a sort of disease (see Weinreich 1953). When so-called globalization set out to conquer the world during the 1990s and early 21st century, the emergence of (American) English was thought to be the only possible outcome in the new economic and social panorama (see for example Crystal 2003). And indeed, English—at least in the Western world—is today the ‘lingua franca’ of politics, business, science, etc. Nonetheless, the change resembles a cycle, or, again, a spiral, insofar as multilingualism appears to be acquiring a new and more positive profile in line with the rise of new superpowers, such as Brasil, China, or Russia, and, consequently, with the strengthening of a positive attitude towards glocalization (i.e. globalization with localization, whereby global content is adapted to a specific, local culture) and so-called language ecology (see the earlier work of Haugen 1972; see also the papers in Valentini, Molinelli, Cuzzolin and Bernini 2003). The possible coming to the fore of regional ‘linguae francae’ (such as Chinese or German: see Janssens, Mamadouh and Marácz 2011) and the maintenance of a fruitful bi- or multilingualism seems to be what both linguists and laypeople should be aiming for. This is, for instance, the opinion of De Mauro (2014), who points out that in Europe there are already 103 national languages, all potentially used for political and high-domain communication. This situation could facilitate the borrowing of constructions and words from English, as well as the contribution of them to the supranational lingua franca. However, even if a ‘lingua franca’ such as English is required, there is and always will be a need for experienced, professional translators into and from English, in the interest of avoiding misunderstandings, grasping the ‘nuances’ of both the source and target languages, and capturing the very spirit (‘Geist’) of both the original and the translated text, be it literary, political, or scientific in nature. In this regard, specifically concerning European linguistic integration Jacqueline Visconti (2013), based on the studies of international institutions such as the Study Group on a European Civil Code (http://www. sgecc.net), has recently tackled the question of how a term used in a European Union (con)text relates to the corresponding terminology in a national (con)text. Adopting a ‘vergleichende [comparative] Textlinguistik’ approach in the multilingual EU context, she concentrates on the logicosemantic level of legal texts, with a special focus on connectives, such as notwithstanding or subject to, that play a crucial role in the interpretation of a text. She notes a huge lack of consistency in the translations of such

Emanuele Miola and Paolo Ramat

3

connectives and concludes that the European Court of Justice needs to resolve linguistic uncertainties in order to ensure the uniform interpretation that the law requires: the court must disambiguate and choose one interpretation, given that very different legal consequences would result if the ambiguity were not clarified (see Stephany, this volume, and Ramat, this volume, for other examples on this topic). On the other hand, large-scale comparisons of entire books—e.g. Le petit prince, or, for historical linguists, the Bible—are nowadays very popular among linguists thanks to computerized data banks. The results of such cross-linguistic—or, more accurately, cross-textual—comparisons go much further than ‘contrastive grammar’, traditionally the first step in contrastive linguistics, no longer being limited to lexical structures. The globalization of the mass media has speeded up the diffusion of English songs, books, TV programmes, movies and the like. Nevertheless, not everybody—even in developed countries—can read, or properly understand, English. By no means should these people be excluded from knowledge and fruition of the global information made available by the mass media. On the contrary, they should be enabled to enjoy knowledge and fruition of—among others—artistic contents in their own native languages. New practices of translation such as instant subtitling have already come into being with a view to making such contents accessible in local areas. Instant subtitling consists of subtitles released by professionals in order to make TV series and TV shows available to a broader fan base as soon as possible after initial release, via pay-per-view. Take for example Italy. In a nation with a strong tradition of dubbing and dubbing actors, Italian television networks such as Sky-TV have only introduced this practice relatively recently, and only because instant subtitling (fan-subtitling) was becoming widespread on the Internet. Notably, these Internet translations were an outstanding example of the socalled collaborative web: they were made available on the web free of cost, by non-professional fans, on a daily basis, and only a few days after the actual broadcasting of the shows in the US. On the one hand, sharing episodes of a series on-line without the copyright owner’s permission is not legal, but releasing and sharing a file with subtitles, provided that the language is different from the original one, might be legal. Thus, in order to ‘win the race’ against their on-line competitors, professional instant subtitles must be better than fan subtitles, and above all professional subtitlers must be faster that fan subtitlers, in order to make fan-sub addicts shift to pay-per-view (see Massidda 2013). New translation practices, such as the ‘instant translation’ of movies and Internet texts make translations unavoidable: it is evident that there

4

Introduction

will always be the need for translators from “global” English into other languages, at least national ones. In such a globalized world as ours, it is indeed striking that, although Arabic is a very popular and widely spoken language, there is still a “low volume of translations into Arabic, which had been identified as an obstacle to the dissemination of outside knowledge into the Arab world” (Ronen et al. 2014). Bearing this in mind, and to sum up, the challenge of “saying almost the same thing” (Eco 2003) is thus continuously renewed and taken up again and again.

2. A brief survey of the volume Many of these topics are touched on and assessed in the present volume, which revolves around two different, though interconnected, thematic nuclei. The first nucleus refers primarily to linguistic theory with a special focus on languages that are distant from contemporary Western culture in terms of both time and space. Emanuele Banfi’s article compares translation practices adopted in Ancient Rome in transferring Ancient Greek into Latin, with those of the Chinese world when it came, and comes, to translating Western concepts into Chinese. Alessio Muro also deals with so-called ‘exotic languages’ in an insightful typological study of grammatical anamorphism and grammatical differences in selected North American language varieties. He points up the grammaticalization of categories/functions such as ‘visibility’, which are completely absent from European languages. Giulia Petitta tackles translation practices applied to a less usual linguistic code: she is interested in a special kind of ‘intersemiotic translation’, namely in simultaneous interpretation from non-signed to signed languages and vice versa. Other authors decided to focus on theories of translation per se. Michele Prandi addresses the issue of metaphorical language by introducing and discussing the different kinds of consistent and conflictual metaphors, and their implications for translators and translation theory. By ‘consistent metaphors’ Prandi means metaphors that are well integrated into our ways of speaking and thinking, such as scientific revolution. In contrast, the label ‘conflictual metaphor’ (e.g. Winter pours its grief in snow) refers to expressions that strike us as unusual. Much in this vein, Debora Biancheri has contributed a paper on translation strategies and the ‘constructed reader’, an expression that refers to publishers’ and translators’ expectations regarding the readership profile. She exemplifies the much-debated question of poetic translations

Emanuele Miola and Paolo Ramat

5

with some Italian versions of the contemporary Irish poet Derek Mahon. Between the two poles of the ‘bella infedele’ (the beautiful but unfaithful) and the ‘brutta fedele’ (the faithful but ugly, scil. translation) she proposes a ‘third space’, a middle ground for the decodification of what may sound unfamiliar to the target audience. Ursula Stephany provides a clear example of recent advances in contrastive linguistics. Her paper is not limited to a comparison of lexical structures but extends to a global analysis of grammatical categories and their use. Stephany conducts an in-depth analysis of mood and modal verbs and convincingly points out that from a morphosyntactic point of view, even when translating within a shared cultural frame such as the European Union, special attention must be paid to the use and behaviour of grammatical categories viewed as a coherent set and not as the output of isolated items. In her contribution, Maria Pavesi emphasizes a pragmatic issue, namely the difficulty of translating conversations and other transitory language expressions, the speech acts that represent our everyday interactional linguistic behaviour: specifically, the phatic and conative aspects of communication are at risk of getting lost in translation. Pavesi, who has extensive experience in the field of dubbing for cinema and has created, together with her colleague Maria Freddi, the Corpus of Film Dialogue, a bilingual unidirectional parallel corpus of film transcriptions, focuses here on a particular morphological category, namely demonstrative pronoun—a universal feature of language whose pragmatic salience is self-evident. Her analysis enables us to identify key functional differences between source and target languages. Finally, Paolo Ramat’s article re-visits and summarizes all of the above-mentioned viewpoints on translation, while exploring the different words for ‘translator’ used in a range of (ancient and modern) languages, and reflecting different cultures.

3. Envoi History and life—we have said—are made up of cycles, and the present volume ends the cycle of the LETiSS (Lingue d’Europa: Tipologia, Storia, Sociolinguistica—Languages of Europe: Typology, History and Sociolinguistics, a research centre at the Istituto Universitario di Studi Superiori, IUSS, in Pavia) which in its half-decade of life was devoted to the study of the typology, history and sociolinguistics of the languages of Europe. Unfortunately, LETiSS was compelled to close down for economic reasons, and consequently to end its own life cycle.

6

Introduction

The present book also completes a notional trilogy investigating the problems of language decay and the outcomes of language contact (Miola and Ramat 2011), the new challenges posed to linguists by computermediated communication (Miola 2012), and finally the multifaceted topic of translation. As may be guessed from the titles of the three volumes, LETiSS’ attention was always directed towards the sociocultural aspects of language and the impact that these aspects have on general theories of language—via an inductive and reality-bound process. LETiSS must now pass the baton to other scholars and researchers, in the hopes that linguistic research may continue to act as a bridge among different cultures, different worlds and different Weltanschauungen, towards a better understanding of ourselves as human beings.

References Bernini, Giuliano and Paolo Ramat. 1996. Negative Sentences in the Languages of Europe: A Typological Approach. Berlin: Mouton de Gruyter. Crystal, David. 2003. English as a Global Language. 2nd edition. Cambridge: Cambridge University Press. Cuzzolin, Pierluigi, Ignazio Putzu and Paolo Ramat. 2006. “The IndoEuropean Adverb in diachronic and typological perspective.” Indogermanische Forschungen 111: 1-38. De Angelis, Alessandro and Paolo Di Giovine. 2002. “Il mutamento tipologico nella funzionalità dei morfemi verbali: le lingue germaniche e l’indo-iranico.” Paper presented at the congress Typological Change in the Morphosyntax of Indo-European Languages. Viterbo, January, 25-26, 2002. De Mauro, Tullio. 2014. In Europa son già 103. Roma-Bari: Laterza Eco, Umberto. 2003. Dire quasi la stessa cosa. Milano: Bompiani. Haugen, Einar I. 1972. The Ecology of Language. Stanfrod: Staford University Press. Janssens, Rudi, Virginie Mamadouh and László Marácz. 2011. “Languages of Regional Communication (Relan) in Europe: Three Case Studies and a Research Agenda.” In A Toolkit for Transnational Communication in Europe, ed. Jørgensen J. Normann, 69-102. Copenhagen: University of Copenhagen. Jespersen, Otto. 1917. Negation in English and other Languages. Copenhagen: Høst & Søn.

Emanuele Miola and Paolo Ramat

7

Kloss, Heinz. 1967. “‘Abstand languages’ and ‘Ausbau languages’.” Anthropological Linguistics 9 (7): 29-41. Ledgeway, Adam. 2012. From Latin to Romance: Morphosyntactic Typology and Change. Oxford: Oxford University Press. Massidda, S. 2013. The Italian fansubbing phenomenon. PhD Thesis, http://eprints.uniss.it/8183/ Miola, Emanuele and Paolo Ramat, eds. 2011. Language Contact and Language Decay. Pavia: IUSS Press. Miola, Emanuele, ed. 2012. Languages Go Web. Alessandria: dell’Orso. Ronen, Shahar, Bruno Gonçalves, Kevin Z. Hu, Alessandro Vespignani, Steven Pinker and César A. Hidalgo. 2014. “Links that speak: The global language network and its association with global fame.” Proceedings of the National Academy of Sciences. 201410931. Schlegel, August Wilhelm von. 1818. Observations sur la langue et la littérature provençales. Paris: Librairie grecque-latine-allemande. Schwegler, Armin. 1990. Analyticity and syntheticity: a diachronic perspective with special reference to Romance languages. Berlin-New York: Mouton de Gruyter. Valentini, Ada, Piera Molinelli, Pierloigi Cuzzolin and Giuliano Bernini, eds. 2003. Ecologia linguistica. Atti del XXXVI Congresso Internazionale di Studi, Bergamo 26-28 settembre 2002. Roma: Bulzoni. Visconti, Jacqueline. 2013. “European integration: connectives in EU legislation.” International Journal of Applied Linguistics 23: 44-59. Weinreich, Uriel. 1953. Languages in Contact: Findings and Problems. New York: Linguistic Circle of New York.

PART I: LANGUAGE ACROSS SPACE, TIME AND CULTURE

WHEN ROME MET GREECE AND WHEN CANTON, BEIJING AND SHANGHAI MET WESTERN CULTURES: TRANSLATIONS AND… LOST IN TRANSLATION EMANUELE BANFI UNIVERSITY OF MILANO-BICOCCA

In this paper, I examine two particular cultural and linguistic situations that are distant from each other in terms of time and space but share, so to speak, similar issues regarding how to translate foreign language texts. Specifically, after reviewing the scant attention paid by the Greek world to peoples speaking other languages (§ 1), I first focus on the problems faced by representatives of Roman and Latin culture when Rome encountered Greek culture and language between the 3rd and 1st centuries B.C. (§§ 1.1, 1.2, 1.3, 1.4). Second, I outline what occurred, in a broadly similar fashion, in the Chinese world when, in the mid-19th century, Chinese intellectuals of the late Qing dynasty encountered Western cultures and began translating mainly English (and French) books into Chinese (§§ 2, 2.1, 2.2).

1. Greek and Roman worlds and foreign languages Before considering the attitude of the Roman world towards the Greek culture and language, it is of interest to focus on the scant attention paid by the Ancient Greeks to peoples speaking other languages. The Greeks viewed these peoples as mere ȕȐȡȕĮȡȠȚ “barbarians”, and they neither had any linguistic politics, so to speak, nor felt the need to translate foreign texts into Greek. Indeed, they saw themselves and their own culture and language as “superior” and therefore as not requiring any “apport” from the outside (Horrocks 20102: 67). Furthermore, prior to the 5th century B.C., there is scant evidence of contact between the Greek world and other languages. Only two passages in the Iliad mention linguistic diversity among the Trojans’ allies:

12 (1)

Translations and… Lost in Translation Homer Il. 2. 803-805 ਯțIJȠȡ, ıȠ੿ į੻ ȝȐȜȚıIJૃ ਥʌȚIJȑȜȜȠȝĮȚ, ੰįİ į੻ ૧ȑȟĮȚ. ȆȠȜȜȠ੿ Ȗ੹ȡ țĮIJ੹ ਙıIJȣ ȝȑȖĮ ȆȡȚȐȝȠȣ ਥʌȓțȠȣȡȠȚ, ਙȜȜૉ įૃਙȜȜȦȞ ȖȜ૵ııĮ ʌȠȜȣıʌİȡȑȦȞ ਕȞșȡȫʌȦȞ. “Hector, to thee beyond all others do I give command, and thou even according to my word. Inasmuch as there are allies full many throughout the great city of Priam, and tongue differs from tongue among men that are scattered abroad” (Murray 198810: 1, 111).

(2)

Homer Il. 4. 436-438 ੮Ȣ ȉȡȫȦȞ ਕȜĮȜȘIJઁȢ ਕȞ੹ ıIJȡĮIJઁȞ İ੝ȡઃȞ ੑȡȫȡİȚ Ƞ੝ Ȗ੹ȡ ʌȐȞIJȦȞ ਷İȞ ੒ȝઁȢ șȡȩȠȢ Ƞ੝įૃ ੅Į ȖોȡȣȢ, ਕȜȜ੹ ȖȜ૵ııૃਥȝȑȝȚțIJȠ, ʌȠȜȪțȜȘIJȠȚ įૃ ਩ıĮȞ ਙȞįȡİȢ. “Even so arose the clamour of the Trojans throughout the wide host; for they had no all like speech or one language, but their tongues were mingled, and they were a folk summoned from many lands” (Murray 198810: 1, 185).

Herodotus (484-425 B.C.) was the first Greek author to manifest an interest in foreign languages and bilingualism. By his account, the Pharaoh Psammetichus instructed Ionians and Carians to teach Greek to young Egyptians who intended to become interpreters in Egypt. One of these read and translated for Herodotus in person a hieroglyphic inscription engraved on the walls of Cheop’s pyramid: (3)

Herod., 2.125.6 ıİıȒȝĮȞIJĮȚ į੻ įȚ੹ ȖȡĮȝȝȐIJȦȞ ਥȞ IJૌ ʌȣȡĮȝȓįȚ ੖ıĮ ਩Ȣ IJİ ıȣȡȝĮȓȘȞ țĮ੿ țȡȩȝȝȣĮ țĮ੿ ıțȩȡȠįĮ ਕȞĮȚıȚȝȫșȘ IJȠ૙ıȚ ਥȡȖĮȗȠȝȑȞȠȚıȚ. ȀĮ੿ ੪Ȣ ਥȝ੻ İ੣ ȝİȝȞોıșĮȚ IJ੹ ੒ ਦȡȝȘȞİȪȢ ȝȠȚ ਥʌȚȜİȖȩȝİȞȠȢ IJ੹ ȖȡȐȝȝĮIJĮ ਩ijȘ, ਦȟĮțȩıȚĮ țĮ੿ ȤȓȜȚĮ IJȐȜĮȞIJĮ ਕȡȖȣȡȓȠȣ IJİIJİȜȑıșĮȚ. “There are writings on the pyramid in Egyptian characters showing how much was spent on purges and onions and garlic for the workmen; and so far as I well remember, the interpreter when he read me the writing said that sixteen hundred talents of silver had been paid.” (Godley 19819: 1, 429).

Herodotus also alludes to contacts among various other languages, such as Lydian and Persian:

Emanuele Banfi (4)

13

Herod., 1.86 4.6 ȀĮ੿ IJઁȞ Ȁ૨ȡȠȞ ਕțȠȪıĮȞIJĮ țİȜİ૨ıĮȚ IJȠઃȢ ਦȡȝȘȞȑĮȢ ਥʌİȚȡȑıșĮȚ IJઁȞ ȀȡȠ૙ıȠȞ IJȓȞĮ IJȠ૨IJȠȞ ਥʌȚțĮȜȑȠȚIJȠ, țĮ੿ IJȠઃȢ ʌȡȠıİȜșȩȞIJĮȢ ਥʌİȚȡȦIJ઼Ȟ … “Cyrus heard it, and bade his interpreters ask Croesus who was this on whom he called; they came near and asked him ...” (Godley 19819: 1, 111).

or Greek and Persian: (5)

Herod., 3.38. 4 ǻĮȡİ૙ȠȢ į੻ ȝİIJ੹ IJĮ૨IJĮ țĮȜȑıĮȢ ੉Ȟį૵Ȟ IJȠઃȢ țĮȜİȠȝȑȞȠȣȢ ȀĮȜȜĮIJȓĮȢ, Ƞ੄ IJȠઃȢ ȖȠȞȑĮȢ țĮIJİıșȓȠȣıȚ, İ੅ȡİIJȠ, ʌĮȡİȩȞIJȦȞ IJ૵Ȟ ਬȜȜȒȞȦȞ țĮ੿ įȚૃ ਦȡȝȘȞȑȠȢ ȝĮȞșĮȞȩȞIJȦȞ IJ੹ ȜİȖȩȝİȞĮ… “Then Darius summoned those Indians who are called Callatiae, who eat their parents, and asked them, the Greeks being present and understanding by interpretation what was said…” (Godley 19819: 2, 51).

In his work we also find bilingual people, for example the Scythian king Scyles, born from a woman of Istria. His mother, who was presumably Ionian, taught him the Greek language and letters: (6)

Herod., 4.78.1 ਝȡȚĮʌİȓșİȧ Ȗ੹ȡ IJ૶ ȈțȣșȑȦȞ ȕĮıȚȜȑȚ ȖȓȞİIJĮȚ ȈțȪȜȘȢ ‫ ڶ‬ਥȟ ੉ıIJȡȚȘȞોȢ į੻ ȖȣȞĮȚțઁȢ Ƞ੣IJȠȢ ȖȓȞİIJĮȚ țĮ੿ Ƞ੝įĮȝ૵Ȣ ਥȖȤȦȡȓȘȢ. TઁȞ ਲ ȝȒIJȘȡ Į੢IJȘ ȖȜ૵ııȐȞ IJİ ਬȜȜȐįĮ țĮ੿ ȖȡȐȝȝĮIJĮ ਥįȓįĮȟİ … “Scyles was one of the sons born to Ariapithes, king of Scythia; but his mother was of Istria, and not nativeborn; and she taught him to speak and read Greek …” (Godley 19819: 2, 277).

Thucydides (460-404 B.C.) also provides evidence of the fact that the Persian language was known in Athens, via a reference to Artaphernes, who was sent to Sparta by the Great King and led to Athens as a prisoner in 424 B.C. The Athenians read the letters carried by Artaphernes after translating them from the Assyrian:

14 (7)

Translations and… Lost in Translation Thucid., 4.50.2. ȀĮ੿ Į੝IJȠ૨ țȠȝȚıșȑȞIJȠȢ Ƞੂ ਝșȘȞĮ૙ȠȚ IJ੹Ȣ ȝ੻Ȟ ਥʌȚıIJȠȜ੹Ȣ ȝİIJĮȖȡĮȥȐȝİȞȠȚ ਥț IJ૵Ȟ ਝııȣȡȓȦȞ ȖȡĮȝȝȐIJȦȞ ਕȞȑȖȞȦıĮȞ … “He was conveyed to Athens, and the Athenians caused his letters to be transcribed from the Assyrian characters and read them…” (Smith 19887: 2, 297).

Xenofon’s (430?-355 B.C.) Anabasis contains a number of references to bilingualism on the part of interpreters, especially between Persian and Greek: (8)

Xenoph. Anab. 2.3.17 ਫʌİ੿ į੻ ਕʌȒȞIJȘıĮȞ Į੝IJȠ૙Ȣ Ƞੂ IJ૵Ȟ ਬȜȜȒȞȦȞ ıIJȡĮIJȘȖȠȓ, ਩ȜİȖİ ʌȡ૵IJȠȢ ȉȚııĮijȑȡȞȘȢ įȚૃ ਦȡȝȘȞȑȦȢ IJȠȚȐįİ … “When the Greek generals met them, Tissaphernes, through an interpreter, began the speaking with the following words…” (Brownson 19927: 129).

In the Lives of Plutarch (46?-125? A.D.), among other attestations of bilingualism (Greek and Persian: Themistocles, 28.1; Greek and Latin: Sulla, 27.2 and Cato Maior, 12.5), we find a highly interesting mention of Queen Cleopatra’s multilinguistic abilities: (9)

Plut., Antonius 27.4 ਲįȠȞ੽ į੻ țĮ੿ ijșİȖȖȠȝȑȞȘȢ ਥʌોȞ IJ૶ ਵȤ૳. ȀĮ੿ IJ੽Ȟ ȖȜ૵IJIJĮȞ, ੮ıʌİȡ ੕ȡȖĮȞȩȞ IJȚ ʌȠȜȪȤȠȡįȠȞ, İ੝ʌİIJ૵Ȣ IJȡȑʌȠȣıĮ țĮșૃ ਴Ȟ ȕȠȪȜȠȚIJȠ įȚȐȜİțIJȠȞ ੑȜȓȖȠȚȢ ʌĮȞIJȐʌĮıȚ įȚૃ ਦȡȝȘȞȑȦȢ ਥȞİIJȪȖȤĮȞİ ȕĮȡȕȐȡȠȚȢ, IJȠ૙Ȣ į੻ ʌȜİȓıIJȠȚȢ Į੝IJ੽ įȚૃĮਫ਼IJોȢ ਕʌİįȓįȠȣ IJ੹Ȣ ਕʌȠțȡȓıİȚȢ, ȠੈȠȞ ǹੁșȓȠȥȚ, ȉȡȦȖȜȠįȪIJĮȚȢ, ਬȕȡĮȓȠȚȢ, ਡȡĮȥȚ, ȈȪȡȠȚȢ țĮ੿ ਙȜȜȦȞ ਥțȝĮșİ૙Ȟ ȖȜȫIJIJĮȢ, IJ૵Ȟ ʌȡઁ Į੝IJોȢ ȕĮıȚȜȑȦȞ Ƞ੝į੻ IJ੽Ȟ ǹੁȖȣʌIJȓĮȞ ਕȞĮıȤȠȝȑȞȦȞ ʌĮȡĮȜĮȕİ૙Ȟ įȚȐȜİțIJȠȞ, ਥȞȓȦȞ į੻ țĮ੿ IJઁ ȝĮțİįȠȞȓȗİȚȞ ਥțȜȚʌȩȞIJȦȞ. “There was sweetness also in the tones of her voice; and her tongue, like an instrument of many strings, she could readily turn to whatever language she pleased, so that in her interviews with barbarians she very seldom had need of an interpreter, but made her replies to most of them herself and unassisted, whether they were Ethiopians, Troglodytes, Hebrews, Arabians, Syrians, Medes or Parthians. Nay, it is said that she knew the speech of many other

Emanuele Banfi

15

peoples also, although the kings of Egypt before her had not even made an effort to learn the native language, and some actually gave up their Macedonian dialect” (Perrin 19885: 197).

1.1. When Rome met Greece Latin literature—as is well known—was a late phenomenon. Five centuries passed from the mythical foundation of Rome (8th century B.C.) until the need was felt to establish an artistic literature in Latin to compete with the Greek models. Nonetheless, in every period of Roman culture Greek literature must be viewed as a dominant influence on Roman writers: the mark of Hellenic thought and myth was ever-present in the Roman mind, and in many ways it is possible to speak of a “Greco-Roman tradition” in literature and the arts, given that the Romans fused everything they did with what the Greeks had done centuries before (Conte 1994). In the 2nd century B.C., we find a famous representative of traditional Roman culture, Cato Censor (234-139 B.C.), still protesting against the influx of – in his view – “debilitating” Greek, yet the presence in Rome of hundreds of Greek schoolmasters hired to teach the youth Greek, the language of high culture, made Greek a familiar part of Roman education: by the 1st century B.C., no educated person could afford to lack a good knowledge of Greek. Thus, Caesar and Cicero were among the flood of aspirants to a superior education who rushed to Athens to become educated and cultured… and Caesar, when stabbed to death in 44 B.C., did not utter the famous sentence “tu quoque Brute, fili mi” in Latin, but gasped out in informal Greek “ȀĮ੿ ıȪ İੇ ਥțİȓȞȦȞ, ੯ ʌĮ૙; (You are one of them, man?)”. By the middle of the 1st century B.C., Roman society had become bi-cultural and was to remain bi-cultural/multicultural later on, due to the vast extent of immigration from Greece and the Near East that took place under the Empire: in Rome, bi-culturalism and multiculturalism were never to be lost (Kaimio 1979; Adams 2002). Roman literature was made, not born. It was the first “derived” literature and its authors consciously viewed themselves as “indebted” to the tradition of another people, whom they acknowledged to be culturally superior. In thus differentiating itself from earlier traditions (von Albrecht 1997: 12), Roman literature found its own identity and specific selfawareness. In this regard, it paved the way for later European literature and became its model. In Rome, literary dependence (imitatio) did not have a bad reputation: artistic borrowing and transfer into a new context was not considered theft but a loan intended to be easily recognized as such. Creation of a new literary work was based on a sort of “competition” with a model, and the more significant the model, the greater the challenge

16

Translations and… Lost in Translation

and, in the case of success, the greater the emulation’s gain in artistic capability. A writer was expected to refer to a series of ancestors and, if needs be, to invent them. The principle and practice of imitatio produced an intellectual relationship binding author to author and period to period (Seele 1995).

1.2. Livius Andronicus and his translation of Homer’s Odyssey It is possible to precisely date the beginning of Roman literature. According to Cicero (Brutus, 72), in 240 B.C. during the Roman games celebrating Rome’s victory over Carthage, Livius Andronicus, a Greek poet from Tarentum, produced the first Latin drama. The same century had seen a series of key historical events: in 282 B.C. the former, highly cultured, Etruria had been vanquished; in 272 B.C. the city of Tarentum, an ancient Greek colony famous for its rich theatrical life, had been conquered; in 242 B.C., the First Punic War had been won. In that historical moment, Rome was the most powerful centre in the Western Mediterranean area: the Urbs possessed, for the first time in its history, both a unified territory and a new identity. For the first time ever, the whole peninsula took the name of its southern region: Italia, Italy. Rome’s growing power as a political centre was so strong that not only the Romans’ Italic kinsfolk, but also Greeks, began to write in Latin. One of the last mentioned was precisely Livius Andronicus. Titus Livius’ ab Urbe condita libri (7.2.8) tells us something of Livius Andronicus’ life. While Andronicus likely came to Rome as a prisoner of war, he possessed stage experience as actor. He was employed as a tutor by the influential Livii family, and we know that the Livii granted him his freedom. Livius Andronicus was the first author to write Latin drama: he transposed Greek structures into a society characterized by the mingling of Italic, Etruscan, and Hellenistic stage practices. He gave Latin titles to his comedies, which were based on Hellenistic models; his tragedies may have followed classical Greek dramatists. As regards epic poetry, Livius took Homer’s Odyssey as his model. The reason for this choice was twofold: first, the Homeric poem was considered part of early Italian history (some episodes of the Odyssey took place in Italy and Sicily); second, given the tradition of Hellenistic schools in which Homer was the key author of reference, Livius made Homer’s great poem accessible to the Latin public for both literary and broader cultural reasons. Homer was viewed as a wise man, a teacher and an educator: his works were the Bible. While a young Greek grew up with the Iliad and the Odyssey, after Livius Andronicus a young Roman grew up with Livius’

Emanuele Banfi

17

translation of the Homeric Odyssey, with Ennius and possibly with Virgil. It is evident that the Hellenized Roman élite of Livius’ time were likely to have read Homer in the original – given that Greek was the language of cultured Romans – but we know that Livius’ Odusia was successful as a school text and we also know from Horace that by the first century Roman schoolboys had trouble with Andronicus’s complex and archaic language: (10)

Horatius Flaccus, Ep. II, 1. 69-71 Non equidem insector delendave carmina Livi esse reor, memini quae plagosum mihi parvo Orbilium dictare… “I am not crying down the poems of Livius / I would not doom to destruction verses which I remember ‘Orbilius of the rod’ dictated to me as a boy…” [Fairclough 199113: 403].

The enterprise of translating Homer’s Odyssey into Latin was of major historical importance: in order to translate the Homeric hexameter, Livius adopted an Italic meter, the Saturnian and, because he did not have an epic tradition behind him, he attempted to confer solemnity upon his literary language by using – as Horace clearly recognized two centuries later – deliberately archaizing language: according to Horace the language chosen by Livius was “archaizing” right from in his own era. However, there is another factor that must be taken into account when analysing the translation technique adopted by Livius: for him, translating meant both preserving what could be assimilated and altering what appeared to be untranslatable, either because of the “poverty” of the linguistic instrument or because of differences in culture and mentality between the Greek and Roman worlds. Furthermore, Livius never indulged in arbitrary alterations of the original Homeric text, but was constantly guided by his own original and by his readership’s mental horizon. Thus, in choosing a “native” meter (the Saturnian) for his epic, he surely had his readers in mind. We know that Naevius was later to use the same meter and Ennius was the first to replace it with the hexameter. Let us examine Livius’ translation of the famous incipit of the Homeric Odyssey: (11)

Homer, Od. 1.1 ਡȞįȡĮ ȝȠȚ ਩ȞȞİʌİ ȂȠ૨ıĮ ʌȠȜȪIJȡȠʌȠȞ “Tell me, o Muse, of the cunning man” Livius 1: Virum mihi Camena insece versutum

18

Translations and… Lost in Translation

In translating this verse, Livius not only tried to maintain the Homeric word order but also used archaic forms such as insece “tell” to render Homer’s ਩ȞȞİʌİ (Aeolian form -ȞȞ- < *-ns-: *in-sekw-e); and Camena – the “ancient name of an Italic water divinity” – as equivalent to Homeric ȂȠ૨ıĮ, here relying on the contemporary etymology, according to which Camena came from Casmena/Carmena and thus from carmen “poem”. The first and last words were linked to one another through alliteration (Virum … versutum). The proper name Camena stood in the middle of the verse, creating a symmetrically balanced structure as demanded by Saturnian metre which generally comprised a “rising” and a “falling” half; in keeping with another rule of Saturnian verse, Livius opted to translate the Homeric adjective ʌȠȜȪIJȡȠʌȠȞ as versutum “cunning”, a derived form of the Latin noun versus in place of a compound adjective (bahuvrihi: as in the Greek ʌȠȜȪIJȡȠʌȠȞ). Livius simplified the Homeric expression ਪȡțȠȢ ੑįȩȞIJȦȞ (the barrier of teeth), adopting a–so to speak–“prosaic” solution: (12)

Homer, Od. 1.64 IJȑțȞȠȞ ਥȝȩȞ, ʌȠ૙ȩȞ ıİ ਩ʌȠȢ ij૨Ȗİ ਪȡțȠȢ ੑįȩȞIJȦȞ; “my child, what a word has escaped the barrier of your teeth?” Livius 3: mea puera quid verbi ex tuo ore supra fugit?

Thus, the phrase ਪȡțȠȢ ੑįȩȞIJȦȞ “barrier of (your) teeth”, which would have sounded very strange in Latin, was simply rendered as ex tuo ore “from your mouth”. Livius’ translation of șİȩijȚȞ ȝȒıIJȦȡ ਕIJȐȜĮȞIJȠȢ (“peer of the gods”), the Homeric expression describing Patroclus, significantly altered the “spirit” intended by Homer: (13)

Homer, Od. 3, 110 ਩ȞșĮ įİ ȆȐIJȡȠțȜȠȢ, șİȩijȚȞ ȝȒıIJȦȡ ਕIJȐȜĮȞIJȠȢ “there Patroclus, the peer of the gods as a councellor” Livius 10: ibidemque vir summus adprimus Patroclus.

Homer spoke of a hero “equal to the gods”, but such a notion was unacceptable to the Roman mentality. This explains Livius’ translation strategy: he modified the Homeric concept and, without any loss of poetic solemnity, translated șİȩijȚȞ ȝȒıIJȦȡ as summus adprimus “greatest and of first rank”. Livius replaced the common Homeric images Ȝ૨IJȠ ȖȠȪȞĮIJĮ țĮ੿ ijȓȜȠȞ ਸIJȦȡ (his knees and heart were loosed) with the impressive

Emanuele Banfi

19

phrase cor frixit prae pavore “(Odysseus’) heart was broken by the fear” (lit. “in front of the fear”): (14)

Homer, Od. 5, 297 țĮ੿ IJȩIJİ ੗įȣııોȠȢ Ȝ૨IJȠ ȖȠȪȞĮIJĮ țĮ੿ ijȓȜȠȞ ਸIJȦȡ “then were the knees of Odysseus loosened, and the heart within him” Livius 16: igitur demum Ulixi cor frixit prae pavore.

Livius translated the expression İ੝ȩʌȚįĮ țȠȪȡȘȞ (“nice-looking maiden”) as virgo (virgin), without rendering İ੝ȩʌȚįĮ “nice-looking”, a frequently recurring adjective in the language of Homer: (15)

Homer, Od. 6. 142-143 ੒ į੻ ȝİȡȝȒȡȚȟİȞ ੗įȣııİȪȢ ਲ਼ ȖȠȣȞ૵Ȟ ȜȓııȠȚIJȠ ȜĮȕ૵Ȟ İ੝ȩʌȚįĮ țȠȪȡȘȞ… “and Odysseus pondered whether he should clasp the knees of the fairfaced maiden…” Livius 17: utrum genua amplectens virginem oraret.

In another case, Andronicus interpreted rather than translated a verse of Homer’s describing a situation fraught with emotion and with irony. In Homer, the swineherd Eumaios speaks to the disguised Odysseus, and Odysseus is listening but not yet ready to reveal his identity. Eumaios says to Odysseus (calling him “son of Laertes”!) neque tamen te oblitus sum, Laertie noster (“I have not forgotten you, o son of Laertes”): this expression was more emphatic than the Homeric ੗įȣııોȠȢ ʌȩșȠȢ Į੅ȞȣIJĮȚ (“grief for Odysseus takes hold of me”): (16)

Homer, Od. 14, 144 ਕȜȜȐ ȝૃ ੗įȣııોȠȢ ʌȩșȠȢ Į੅ȞȣIJĮȚ ȠੁȤȠȝȑȞȠȚȠ “instead, it is longing for Odysseus, who is gone, that seizes me” Livius 18: neque tamen te oblitus sum, Laertie noster.

In Livius’ translation, the Greek goddess of fate, ȂȠ૙ȡĮ, became Morta (Ramat 1960a; Ramat 1960b), a choice criticized by Aulus Gellius who maintained that the more appropriate translation would have been Moeram… while the complex Homeric expression İੁȢ ੖IJİ (“until/when”) ȝȠ૙ȡૃ ੑȜȩȘ țĮșİȜૌıȚ IJĮȞȘȜȑȖİȠȢ șĮȞĮIJȠ૙Ƞ (“the cruel fate of pitiless death”) was simply translated as dies:

20 (17)

Translations and… Lost in Translation Homer, Od. 19, 144-145 … İੁȢ ੖IJİ țİȞ ȝȚȞ [sc. Laertes] ȝȠ૙ȡૃ ੑȜȩȘ țĮșİȜૌıȚ IJĮȞȘȜȑȖİȠȢ șĮȞĮIJȠ૙Ƞ “until the time when the cruel fate of pitiless death shall strike him down” Livius 11: quando dies adveniet quem profata Morta est.

1.3. Epicurus’ Ȇİȡ੿ ĭȪıİȦȢ and Lucretius’ De rerum natura Concerning the life of Titus Lucretius Caro, poet and philosopher (9450 B.C.), we know almost nothing apart from the odd fact that he may have died at the age of forty-four as the indirect result of having taken a “love potion” (Conte 1994, 155). He lived in politically troubled times in which the old traditional religion had largely declined, and for the first time the full force of the Greek philosophical tradition, particularly the work of Epicurus, was available to Romans. The title of Lucretius’ poem De rerum natura (On the nature of things) faithfully translates the title of Epicurus’ most important work Ȇİȡ੿ ĭȪıİȦȢ comprising thirty-seven books. From these a now lost ȂȚțȡ੽ ਫʌȚIJȠȝȒ and a ȂİȖȐȜȘ ਫʌȚIJȠȝȒ were derived. The latter was probably the outline chiefly followed by Lucretius. The date of the poem’s composition is not certain. In it, Lucretius appeals for Gaius Memmius not to abandon his efforts for the public good at a difficult moment for the country (De rerum natura, I, 41: hoc patriai tempore iniquo “in this time of our country’s troubles”): the entire first half of the century was ravaged by wars and Gaius Memmius was “praetor” in 58. There is a tendency to believe that the reference is to internal disagreements in the years after 59. However, earlier dates cannot be ruled out. Lucretius’ aim was to explain Epicurean philosophy to a Roman audience by means of a didactic poem comprising some 7,400 dactylic hexameters that was divided into six untitled books and explored Epicurean physics through richly poetic language and metaphors. Lucretius explained the principles of atomism, the nature of the mind and soul, the nature of sensation and thought, the birth of the world and its phenomena, both celestial and terrestrial. The subtitle of the poem was “Against superstition”, which is exactly what Lucretius understood “religion” to be: the mysteries that “bound back” the mind of men before Epicurus all disappeared when faced with thought, logical reason, and above all “science”. Lucretius displayed excellent knowledge of Greek literature: his poem contained many allusions to Homer, Plato, Aeschylus and Euripides, and he presented himself as the first poet to reach the

Emanuele Banfi

21

“trackless land of the Pierian Muses” in order to draw on a new source of poetry and win glory. In so doing, he reproduced the attitude of selfconsciousness that Callimachus had made a commonplace in Hellenistic poetry: (18)

Lucr., De rerum natura, I, 925-934/IV, 1-9 Avia Pieridum pearagro loca nullius ante trita solo. Iuvat integros accedere fontis atque haurire, iuvatque novos decerpere flores insignemque meo capiti petere inde coronam unde prius nulli velarint tempora Musae: primum quod magnis doceo de rebus et artis religionum animum nodis exsolvere pergo, deinde quod obscura de re tam lucida pango carmina, musaeo contingens cuncta lepore. “I traverse pathless tracts of the Pierides never yet trodden by any foot. I love to approach virgin springs and there to drink; I love to pluck new flowers, and to seek an illustrious chaplet for my head from fields whence before this the Muses have crowned the brows of none: first because my teaching is of high matters, and I proceed to unloose the mind from the knots of superstition ; next because the subject is so dark and the lines I write so clear, as I touch all with the Muses’ grace” (Rouse and Smith 199212: 77).

As to linguistic choices (McIntosh Snyder 1980; Dionigi 1988), Lucretius deplored the limited nature of his ancestral vocabulary (patrii sermonis egestas): (19)

Lucr., De rerum natura, I, 830-833 Nunc et Anaxagorae scrutemur homoeomerian quam Grai memorant nec nostra dicere lingua concedit nobis patrii sermonis egestas, sed tam ipsam rem facilest exponere verbis. “Now let us also examine the homoeomeria of Anaxagoras, as the Greeks call it, which cannot be named in our language because of the poverty of our mother speech, but yet it is easy to explain the thing itself in words” (Rouse and Smith 199212: 69).

Furthermore, given that certain philosophical concepts could not be expressed in Latin, in order to designate the notion of “atoms” (IJ੹ ਙIJȠȝĮ), Lucretius had to fall back on generic nouns such as semina “seeds”,

22

Translations and… Lost in Translation

primordia “primary things”, corpora prima “first bodies”; alternatively, in order to designate the notions of “homogeneity” (੒ȝȠİȠȝȑȡȚĮ) and “air” (ਕȒȡ), he supplied coinages of his own such as homoeomeria [De rerum natura, I, 830] and aer [De rerum natura, I, 1000]. In addition, in order to compensate for the egestas of his patrii sermonis, Lucretius drew on a large corpus of poetic words made available to him by the archaic tradition as well as on rhetoric strategies such as alliteration, assonance, archaic constructions, and in general the “sound effects” characterising the expressive-pathetic style of early Roman poetry. Above all, Lucretius used compound adjectives drawn from Ennius’ epic lexicon (De rerum natura, I, 117-118: Ennius ut noster cecinit, qui primus amoeno / detulit ex Helicone perenni fronde coronam “as our Ennius sang, who was the first to bring down / from lovely Helicon the crown of perennial leaves”) such as suaviloquens “sweet-speaking”, altivolans “high-flying”, navigerum “ship-carrying”, frugiferens “fruit-bearing”); or created adverbs of his own such as filatim “thread by thread”, moderatim “gradually”, praemetuenter “with anticipatory fear” and new periphrases based on Homeric models such as natura animi “soul” vs animus or equi vis “strength of the horse” vs equus.

1.4. Sappho’s ĭĮȓȞİIJĮȚ ȝȠȚ țોȞȠȢ ੅ıȠȢ… vs Catullus’ Ille mi par esse deo videtur… Another good example of the close relationship (and “competition”) between Greek poetic models and their translation into Latin is provided by the translation of an extremely well-known Sappho’s Ode by Gaius Valerius Catullus (84-54 B.C.), a famous Roman poet of the late Roman Republic who wrote in the Neoteric style, that is to say, in the manner of the so called “Poetae novi” (new poets) who flourished during the late Roman Republic. I first quote the text of Sappho’s Ode (with a “word for word” translation followed by both Arieti and Crossett’s and Barnstone’s poetic translations) and second its translation by Catullus followed by some linguistic considerations: (20) Sapph. Fr. 2 (handed on from the Ȇİȡ੿ ੢ȥȠȣȢ c. 10, 1-2): ĭĮȓȞİIJĮȚ ȝȠȚ țોȞȠȢ ੅ıȠȢ șȑȠȚıȚȞ seems to me he equal to gods ਩ȝȝİȞ’ ੭ȞȘȡ, ੕IJIJȚȢ ਥȞȐȞIJȚȩȢ IJȠȚ to be the man in front of you ੁıįȐȞİȚ țĮ੿ ʌȜȐıȚȠȞ ਛįȣ ijȦȞİȓis sitting and nearby (you) 4 ıĮȢ ੝ʌĮțȠȪİȚ sweetspeaking listens țĮ੿ ȖİȜĮȓıĮȢ ੁȝȑȡȠİȞ, IJȩ ȝ’ ਷ ȝ੹Ȟ and sweetly laughing which my heart in breast passionately excited țĮȡįȓĮȞ ਥȞ ıIJȒșİıȚȞ ਥʌIJȩĮȚıİȞ.. ੩Ȣ Ȗ੹ȡ ਥȢ ıૃ ੅įȦ ȕȡȠȤȑȦȢ ȝİ ijȫȞĮȢ as soon as I see you hardly to me sound 8

Emanuele Banfi Ƞ੣įİȞ ਩IJ İ੅țİȚ, ਕȜȜ੹ ț੹ȝ ȝ੻Ȟ ȖȜ૵ııĮ ਩ĮȖİ, ȜȑʌIJȠȞ į’ Į੡IJȚțĮ Ȥȡ૵Ț ʌ૨ȡ ੝ʌĮįİįȡȩȝĮțİȞ, ੑʌʌȐIJİııȚ į Ƞ੣įİȞ ੕ȡȘȝȝ’, ਥʌȚȡȡȩȝȕİȚıȚ į ਙțȠȣĮȚ, ਕ įȑ ȝ’ ੅įȡȦȢ țĮțȤȑİIJĮȚ, IJȡȩȝȠȢ į੻ ʌĮ૙ıĮȞ ਙȖȡİȚ, ȤȜȦȡȠIJȑȡĮ į੻ ʌȠȓĮȢ ਩ȝȝȚ, IJİșȞȐțȘȞ įૃ ੑȜȓȖȦ ૃʌȚįİȪȘȢ ijĮȓȞȠȝ’, AĮȜȜ. ਕȜȜ੹ ʌ઼Ȟ IJȩȜȝĮIJȠȞ, ਥʌİȓ … (versus 18-20 perierunt)

23

nothing still come but the tongue grows thick, thin at once skin flame runs down with eyes nothing I see, 12 ring ears sweat me pours down, trembling all (me) makes wild, greener than grass I am, to die little in want of 16 I seem, Agalli but everything must be undertaken, since…

(Lobel 1925: 16-17; Brunet 2001: 21-22; Pigeaud 2004: 114). (21)

Arieti and Crossett’s poetic translation: Barnstone’s poetic translation:

Seems to me that man to the gods is equal Who sits across from you near and hears

To me he seems equal to gods, that man who sits facing you and hears you near Your sweet voice. as you speak softly and laugh Laughter of love. ’Tis a cause to flutter in a sweet echo that jolts Heart within rib-cage; should I merely the heart in my ribs. Now Behold you, the voice within me sounds when I look at you a moment No longer. my voice is empty Yet, the tongue is broken; a gentle fire and can say nothing as my tongue Runs beneath my flesh in a rush; seeing cracks and slender fire races Leaves my eyes, my ears echo in a boom under my skin. My eyes are dead Of humming. to light, my ears Sweat upon me pours, as a tremble seizes me pound, sweat pours over me. All over, I seem wanner than the pale green grass, I convulse greener than grass To be near dying, lost in and feel my mind slip as I go A weakness. close to death, All must be endured, since as a wretch… yet I must suffer all, even poor… (Arieti and Crossett 1985: 66) (Barnstone 20104: 57-58). (22)

Catullus, LI

lle mi par esse deo videtur, ille, si fas est, superare divos, qui sedens adversus identidem te spectat et audit dulce ridentem, misero quod omnes eripit sensus mihi: nam simul te, Lesbia, aspexi, nihil est super mi

lingua sed torpet, tenuis sub artus

He seems to me to be equal to a god, he, if it may be, seems to surpass the very gods who sitting opposite you again and again 4 gazes at you and hears you sweetly laughing. Such a thing takes away all my sense, alas! For whenever I see you, Lesbia, at once no sound of voice remains 8 within my mouth, but my tongue falters,

24

Translations and… Lost in Translation

flamma demanat, sonitu suopte tintinant aures, gemina teguntur lumina nocte. otium, Catulle, tibi molestum est: otio exsultas nimiumque gestis: otium et reges prius et beatas perdidit urbes.

a subtle flame steals down through my limbs, my ears ring with inward humming, my eyes Are shrouded in twofold night. Idleness, Catullus, does you harm, you riot in your idleness and wanton too much. Idleness ere now has ruined Both kings and wealthy cities.

12

16

(Cornish, Postagate and Mackail 198818: 59-61)

Catullus, in line with the Neoteric style of poetry, wished to “compete” with the Greek model and, in so doing, “reproduced” Sappho’s stylistic moods: the well-known incipit of Sappho’s poem ĭĮȓȞİIJĮȚ ȝȠȚ țોȞȠȢ ੅ıȠȢ șȑȠȚıȚȞ ਩ȝȝİȞૃ ੭ȞȘȡ became “Ille mi par esse deo videtur (He seems to me to be equal to a god)”; the verses ੕IJIJȚȢ ਥȞȐȞIJȚȠȢ IJȠȚ / ੁıįȐȞİȚ țĮ੿ ʌȜȐıȚȠȞ ਛįȣ ijȦȞİȓ- / ıĮȢ ੝ʌĮțȠȪİȚ / țĮ੿ ȖİȜĮȓıĮȢ ੁȝȑȡȡȠİȞ were rendered as “qui sedens adversus identidem te / spectat et audit / dulce ridentem (who is sitting opposite watches and listens / to you again and again / sweetly laughing)”, etc. The only variatio in Catullus’ poetic translation concerned the verse ੑʌʌȐIJİııȚ įૃ Ƞ੣įİȞ ੕ȡȘȝȝȚ, which he translated as “gemina teguntur / lumina nocte (our lights (eyes) are covered / by twin night.)”: a good example of imitatio cum variatione, typical of Roman Neoteric poetry.

2. Late Qing’s China, an “isolated” world, and the Yi ⣟ “barbarians” Before dealing with the problems faced by Chinese intellectuals of the Qing dynasty (Ύᮅ.Qing Chao) in translating Western works in the the mid-19th century, it is important to remember that over the previou centuries China’s contacts with the West had had very little impact on the languages of the Chinese Empire. Indeed, prior to the 19th century, very few Chinese had undertaken any formal study of Western languages: the Chinese had always considered Westerners to be Yi ዀ “barbarians”, just like all the other populations of the Empire with whom they had come into contact over the centuries. The imperial court of Beijing was thousands of miles away from the coastal provinces: it did not need, fear, or even want to come into contact with foreigners. Chinese mandarins therefore took no interest in foreigners and in their countries of origin.

Emanuele Banfi

25

Because of the strong contempt for the Yì “barbarians”, the Qing prince Gong Wang ᜤ⋤ said that “no scholar would have lowered himself to studying the languages of the Yi ዀ” (Masini 1993: 32), hence contacts between Chinese and foreigners were mediated by local servants, mostly of Cantonese origin. Indeed, relations between the Chinese world and the Westerners were purely commercial and took place through a special class of Chinese interpreters: there was a special merchant guild – Gonghang බ ⾜ “official Corporation” – for this purpose, generally made up of about ten Chinese merchants. Among them were also the so-called shilaofu ᘧ⪁ ኵ (this word is the Chinese adaptation of the Arabic word sharaf “moneychanger”): the shilaofu were responsible for checking the value of the precious metals used for transactions. We know that the Cantonese were the first to come into direct contact with the Yì “barbarians”, and that when Westerners travelled to other ports/parts of China, they were accompanied by Cantonese interpreters speaking a sort of “Pidgin English”. Thus, over the decades, many words of English origin, related to products or objects previously unknown in China, became part of the local dialect of Canton: for example sanwenzhi ୕ ᩥ ἞ “sandwich” (cant. sammenji), bashi ᕮኈ “bus” (cant. baxi); doxi ከኈ “toast” (cant. doxi); zhisi அኈ “cheese” (cant. jisi), etc. As well as through merchants, basic linguistic contact between China and the West took place through Catholic and Protestant missionaries, the only Westerners to travel to China other than diplomatic envoys sent directly by sovereign states. At the beginning of the 19th century, in 1809 to be precise, Robert Morrison, a missionary with the London Missionary Society, was appointed interpreter to the East India Company office in Canton (Bertuccioli and Masini 1996, 221). In 1815, he moved to Malacca where he founded the Anglo-Chinese College (Yinghua Shuyuan ⱥ⳹᭩ 㝔) that came to be considered the best Chinese school in the Far East. Morrison was also responsible for the appearance of the Chinese Monthly Magazine (Chashisu Meiyue Tongjizhuan ᐹୡ಑ẗ᭶⤫グബ), the first foreign magazine to be published in China. With the support of the Royal Asiatic Society of London, he also published one of the first Chinese translations of the Bible (Shentian Shengshu ⚄ኳ䏆᭩). The early Catholic and Protestant missionaries were particularly prolific in publishing works in Chinese: of the 777 works published prior to 1867, 474 were texts of religious propaganda (excerpts of the Holy Scripture, catechisms, moral precepts, theological essays, etc.). Prior to 1854 only 13 works on geography, history and political economy had been published. Generally speaking, however, with the exception of

26

Translations and… Lost in Translation

geographical and historical works, Chinese scholars did not show much interest in the texts produced in Chinese by the missionaries. They were suspicious of the outside world and considered these works to be poorly written in any case (Masini 1993: 38). It is well-known that until the middle of the 19th century historical China was, essentially, an extremely “isolated” country (Bertuccioli and Masini 1996: 245-248; Gernet 2005: 307).

2.1. The Beijing Tongwenguan ྠᩥ㤋, “The College for Foreign Languages” and the Shanghai Jiangnan jiqi zhizao ju Ụ༡ᶵჾ〇㐀ᒁ “The Arsenal of Shanghai” The period of isolation of the Chinese world came to an end when foreign concessions were established in China’s main trading centres, as a consequence of treaties imposed by Great Britain, France and the United States. Foreign merchants and traders now had the incentive to settle in the concessions and in the newly opened ports: in 1861, the Zongli geguo shiwu yamen ⦻⌮ྛᅧ஦ົ⾦㛛 “The Office for the Affairs of All Countries”, the first modern Chinese ministry for foreign affairs, was set up in Beijing. Translation activities on a wider scale were only initiated after 1862, with the founding in Beijing of the Tongwenguan ྠᩥ㤋 “The College for Foreign Languages” which was also a research institute for the dissemination of Western knowledge in China. The Beijing Tongwenguan was mainly responsible for translating Western works on subjects hitherto unknown in China: about twenty Western books, mostly on the subject of international and foreign law, were translated at the Beijing Tongwenguan from its foundation up to 1898. One of the first works was Henry Wheaton’s Elements of International law (London-Philadelphia 1836) translated into Chinese in 1864 under the title Wanguo gongfa ⴙᅧබἲ (“The Public Law of Ten Thousand Countries”). In 1869, the Jiangnan jiqi zhizao ju Ụ ༡ ᶵ ჾ 〇 㐀 ᒁ (“The Arsenal of Shanghai”), another important institute for the dissemination of Western knowledge in China, was opened in Shanghai. In Beijing, the translators of the Tongwenguan mostly focused on legal and historical topics, whereas in Shanghai the translators of the Jiangnan jiqi zhizao ju Ụ༡ᶵჾ〇㐀ᒁ concentrated on technical and scientific works: they had grasped the importance of developing a Chinese scientific terminology both by coining new terms and drawing on traditional Chinese lexicon (Alleton 2001; Lippert 2001).

Emanuele Banfi

27

Although by the early 1880s a good deal of material was available on Western geography, technology and science, and to a lesser extent on history and law, Chinese scholars showed little interest in Western texts. Knowledge of “Western subjects” was not awarded points within the civil service system, and young people sitting state examinations did not have much interest in studying Western disciplines, as science was not included on the traditional curriculum. Things changed when in 1896 Liang Qichao ᱱ ྐྵ㉸ (1873-1929), a distinguished reformist scholar, edited in Shanghai the famous Xixue shumu biao す Ꮵ ᭩ ┠ ⾲ “Bibliography of Western Learning”, an annotated list of approximately 329 Chinese works on Western subjects (Masini 1993: 82; Bertuccioli and Masini 1996: 287). The list was accompanied by an article on how to approach the study of Western subjects entitled Du xixueshu fa ㆫすᏥ᭩ἲ “Method of reading Western learning”. The composition of the list clearly reflected the shift in interest. It was no longer exclusively technical but extended to other aspects of Western learning. Divided by subject area – technical (xue Ꮵ), political (zheng ᨻ), and religious (jiao ᩍ) – the list was the first bibliographical guide available in Chinese for the study of Western knowledge. Liang Qichao believed that a reform of China’s traditional education system was essential for political transformation. At that time, education was only for an élite and was based solely on knowledge of the Chinese classics. In reference to the fact that in the West, women, farmers, merchants and soldiers were literate, Liang Qichao suggested that schools be set up in China for the social classes that had traditionally been excluded from the study of the written language. He believed that in China literacy was greatly hindered by the fact that the written language was based on form rather than on sound as in Western languages. Liang Qichao appears to have advocated a system of phonetic transcription of characters in order to help children learn the Chinese language more easily. Above all, he believed that the study of Western law would facilitate reform, and that for this purpose it was imperative to be familiar with Western languages. He observed that as soon as relations had been established with China, Western missionaries immediately translated Chinese works into Latin, English and French. The Chinese had not reciprocated. Therefore, he held that Western languages should be taught to the Chinese from childhood, and that Western works of particular interest should be translated into Chinese without delay.

28

Translations and… Lost in Translation

2.2. How the Western notions of “liberty”, “democracy”, “president” were rendered in Chinese It is generally known that the Chinese language did not lend itself to the large-scale borrowing of terms from other languages. The reasons for this lie in the nature of the Chinese syllable structure and in the special features of the Chinese script (Novotná 1967; Novotná 1975; Godwin 1979). The Chinese language displayed (and continues to display…) a different method of assimilating Western concepts: specifically, it has tended to draw on native morphemes to build a descriptive replica of Western terms. In some instances, the morphemes of a Western model were translated on a one-to-one basis, giving rise to loan-translations. More commonly, however, the rendering was so free that we may speak of loan-creations induced by foreign concepts. Chinese intellectuals, because of the egestas of the Chinese language had numerous issues in translating Western notions into Chinese: they were faced with the same problems that Roman poets faced when they tried to translate Greek poetic texts into Latin. A review of the various replicas that were suggested as possible Chinese equivalents of the English words liberty, democracy, and president will shed some light on the way in which the originally Western notions they represent were received when first introduced into the Chinese lexicon in the late Qing period (Xiong 2001). In his Dictionary of the Chinese Language (1815-1823) Robert Morrison paraphrases “liberty” as the “principle of self-determination” and translates the Western notion into Chinese as Zizhu zhili ⮬୺அ⌮. In Walter H. Medhurst’s English and Chinese Dictionary (1847), “liberty” is translated by the Chinese term ⮬୺ Zizhu “self determination”. Walter H. Medhurst was the first to use the two characters ⮬⏤ Ziyou that were to become the standard translation of the term “liberty”. He had recourse to a bisyllabic word that appeared early in traditional Chinese texts, but the meaning of the traditional Chinese word ⮬ ⏤ Ziyou does not correspond at all to the modern Western concept of “liberty”, in the sense of the “liberty of individuals”. On the contrary, the Chinese term has a strongly negative connotation indicating a person “acting in a individual manner”, an attitude that is against the moral code of a Confucian society: the original meaning of Chinese Ziyou was more precisely something like “to behave in accordance with an excessively self-centred attitude”. This explains why the notion of “liberty” was not (and still is not!) greatly “appreciated” in Chinese culture… Indeed, all the sages in ancient Chinese history were

Emanuele Banfi

29

afraid of the theory of liberty and therefore never established it as a doctrine (Xiong 2001: 72). Robert Morrison, in his aforementioned Dictionary of the Chinese Language, states that democracy “is improper, since it is improper to be without a leader” (Xiong 2001: 73) and he paraphrases this statement in Chinese as ji buke wuren tongshu yi buke duoren luanguan ᪤୙ྍ↓ே⤫ ⋡஼୙ྍከே஠⟶ “if it is improper that nobody leads, it is equally improper that a multitude of people govern disorderly” (Xiong 2001: 73). Given that there was apparently no single term available in the Chinese lexicon to translate the word “democracy”, Robert Morrison was obliged to use a full sentence express his unfavourable view of the concept. Other negative connotations of the notion of “democracy” were featured in Walter H. Medhurst’s English and Chinese Dictionary, where we find the following explanations: Zhongren de guotong ╓ ே ⓗ ᅧ ⤫ “administration of the state by the multitude”, Zhongren de zhili ⤊ேⓗ἞ ⌮ “rule of the multitude”, Duoren luanguan ከ ே ஠ ᐁ “disorderly administration by many”, Xiaomin nongquan ᑠẸᘝḒ “abuse of power by worthless people”. In Chinese the original meaning of the word Minzhu Ẹ ୺ , used today to signify “democracy”, was “lord of the people” (Minzhizhu Ẹஅ୺), and only in the late Qing did the term come to be used as a translation for “democracy” or to denote a democratic politic system. The term Minzhu is morphologically identical to the “lord of the people” in ancient Chinese, but between the two words there is a key semantic difference given that the modern term simultaneously implies the principle of “rulership by the people”. This new usage of the ancient word Minzhu may be viewed as an ingenious application of the flexibility of the Chinese language in creating new words. The presidential system of democratic countries was unfamiliar to Chinese intellectuals of the Qing period and therefore a large number of different renderings and designations for the word “president” were suggested with a view to translating the Western notion of “president”: Touren 㢌ே “leader”, Zongli ⦻⌮ “superintendent”, Guozhu ᅧ୺ “lord of the state”, Qiu 㓂 “chief”, Bangzhang 㑥㛗 “leader of the country”, Tongling ⤫ 㡿 “commander”, Zongtong ⦻ ⤫ “commander-general”. Zongtong is an old Chinese word with two meanings: the first is the verbal meaning of “to manage” or “to control”, while the second meaning comes from the military realm and denotes the commander-general of a brigade. The modern meaning of the word was obviously derived from both these ancient meanings, via transformation of the verb into a noun and abbreviation of the expression Zong tongling ⦻ ⤫ 㡿 “commander-

30

Translations and… Lost in Translation

general”. Another word used in the Qing era to translate the notion of “president” was Shouling 㤳 㡿 “leader”, but this word designated a particular leader, namely the “chief eunuch”. Thus, in 1879, a foreigner residing in Shanghai stated in a letter to the editor of the best-known English newspaper, the North China Daily News, that the Chinese word Shouling implied at best the captain of a gunboat, but more commonly referred to the head of a gang of robbers. Further, he wrote that it had been a Western or, more precisely, an American missionary who had first used this term as a translation of the English “president”, a suggestion which he – the reader – considered ridiculous and offendsive… In the Jiaohui xinbao ᩍ᭳᪂ሗ (Church News) and the Wanguo gongbao ⴙᅧබሗ (News of Ten Thousand Countries) during the 1860s and 1870s, the notion of “president” was rendered by the terms Huangdi ⓚ ᖇ “emperor”, Guojun ᅧྩ “monarch” or Guohuang ᅧⓚ “emperor of the state”. When the American president Ulysses Simpson Grant (1822-1885) visited Shanghai in May 1879, he was warmly received and, in order to express their respect, the Chinese referred to him as Guohuang ᅧⓚ “emperor of the state” or simply Huangdi ⓚᖇ “emperor”. It is interesting to note that in the Chinese texts containing the word Huangdi “emperor”, two empty spaces were inserted before the word as was customary in Chinese documents as an expression of respect. Finally, it is also worth recalling that, in order to translate the word “president”, the earlier-mentioned traditional and ancient Chinese words were preferred to proposed phonemic loanwords such as bolixidun ఑⌮႐㡻 or bolixitiande ఑⌮ታ ኳᚫ (< Engl. politician) or qian bolixitiande ๓఑⌮ታኳᚫ “former (qian ๓) president, that is to say, politician”. Ultimately, Chinese forms such as bolixidun ఑⌮႐㡻, bolixitiande ఑⌮ታኳᚫ were not a good choice for translating the English word “president” given that “a politician” is not necessarily “a president”…

3. Some final remarks Translating is both inborn in the very nature of human language and ancient and omnipresent within the history of human species and cultures (De Mauro 1994; Silvestri 2002; Ramat, this vol.). All languages feature a series of “semantic primes”, that is to say, concepts that cannot be explained in terms of other concepts and are therefore common to all languages (Wierbizcka 1996). Furthermore, Claude Hagège argues that “[…] au prix de quelque infidélité, on peut [...] traduire de n’importe

Emanuele Banfi

31

quelle langue en n’importe quelle autre, et quel que soit le type de texte, c’est bien parce qu’un fonds commun les relie toutes” (Hagège 1987 : 34). But, contrary to Hagège’s optimistic opinion, it is more appropriate to acknowledge that the operation of translating may be situated along a sort of “scale”, which starts from the mere literal expression of some denotata (“Bezeichnung” in Coseriu’s terms) and ends, at the opposite pole, with the expression of inferential connotation (“Bedeutung”, in Coseriu’s terms): Bernard Comrie (Comrie 1989) speaks in this regard of a “weak version of translatability”, which is always possible inasmuch it is only based on the denotata of a text or of a word. However, in the case of either “complicated” texts or culturally “sophisticated” words, a “good” translation seems to be unattainable. Paolo Ramat states that “translating means to transfer the contents of one and the same text in the phonological, morphological, syntactic and semantic rules of the target language, in order to get the same total effect”. But he also acutely and rightly adds that “the operation of translating is even more complex than this” (Ramat 2007: 21). In comparing the translation strategies adopted in two very different linguistic and cultural environments, we have noted that the cultural affinity between the Greek and Latin worlds certainly “facilitated” the production of Roman translations capturing the connotations of the original Greek texts and words, while, at the same time, we have observed the difficulties encountered by Roman poets in rendering Greek models. Equally in late Qing China, when the country opened up to Western nations in the mid-19th century, we have looked at the great difficulty associated with rendering, in Chinese, Western concepts and notions that did not feature in traditional Chinese culture. Both in ancient Rome and in 19th century Canton, Beijing and Shanghai – as well today in other cultural contexts – translating mostly meant (and means) “losing” much of the original texts and words, so that much of what was (and is) translated… was (and is)… lost in translation… To conclude, I think that Michele Prandi is right when, in relation to the issue of translating metaphors documented in the lexical structure of a language, he claims that this operation “is a specific case of the more general question of anisomorphism [… pertinent] to the fact that meanings have different forms in different languages [… and consequently that] it is not always possible to directly compare a meaning in the source language and a meaning in the target language, and to find an immediate equivalent in the target language” (Prandi, this vol.).

32

Translations and… Lost in Translation

Acknowledgments I am very grateful to my colleagues Hermann W. Haller and Elisabetta Jezek for revising the English expression.

References Adams, James N. 2002. Bilingualism and the Latin Language. Oxford: Oxford University Press. Alleton, Viviane. 2001. “Chinese Terminologies: on Preconceptions.” In New Terms for New Ideas. Western Knowledge and Lexical Change in the Late Imperial China, ed. Michael Lackner, Iwo Amelung and Joachim Kurtz, 15-34. Leiden-Boston: Brill. Arieti, James A. and Crossett, John M. 1985. Longinus ‘On the Sublime’. New York-Toronto: The Edwin Mellen Press. Barnstone, Willis. 20104. Ancient Greek Lyrics translated and annotated by W. Barnstone. Bloomington-Indianapolis: Indiana University Press. Bertuccioli, Giuliano and Federico Masini. 1996. Italia e Cina. RomaBari: Laterza. Brownson, Carleton L. 19927. Xenophon. Anabasis. Cambridge, Mass.London: Harvard University Press-William Heinemann. Brunet, Philippe. 2001. Sappho. L’Égal des dieux. Paris: Éditions Allia. Comrie, Bernard. 1989. “Translatability and language universals.” Belgian Journal of Linguistics 4: 53-67. Conte Gian Biagio. 1994. Latin Literature. A History. Baltimore-London: The John Hopkins University Press. Cornish Francis W., Postgate John Percival and Mackail John William. 198818. Catullus, Tibullus, Pervigilium Veneris. Cambridge Mass.London: Harvard University Press-William Heinemann. De Mauro, Tullio. 1994. Capire le parole. Bari: Laterza. Dionigi, Ivano. 1988. Lucrezio. Le parole e le cose. Bologna: Pàtron. Fairclough, Henry R. 199113. Horace. Satires, Epistles and Ars Poetica. Cambridge Mass.-London : Harvard University Press-William Heinemann. Gernet, Jacques. 2005. Le monde chinois. 2. L’époque moderne (Xe siècleXIXe siècle). Paris: Colin. Godley Alfred Denis. 19819. Herodotus Historiae. Cambridge, Mass.London: Harvard University Press-William Heinemann. Godwin Charles D. 1979. “Writing Foreign Terms in Chinese.” Journal of Chinese Linguistics 7 (2): 246-77.

Emanuele Banfi

33

Hagège, Claude. 1987. “La traduction, le linguiste et la rencontre des cultures.” Diogène 137: 24-34. Horrocks, Geoffrey C. 20102. Greek: a History of the Language and its Speakers. Maiden, Mass.-Oxford-Chichester: Wiley-Blackwell. Kaimio, Jorma. 1979. The Romans and the Greek Language. Helsinki: Societas Scientiarum Fennica. Lippert, Wolfgang. 2001. “Language in the Modernization Process: the Integration of Western Concepts and Terms into Chinese and Japanese in the Nineteenth Century.” In New Terms for New Ideas. Western Knowledge and Lexical Change in the Late Imperial China, ed. Michael Lackner, Iwo Amelung and Joachim Kurtz, 57-66. LeidenBoston: Brill. Lobel, Edgar. 1925. ȈĮʌijȠࠎȢ ȂȑȜȘ. The fragments of the lyrical poems of Sappho. Oxford: Clarendon Press. Masini, Federico. 1993. The Formation of Modern Chinese Lexicon and its Evolution towards a National Language. The Period from 1840 to 1898. Berkeley: Journal of Chinese Linguistics-Monograph Series n. 6. McIntosh Snyder, Jane. 1980. Puns and Poetry in Lucretius’ De Rerum Natura. Amsterdam: Grüner Publishing. Murray, Augustus Taber. 198810. Homer. The Iliad. Cambridge, Mass.London: Harvard University Press-William Heinemann. Novotná, Zdenka.1967. “Linguistic Factors of Low Adaptability of loanwords in the Lexical System of Modern Chinese.” Monumenta Serica 26: 103-18. —.1975. “Morphemic Reproductions of Foreign Lexical Models in Modern Chinese.” Archiv Orientálni 43: 146-71. Perrin Bernadotte. 19885. Plutarch’s Lives. Demetrius and Antony, Pyrrhus and Caius Marius. Cambridge, Mass.-London: Harvard University Press-William Heinemann. Pigeaud Jackie. 2004. Sappho. Poèmes. Paris: Rivage. Ramat, Paolo. 1960a. “L’etimologia del nome latino Morta”. Archivio Glottologico Italiano 45: 61-7. —. 1960b. “La figura di Moira in Omero alla luce dell’analisi linguistica”. Studi Italiani di Filologia Classica 32: 215-48. —. 2007. “The problem of (un)translatability”. Rassegna Italiana di Linguistica Applicata 39: 21-32. Rouse, William Henry Denham & Smith, Martin Ferguson. 199212. Lucretius de Rerum Natura. Cambridge, Mass.-London: Harvard University Press-William Heinemann.

34

Translations and… Lost in Translation

Smith, Charles Forster. 19887. Thucydides. History of the peloponnesian war. Cambridge, Mass.-London: Harvard University Press-William Heinemann. Seele, Astrid. 1995. Römische Übersetzer. Nöte, Freiheiten, Absichten. Verfahren des literarischen Übersetzens in der griechisch-römischen Antike, Darmastd: Wissenschaftliche Buchgesellschaft. Silvestri, Domenico. 2002. “Linguistica ‘contestuale’ e traduzione come operazione interlinguistica.” Rivista Italiana di Linguistica e Dialettologia 2: 89-102. von Albrecht, Michael. 1997. A History of Roman Literature. I: from Livius Andronicus to Boethius. Leiden: Brill. Wierzbicka, Anna. 1996. Semantics: Primes and Universals. Oxford-New York: Oxford University Press. Xiong, Yuezhi. 2001: “Liberty, Democracy, President: the Translation and Usage of some Political Terms in Late Qing China.” In New Terms for New Ideas. Western Knowledge and Lexical Change in the Late Imperial China, ed. Michael Lackner, Iwo Amelung and Joachim Kurtz, 68-89. Leiden-Boston: Brill.

“LOST IN TRANSLATION” BETWEEN TYPOLOGICALLY DIFFERENT GRAMMARS ALESSIO MURO UNIVERSITY OF PADUA

Grammatical (morphosyntactic) anisomorphism is, together with lexical anisomorphism, a key factor which must be dealt with in describing, learning and teaching languages. Beside being one of the main interests of linguistic theory, it also represents one of the main difficulties translators must face. However, notwithstanding the importance of the phenomenon, the attention paid to cross-linguistic variation in the field of morphosyntax is rarely complemented with considerations regarding the impact of such variation on the translation process. This article aims to partly fill this gap starting from a classical illustration of grammatical anisomorphism, presented by Franz Boas in the Introduction to the Handbook of American Indian Languages (1911). The grammatical categories described therein will first be exemplified in the light of recent linguistic research (Section 2). After this (Section 3), their phenomenology will be studied in texts of two genetically unrelated and typologically divergent North American languages (Oowekyala and Oneida). The article offers preliminary generalizations about how the problems posed by the morphosyntactic properties of these languages are dealt with in translating texts into English.

1. Introduction The question of the implications of linguistic diversity has been of central interest to Western thought at least since the times of Wilhelm von Humboldt (Brown 1967); the most debated domain within the topic is certainly linguistic relativism, which found its most influential enunciation in the so-called Sapir-Whorf hypothesis. According to the most quoted variant of such proposal, language may to some extent

36

“Lost in Translation” between Typologically Different Grammars

influence a person’s thought and world view.1 Preliminary to this debate, however, was a large-scale phase of language documentation, mainly carried out by Franz Boas and his cooperators, and mainly concerned with the languages of North America. This endeavour resulted in the publication of the monumental Handbook of American Indian Languages (4 vols., 1911-1941). Boas’ Introduction to the first volume is still considered a milestone in the history of linguistics: it contains clear exemplifications of linguistic diversity both in the field of the lexicon (e.g. it introduces the still highly controversial problem of the Eskimo words for ‘snow’) and in the field of morphosyntax. In the light of contemporary linguistic theory, it is important to distinguish lexical from grammatical anisomorphism: in this work I will be concerned with the latter. The aim of the present paper is twofold: first, I will attempt a reconstruction of the grammatical patterns mentioned by Boas, but not exemplified with concrete linguistic material; second, I will provide a synthetic survey of selected grammatical patterns found in texts of two North American languages, both very familiar to Boas himself (Oowekyala, Upper North Wakashan; Oneida, Northern Iroquoian), paying special attention to the consequences of their morphosyntactic diversity when translating texts from these languages into English. The languages are extremely divergent grammatically and unrelated genetically. 2 From the analysis it should emerge clearly that Boas’ insight into grammatical anisomorphism still has a lot to offer to both linguistic theory and translation studies.

2. Boas and grammatical anisomorphism The first anisomorphous category considered by Boas is tense. The author writes (Boas 1911: 42): In the sentence The man is sick we really express the idea, The single definite man is sick at the present time. This strict expression of the time relation of the occurrence is missing in many languages. The Eskimo, for instance, in expressing the same idea, will simply say, single man sick, 1

For a concise introduction to the problem of linguistic relativism and a state of the art in this research domain see a.o. Beek (2006); McWhorter (2014). Chafe (1998) also explores the issue, focalizing on Seneca (Northern Iroquoian). 2 Greenberg and Ruhlen (2007: 6) posit the existence of an “Almosan-Keresiouan” macro-family which would include both languages. The (putative) evidence they present is predominantly lexical; there is practically no unambiguous grammatical evidence that could justify postulating a relationship.

Alessio Muro

37

leaving the question entirely open whether the man was sick at a previous time, or is going to be sick in the future.

Whether Eskimo languages can be considered tenseless is still a matter of dispute among specialists (Swift 2004 argues for tense as an active category in Inuktitut, and so does Fortescue 1984 for West Greenlandic; the latter language is conversely described as tenseless by Shaer 2003 and most recently Bittner 2014). The pattern referred to by Boas can be exemplified by the West Greenlandic example (1) (from Berthelsen et al. 1997: 198, my glosses):3 (1)

napparpoq nappar -puq become.sick -3SG.IND.S ‘S/he became sick.’

The root nappar- is inherently telic, and in the indicative mood it can be understood by default as indicating a relative past (a default present interpretation is instead reserved to atelic predicates). The ingressive meaning of the root could justify the reading purported by Boas as future, but future tense is actually obligatorily marked by overt morphology in 3

In the examples, a separate line of morphological analysis is provided whenever the surface representation of one or more morphemes differs significantly from their inherent phonological shape. The following abbreviations are used in morpheme glosses: 1,2,3 – persons of verb agreement and deictic degrees (resp. proximal, medial, distal; when the two features co-occur, they are linked with ‘>’, e.g. 3>3S – 3rd distal subject); A – agent; AB – absentive (suffix and deictic degree); ACC – accusative; AH – addressee honorific; AN – animate; CIS – cislocative; CONTR – contrastive; DCS – decessive; DECL – declarative; DEF – definite; DEM – demonstrative; DET – determiner; DIST – distributive; DU – dualic; EG – epenthetic glide; EV – epenthetic vowel; EXCL – exclusive; F – feminine-indefinite; FACT – factual mood; FOC – focalizer; HAB – habitual; HRS – hearsay; INCL – inclusive; IND – indicative; ING – ingressive; INS – instrumental applicative; INV – invisible; IPFV – imperfective; IRR – irrealis mood; ITER – iterative; LNK – linker between components of a noun incorporation construction (Oneida); M – masculine; MOV – moving; N – neuter; NFS – noun forming suffix; NMLZ – nominalizer; NMOV – nonmoving; NOM – nominative; P – patient (in Oowekyala also complement of a preposition); PART – partitive; PASS – passive; PFV – perfective; PL – plural; PROX – proximate; Ps – possessor; PST – past; RDET – restricting determiner; REL – relativizer; S – intransitive subject (for languages without split intransitivity: Oowekyala, Korean); SDET – saturating determiner; SG – singular; SH – subject honorific; SRF – semireflexive; TOP – topic; VIS – visible.

38

“Lost in Translation” between Typologically Different Grammars

Eskimo languages.4 For the purposes of the present paper, we can safely assume a basic non-future vs. future tense distinction for Greenlandic (as well as for the North Wakashan languages to be discussed next). The Introduction goes on to illustrate the very rich article system of Kwakw’ala (called by Boas Kwakiutl, Lower North Wakashan). Boas (1911: 43) writes: In Kwakiutl this sentence [i.e. “The man is sick”, AM] would have to be rendered by an expression which would mean, in the vaguest possible form that could be given to it, definite man near him invisible sick near him invisible. Visibility and nearness to the first or second person might, of course, have been selected in our example in place of invisibility and nearness to the third person.

In this case two different grammatical domains are at issue: pronominal enclitics on verbs and articles on nouns. Like demonstratives, these two categories display a three-way system of deictic degrees (with proximal, medial and distal values); most unusually, they even qualify their referent in terms of visibility (but not number).5 The visible vs. invisible dichotomy applies to each of the three deictic degrees, yielding six possible combinations (i.e. determiners can be proximal and visible, proximal and invisible, etc.). The pattern can be observed in the following example (Anderson 1992: 28, glosses adapted):6,7

4

I am grateful to Michael Fortescue for clarifying this point to me. As Bach (2006) notes, the visibility features on articles are likely to constitute a nominal equivalent of the verbal category of evidentiality. 6 The Kwakw’ala and Oowekyala examples are given in the americanist transcription used in Howe (2000). Epenthetic schwas have been inserted where appropriate. In addition, for Oowekyala, word stress is indicated by means of accents: acute for primary and grave for secondary stress, as in Hilton and Rath (1982). The epiglottal plosive symbol ʡ is used to indicate a glottal stop with a lowering effect on the following vowel, as in Flynn and Smith (in progress). 7 In the glosses I depart from Anderson’s analysis of -ida as a single-morpheme determiner and prefer Boas’ (1947) original analysis in terms of a subject clitic (-i) plus a nominal determiner (-da) which, like the Persian ezāfe construction, is linked to the following noun phrase syntactically, but to the preceding verb phonologically. 5

Alessio Muro ə

(2)

39

w

əg anəm w

-ə (a) -i -da bəg anəm cover.with.blanket -on.cheek -3>3S -DEF.NOM man.3VIS.DET ‘The man (distal, visible) covered his cheek with a blanket.’

According to Boas (1947: 252), in order to denote an invisible subject, it would be sufficient to use the determiner -a, leaving the verbal complex unchanged (the expected form for “definite man near him invisible” w would thus be bəg anəma). Typologically, apart from visibility, it is also unusual to find articles distinguishing deictic degrees the way demonstratives do (even though the latter often provide the source from which articles are derived). Still with regard to article systems, Boas goes on to describe a different situation for the Siouan language Ponca: In Ponca, one of the Siouan dialects, the same idea [i.e. “The man is sick”, AM] would require a decision of the question whether the man is at rest or moving, and we might have a form like the moving single man sick.

The author here mentions two features of Omaha-Ponca articles, i.e. number (“single”) and movement status (“at rest” vs. “moving”); however, from Eschenberg’s recent study of the Omaha-Ponca article system (Eschenberg 2005) we learn that things are much more complex, as the additional features of obviation and animacy are distinguished. 8 The following examples illustrate the cases hinted at by Boas (I give the proximate forms only) (Eschenberg 2005: 14, 77): (3)

h

a.

nu -ak a nushiaha man-AN.PROX.NMOV.SG.DET be.short ‘The man (at rest) is short.’

b.

Thomas-ama wasnide tama T. -AN.PROX.MOV.SG.DET be.late IRR ‘Thomas (in motion, in a rush to get ready) will be late.’

This article system is profoundly different from that of Kwakw’ala: the grammaticalization paths which are likely to have generated the two 8

Obviation is a differential marking system for third person arguments on the basis of their discourse status; it allows to distinguish more prominent arguments, called proximates, and less prominent ones, called obviatives.

40

“Lost in Translation” between Typologically Different Grammars

systems are also different (articles are assumed to derive from demonstratives in Kwakw’ala, but from verbs in Omaha-Ponca, with a typologically very unusual grammaticalization path; such verbs also grammaticalize into evidentials). The last arguments exposed by Boas on page 43 seem to coalesce different categories under the name of “modalities”: “If we take into consideration further traits of idiomatic expressions, this example might be further expanded by adding modalities of the verb; thus the Kwakiutl, whose language I have used several times as an example, would require a form indicating whether this is a new subject introduced in conversation or not [i.e. topicality, expressed, if I understand correctly, by auxiliary verb constructions, AM]; and, in case the speaker had not seen the sick person himself, he would have to express whether he knows by hearsay or by evidence that the person is sick, or whether he has dreamed it [these words clearly describe the concept of evidentiality, AM].”

Since the two categories of topicality and evidentiality emerge clearly in the examples I selected from my corpus study of the closely related Oowekyala language, a discussion of such categories will be provided in the next section (3.1). To sum up, in Boas’ (1911) formulation, several functional categories were identified as sources of grammatical variation across languages. They are reported in Table 1 (‘yes’ indicates that a category is obligatorily coded, and ‘no’ that it is absent from the morphological system of the language; some categories are indirectly coded or of restricted occurrence, as specified in the table): Table 1. Anisomorphous categories in Boas’ (1911) sample languages Category\Language Temporal location Definiteness Deictic degree

Eskimo Yes No Yes (in DEM)

Kwakw’ala Yes Yes Yes

Visibility Topicality/Obviation

Yes Yes (auxiliaries)

Animacy Movement status Number

No Marginal (anaphoric pronouns and switch-reference) No No Yes (with case)

Ponca Yes Yes Yes (in DEM) No Yes (obviation)

Evidentiality

Yes

No No Yes (via root reduplication) Yes

Yes Yes Yes (in articles) Yes

Alessio Muro

41

Several of the categories in Table 1 have been found to be active in the languages that will be surveyed in the next section.

3. Anisomorphism and translation How does grammatical anisomorphism affect the process of translation between typologically different languages? The second goal of this paper is to provide concrete examples of cross-linguistically anisomorphous structures. I selected a sample of two genetically unrelated source languages, Oowekyala and Oneida, which in my opinion well exemplify the kind of variation at issue. They are indigenous of the Northwest and Northeast of North America, respectively: apart from the common typological feature of polysynthetic morphology (a term which should be understood in a quantitative sense only, i.e. a high morpheme-to-word ratio), their morphosyntactic divergences are impressive. The target language chosen is English in both cases: this was necessary in order to facilitate the comparative process, but also because English is the original target language of the translations provided in the corpora. The material considered here is drawn from the following sources: a) For Oowekyala: 3 texts from Hilton and Rath (1982), a collection of 15 oral narratives and songs, carefully transcribed, annotated with valuable linguistic and cultural observations and provided with interlinear translations.9 b) For Oneida: 1 text from Michelson (1981), 1 text from Abbott (1980). Both corpora contain transcriptions of oral narratives (3 in Michelson (1980), 2 in Abbott (1980)), translations and morpheme glosses.10

3.2 Oowekyala This virtually extinct Upper North Wakashan language was first studied by Boas; it is a close congener of the Kwakw’ala language

9

Since morpheme glosses are not provided, I have personally produced such glosses for the examples reported here. 10 The glosses in the two corpora follow different conventions; in this article they have been modified for the sake of consistency. In the few cases of different orthographical renderings of Oneida words, the orthography proposed in Michelson and Doxtator’s (2002) dictionary has been followed.

42

“Lost in Translation” between Typologically Different Grammars

mentioned above.11 It displays a basic VSO word order. Morphologically, it allows only one root per word (no compounds or true noun incorporation); noun incorporation constructions of other languages are realized by affixal predicates, i.e. suffixes which attach to a nominal, adjectival or quantifier root to derive intransitive predicates. 12 Its TMA morphology mainly indicates aspect (but tense categories such as future and an optional remote past are also represented). Arguments are crossreferenced by enclitic pronouns, as in Kwakw’ala (subject, object and oblique clitics are mostly found in the absence of overt corresponding noun phrases). Polysynthesis in Oowekyala manifests itself as long strings of suffixes: (4)

w

w

dúq əla- alam-a a - ənt -sq -ənƛənc [Salm47] see -FOC -NMLZ-seem -3>2VIS.PS-1INCL.P ‘(The fish) just seem to see us.’

This word-sentence contains a root whose meaning is emphasized by a focalizer (- alam, rendered as just in the translation). The nominalizer a a qualifies it as a case of insubordination (Evans 2007), i.e. a non-finite form used as finite (but translated with a simple finite form into English). The agreement clitics specify: third person, medial deixis, visibility (via w the subject clitic -sq , actually a possessive determiner, since the form is nominalized), first person and clusivity (the object clitic -ənƛənc). Apart from person, all the other categories are lost in the English translation. A more literal translation in the style of Boas might sound like “It seems to be the case that they (close to you, visible) just see us (including you).” The most impressive feature of the language is its extremely rich determiner system: articles are suffixal and bipartite, with restricting and saturating components (the latter may also be replaced by possessive indices, which are suffixes for the third and the second person and proclitics for the first person). 13 In addition to the six combinations of deictic and visibility features displayed by Kwakw’ala determiners there is also an absentive (‘just gone’) category, used for something which used to 11

The language is indigenous to the village of Rivers Inlet, British Columbia. The first text collection appeared in Boas (1928): it consists of 8 texts whose reliability is however doubtful, since they were processed by two speakers of different languages, as Howe (2000:8) explains. 12 For a recent approach to noun incorporation, affixal predication and related phenomena, see Muro (2009). 13 A theory of restriction and saturation, albeit applied to different phenomena, can be found in Chung and Ladusaw (2006).

Alessio Muro

43

be visible but is no longer in view (Boas 1947, Bach 2006). With possessed nouns, things become even more difficult: the 7 categories of the determined possessed noun can generate any logically possible combination with the same categories of a third person possessor. As a result, 49 potential forms can be obtained for expressing, e.g. ‘his/her pencil’.14 The following examples (from Bach 2006: 269, glosses adapted) illustrate the extraordinary complexity of the system: (5)

a.

w

adayug ask adayu -ga -sk -3>1VIS.PS pencil -1VIS.RDET ‘his/her (proximal, visible) pencil (proximal, visible)’

b.

c.

w

adayug ack adayu-gac -sk pencil -1INV.RDET -3>1VIS.PS ‘his/her (proximal, visible) pencil (proximal, invisible)’ w w adayug as w

adayu-ga -s pencil -1VIS.RDET -3>2VIS.PS ‘his/her (medial, visible) pencil (proximal, visible)’

Number is not distinguished in determiners and pronominal enclitics, but first person pronouns and enclitics distinguish inclusive and exclusive forms, which presuppose reference to more than one individual. Affixal predication can be illustrated by the following sentence pair. First of all, a child asks his father for a sling: (6)

14

g -i haúma-su -s [Prim3] and-3>3VIS.S ask -PASS-by w w ənùk -a -si sil-i if -3>3VIS.S son -3>3VIS.RDET-3>3VIS.PS w ǧ -s qə -í way.of.doing -of make for -3>3VIS.P ə sling ‘He was asked by his child if he could make a sling for him.’

In the related Haisla language described by Bach (2006) the ‘just gone’ category seems to be compatible with the visible vs. invisible dichotomy, thus yielding no less that 8 × 8 = 64 third person possessive forms.

44

“Lost in Translation” between Typologically Different Grammars

Note that the verb ‘make’ and the object ə ‘sling’ are expressed by separate words. The narration continues with (7): (7)

g -i ə -gila -xʔit qə -í sling -make-ING for -3>3VIS.P and-3>3VIS.S ‘And he made a sling for him.’

[Prim4]

Here the morpheme expressing the meaning ‘make’ (-gila) is suppletive with respect to the free-standing verb in (6): this is why we must distinguish affixal predication constructions like (7) from true noun incorporation (i.e. the compounding of a full nominal root and a full verbal root). These constructions are always intransitive and have a backgrounding effect on the nouns they involve; this shade of meaning is particularly difficult to render in languages like English. Let us now focus on the two categories termed “modalities” by Boas. Oowekyala has very productive evidential suffixes, one of which (- ənt, ‘seemingly’) appears in (4); the most productive is however the hearsay evidential (- , the same as in Kwakw’ala); it can be observed in (8): (8)

w

wá g -i láka then and-3>3VIS.S reach -HRS 3VIS.DEM w w də -y -a g village -EG-3VIS.RDET -3.SDET ‘... after which they reached this village.’15

[Salm35]

This morpheme is not always used when a situation is reported: especially in traditional narratives, it seems to perform the additional function of marking the peak event of a certain portion of discourse; since its use is much more frequent than that of nearly equivalent English strategies like they say or reportedly, it is usually left out in translation. If a constituent is topicalized, it is fronted and preceded by the topic particle yi. Ex. (9) contains both a Topic and a Focus:

15

The hearsay morpheme has no equivalent in the translation; the presentational character of the sentence is however rendered by the use of indefinite this instead w of the normal translation of the distal demonstrative (‘that’, distal).

Alessio Muro (9)

w

yi

ǧ -s qənc ʡəw -xd-i thing.meant -by 1INCL.PS father -AB-AB.RDET

TOP

45 [Nuw14]

w

qúq 2VIS.DEM ‘This is what our father was talking about.’

The yi particle introduces the topicalized constituent; the deictic pronoun is focalized. The English translation uses a presentational construction. In other cases, sentences introduced by yi are translated by means of English cleft sentences, and in fact yi sentences resemble clefts in structure. This issue needs further study, though. What difficulties will one be confronted with when translating an Oowekyala text into English? To sum up: a) Oowekyala tense categories function differently if compared to English tenses; the tense value which an English translation should feature needs therefore to be determined from the context. b) Definite noun phrases bear features of three deictic degrees, plus either visibility or the ‘just gone’ category. These features are lost in the running translations. Moreover, given the presence in English of forms such as indefinite this, even deictic degrees can be altered in translation. c) For long synthetic verbal complexes periphrastic translations of several English words are provided. These, however, do not allow one to capture the pragmatic difference between structures like affixal predication constructions and their analytic equivalents. d) Hearsay is either conveyed in translations via periphrastic strategies (e.g. it is said or they say) or more often it is lost. e) Topicalizations with the yi particle are usually translated with clefts or presentational constructions. f) To express the clusivity features of first person plural pronouns and clitics, in English we have forms such as we all, which are normally inclusive, but again they are not as systematic as the Oowekyala morphemes, and just as in the case of evidentiality, an excessive use of such strategies may result in a clumsy English text. There are also other categories which are “lost in translation”. To name but one example, Oowekyala has numeral classifiers, which are not too different from those of Korean and Japanese: they alone have the power to disambiguate items in a list, if the items take different classifiers. English

46

“Lost in Translation” between Typologically Different Grammars

numerals have no such potential, and here as well periphrastic strategies are often the only available solution.

3.2 Oneida The Northern Iroquoian Oneida language was also familiar to Boas, who published a short grammatical description of it including a text (Boas 1909). Its main typological properties differ greatly from those found in Oowekyala: word order is very flexible, allowing for the expression of various degrees of prominence of the main constituents (“free” word order). Morphologically, it shows a well-developed system of noun incorporation. Aspect and degrees of factuality (i.e. the truth or different degrees of probability of an event) are indicated morphologically in the verbal complex (temporal location is often indirectly coded by such morphemes; a few optional past tense morphemes are also reported by Abbott 2000). Arguments are cross-referenced by a complex system of obligatory pronominal prefixes: in the case of transitive predicates the pronominals indicate subject and direct object for monotransitives, and subject and indirect object for ditransitives. The features represented by the prefixes include person, three numbers (singular, dual and plural), four genders (masculine, neuter, feminine-indefinite and feminine-zoic16) and clusivity in the dual and plural numbers: a complete inflectional paradigm for a given transitive verb includes 58 pronominal prefixes. In Oneida polysynthesis may manifest itself as noun incorporation as well as rather long strings of both prefixes and suffixes: (10)

sayakwanitskwahlaʔtslotúnihe

[Intr24]

s

-waʔ -yakwaITER -FACT -1EXCL.PL.A>3N.SG.P[at - itskw -a -hl -a -ʔtsl] -ot -u -nyu -heʔ [SRF-haunches-LNK-set-LNK-NMLZ]-stand-DIST-DIST -IPFV ‘We used to set up the chairs again.’

This true NI construction shows a complex incorporated nominal root -anitskwahlaʔtsl- ‘chair’, which is in turn a nominalized noun incorporation construction: it corresponds to the slightly different free16

The feminine-indefinite gender in pronominal prefixes is used to refer to an especially respected female person or with generic reference to persons (‘people’) or indefinite reference to persons (‘someone’); the feminine-zoic gender refers to female persons and animals (Michelson 1981:35).

Alessio Muro

47

standing word anitskwahlákhwaʔ (at-itskw-a-hl-a-hkw-haʔ, SRFhaunches-LNK-set-LNK-INS-HAB, Michelson and Doxtator 2002: 108), lit. “the thing one habitually uses to set one’s haunches on for one’s own benefit”. The free-standing word results from the compounding of the roots -itskw- ‘haunches’ and -hl- ‘set’.17 The complex is integrated by the instrumental applicative suffix -hkw (meaning ‘use for’), the habitual aspect suffix -haʔ (one sets one haunches at rest as a habit) and the semireflexive prefix at- (an- before roots beginning with -i, ‘for one’s own benefit’). The incorporating form for ‘chair’ replaces the instrumental-habitual suffix sequence with the nominalizer -ʔtsl. In (10) the complex root for ‘chair’ is compounded with the verbal root -ot- ‘stand something upright, have’ to yield the meaning ‘set up the chair(s)’. The feature of plurality of the object is clarified by the distributive suffixes -unyu. The imperfective serial suffix -heʔ teams up with the iterative prefix s- to yield the idea of a past customary habit; the factual mood prefix waʔindicates the speaker’s commitment to the truth of the described event. Finally, the transitive pronominal prefix yakwa- indicates a first plural agent with exclusion of the hearer, as well as a neuter singular patient, which registers the presence of the incorporated noun.18 The fact alone that the English translation makes use of no less than eight orthographical words to translate this Oneida word-sentence (losing at the very least the clusivity of the pronominal prefix and the internal structure of the component meaning ‘chair’) should give an idea of the complexity of the task of rendering the subtle grammatical distinctions made by this incorporating language. Unlike Oowekyala, Oneida has no articles in the traditional sense (the neʔn particle sometimes has a meaning comparable to that of an article, yet it is used very sparingly). A category which might to a limited extent approximate the Oowekyala ‘just gone’ category is the - decessive, which carries the meaning conveyed by the English adverbial former: ak-e-ʔsleht(1SG.PS-EV-car-DCS, ‘my former car’). 17

The roots are conjoined by a linker -a-, which should not be considered an epenthetic vowel, since it only appears in noun incorporation constructions, the normal epenthesis being -e- in Oneida, as in Mohawk (Mithun 1984: 876). 18 Incorporated nouns are always treated as neuter singular, regardless of the gender and number features of the referents they represent. This is a major difference from Oowekyala affixal predication constructions, which, as we saw in ex. (7), are always intransitive.

48

“Lost in Translation” between Typologically Different Grammars

A more straightforward case of noun incorporation is illustrated in (11) and (12) (examples by courtesy of Karin Michelson): (11)

swatyell kʌs kiʔ né· ohwístaʔ tayuta·tú· swatyell kʌs kiʔ né· o -hwist -aʔ sometimes customarily actually FOC N -money-NSF t -a -yutat -u -ʔ CIS -FACT-3F.A>3F.P -give -PFV ‘Sometimes though they gave her money.’

The sentence above reports on a habit of the speaker’s grandmother, who used to trade the baskets she made for various things, but sometimes also for money. The money is in focus and appears as a free-standing word. But immediately after, the speaker continues with (12): (12)

tayutathwístuʔ t

-a -yutat -FACT -3F.A>3F.P ‘They gave her money.’ CIS

-hwist -u -money -give

-ʔ -PFV

The normal way to express the meaning ‘they gave her money’ is the incorporating construction in (12); this pattern renders the institutionalized character of giving money, which is not conveyed by (11). It can be observed that in (12) a simple nominal root is compounded with a verbal root; under incorporation, the noun loses its gender prefix. Evidentiality is mainly rendered by particles, such as yakʌʔ ‘hearsay’, khelé, (kw)aʔnyó, wéˑne ‘seemingly’, taˑt nuʔu, taˑt núwaʔ ‘maybe’ (Abbott 2000: 57, 2006: 69). These particles do not seem to alter the syntax of the sentence the way the Oowekyala yi particle does; they simply add a shade of uncertainty to an utterance. Topicalizations and focalizations are not morphologized in the language: word order and prosody determine the prominence of a constituent. All possible orderings of major constituents are permitted. Let us illustrate the phenomenon with intransitive predicates. A highly prominent subject is very likely to be fronted, yielding a surface SV order, as in (13):

Alessio Muro (13)

okhnaʔ kayaʔkwáheleʔ

49

ʔ

[Race20]

okhnaʔ [Ska-yaʔkwáhel-eʔ ] [Vt -a -hʌ -ihlu -ʔ ] and.so N- monster -NFS CIS-FACT-3M.A-say -PFV ‘And the monster replied...’

The monster had not spoken in the text prior to this sentence. If the main emphasis is on the predicate, the latter can be fronted as well, yielding a surface VS order, as in (14): (14)

kwah ok tho thúhtuʔ tsiʔ naʔtehóˑkat neʔn

[Race22]

kayaʔkwáheˑle ok tho [Vth -aʔ -w -ahtu -ʔ ] only there CONTR-FACT-3N.A-disappear-PFV naʔ -te -ho -ʔkat neʔn REL PART-DU-3M.A-run.fast just.that ka-yaʔkwáhel-eʔ ] N -monster -NFS ‘Then he just disappeared, that speedy monster.’

kwah just [Stsiʔ

In this sentence, the disappearing is contrastively focalized against other possible events (such as e.g. hiding). The rich obligatory pronominal prefix system of the language helps to correctly identify the role of each argument. What are then the challenges posed by such a grammatical system when we try to translate texts into English? We can observe the following: a) As in Oowekyala, temporal features are not always directly coded in verbal morphology and need to be reconstructed in order to integrate them in an English text. b) Unlike Oowekyala, Oneida has no true articles: in this case as well articles must be provided in order to produce a correct English translation. c) For noun incorporation constructions periphrastic translations have to be provided. In comparison with Oowekyala affixal predication the loss is still greater in Oneida, since in the latter noun incorporation is more pervasive. d) Evidential particles can be rendered in translation via periphrastic strategies or left untranslated. e) Topic and Focus can be rendered by word order alterations or by other means, e.g. cleft sentences or presentational constructions.

50

“Lost in Translation” between Typologically Different Grammars

Other “lost” categories include: f) The clusivity features of first person dual and plural pronominal prefixes on nouns and verbs, as observed for Oowekyala. g) The four-way morphological gender features of nouns and pronominal prefixes.

3.3 Summary Bringing together the threads of the survey of grammatical features exposed in Section 2, in Table 2 I present a synthesis of the same grammatical features illustrated in Boas’ Introduction, specifying whether they are coded morphologically in Oowekyala and Oneida. It can be easily noticed that the two languages differ in their treatment of practically all the features considered. The values of English have been included in order to provide a bird’s eye view of the problems encountered in translation. While English may use patterns similar to those used by one or the other of the two languages described in this section, grammatical anisomorphism affects the translation process from each of the two languages into English with respect to all features except movement status, which is not morphologized in any of the three languages (in this respect, all three languages contrast with the Ponca article system described in Section 2):19 Table 2. Anisomorphous categories in Oowekyala and Oneida Category\Language Tense

Oowekyala Yes

Definiteness

Yes

Oneida Indirectly (mood, particles) No

Deictic degree Visibility

Yes Yes

DEM only No

19

English Yes

Indirectly (freestanding DET) DEM only No

As in Table 1, ‘yes’ is used to indicate that a category is obligatorily coded in the morphology of the language and ‘no’ that it is not coded; some categories are indirectly coded or of restricted occurrence, as specified.

Alessio Muro

Topicality/Obviation Animacy Movement status Number

Evidentiality

Indirectly (particle) Yes (numeral classifiers) No Indirectly (root reduplication + clusivity) Yes

51

Indirectly (word order) Yes (gender)

Indirectly (word order) No

No Yes (3 numbers)

No Yes (2 numbers)

Indirectly (particle)

No

A few observations are in order. First of all, the values given in the table above can only be approximate: the tense systems of Oowekyala and English, for instance, can hardly be comparable, although both categories are somehow morphologically expressed in the two languages. The same can be said about animacy, which in Oowekyala and Oneida is subsumed within the formally different systems of numeral classifiers and nominal gender. Similar considerations may be made for number as well: while this category is overtly and more pervasively coded in Oneida, in Oowekyala it is only indirectly implied by two independent categories: root reduplication on verbs (which indicates a distributive or collective interpretation of an argument, or else provides conative or imitational interpretations to the predicate) and the presence of inclusive and exclusive forms for first person pronouns and clitics. Of particular interest should be the set of features expressed by the Oowekyala determiner system: to my knowledge, the typologically unusual preservation of deictic degrees even in definite articles, together with the systematic and pervasive distinction of visibility and the absentive category still await garnering the attention they deserve from contemporary theoretical linguistics. The same can be said about the feature of movement status expressed by Ponca articles. These facts should give a clear idea of the fact that Boas’ formulation of grammatical anisomorphism, far from being outdated or obsolete, still represents an open challenge to modern linguistics. As far as noun incorporation and related phenomena are concerned, Oowekyala and Oneida differ both in terms of the morphological properties of the constructions available in their grammars and in terms of the referential functions compatible with these patterns. On the one hand, the Oowekyala affixal predication technique allows one at best to infer

52

“Lost in Translation” between Typologically Different Grammars

specific reference for the individual designated by the host nominal root.20 On the other hand, the true root-compounding noun incorporation constructions of Oneida can be used even when the arguments crossreferenced by the pronominals can be understood as unique (in Oneida the construction can sometimes even be accompanied by a demonstrative which appears to be coreferent with the incorporated noun, a phenomenon called modifier stranding).21 As far as the consequences of morphosyntactic anisomorphism for translation are concerned, we can observe that the English categories of tense, determiners, number and topicality call for the utmost attention in the translation process, since their equivalents in the two languages analysed differ a lot in their manifestations if compared to their normal translational equivalents. Visibility, animacy and movement status are always lost in the translation process. Evidentiality and clusivity may to some extent be rendered by periphrastic strategies (expressions like they say, it seems, etc. for the former; inclusive all for the latter); these strategies are however not as common in English as the corresponding grammatical categories are in the source languages, and therefore they must be used sparingly when translating. Last but not least, noun incorporation and affixal predication can often be translated by means of indefinite noun phrases or bare plurals; but this strategy cannot do justice to the wider array of degrees of individuation of referents which are compatible with Oneida noun incorporation constructions.

4. Conclusion: anisomorphism beyond Boas The variation described in this paper is of course only a small part of what can be found typologically on a cross-linguistic scale. In concluding this survey of diverse grammatical resources of different languages, I would like to add a few notes from my study of yet another typologically different language: Korean. Like Oowekyala and Oneida, this rigid SOV language exhibits a complex morphology, with a large number of grammatical categories not represented in English. First of all, I would like to give a brief example of one category which to my knowledge is not represented in any of the 20

For a recent introduction to referentiality see König and Lehmann (2014). This difference between the two languages is very likely due to the greater richness of the functional load conveyed by the Oneida pronominal prefixes, which are always obligatory (recall that even noun incorporation constructions are not actually intransitive in this language). I thank Karin Michelson for pointing this out to me.

21

Alessio Muro

53

languages dealt with in this paper: honorifics. Korean verbs do not agree with any of their arguments, but certain features of subjects do surface in the morphology of verbal complexes: if the subject is a respected person, the subject honorific suffix -si immediately follows the root. If the speaker is talking to a respected person, the addressee honorific suffix -sup must be used as well (the latter takes scope over tense). 22 The pattern can be appreciated in (15), from a narrative text (the verbal root is a compound root resulting from the combination of ttena- ‘depart’ and peli- ‘leave, throw away’, the latter grammaticalized into an adverbial modifier meaning ‘irreversibly’):23 (15)

kunye-ka seysang-ul [SCh3] she -NOM world -ACC ttena -peli -si -ess-sup-ni -ta depart-irreversibly-SH-PST-AH -IND-DECL ‘She passed away (lit. “she irreversibly left the world”).’

Apart from the use of pass away, which indicates respect for the deceased person, English has no straightforward way of rendering the complex interplay of Korean honorifics. Among the most challenging problems posed by grammatical anisomorphism are those cases in which strings of different affixes are attached to one and the same root, with at least one of the grammatical categories represented in the affix strings having no straightforward equivalent in the target language. This is illustrated in (16) for Korean hearsay evidential morphology ((a) elicited, (b) from Lim 2010: 62): (16)

22

a.

nay-ka ecey I -NOM yesterday

chwihayss get.drunk.PST

-ta -DECL

b.

nay-ka ecey I -NOM yesterday ‘I got drunk yesterday.’

chwihayss -ta -nta get.drunk.PST -HRS-DECL

All the 4 possible combinations of honorific vs. non-honorific subject and addressee marking are possible and in current use. 23 The example is taken from Hwang 1987, which includes a collection of glossed and translated folklore narratives. Korean data are given in the Yale transliteration system. I am grateful to my informants Park Jin-kyung and Cho Sung-yun for helpful advice.

54

“Lost in Translation” between Typologically Different Grammars

(16a) is a plain statement; in the case of (16b), the speaker knows the information by hearsay. The Korean hearsay evidential is often left untranslated in English (although strategies like e.g. I was told that or it seems like are sometimes acceptable). To conclude, on the basis of what has been discussed so far it should be apparent that the difficulties due to morphosyntactic anisomorphism are not insurmountable, yet they require special skills to be dealt with properly. To a certain extent, speakers of different languages do think differently, insofar as they are forced to think about features of individuals and events that need to be expressed in order to fulfill obligatory requirements posed by the grammatical systems of their languages. But they can learn to think differently, and they can do so by learning other languages.

Acknowledgements I would like to express my appreciation to the members of the Organizing Committee of the TLAL workshop for their assistance and interest in my research. For greatly valuable comments and advice on the languages discussed in this paper I am especially grateful to Karin Michelson (Oneida) and Michael Fortescue (West Greenlandic). Special thanks are also due to my Korean informants Park Jin-kyung and Cho Sung-yun. For helpful discussions I am indebted to Paolo Ramat, Anna Giacalone Ramat, Ursula Stephany, Michele Prandi, Alberto Mioni and Rosanna Benacchio. I would also like to extend my thanks to two anonymous referees for their comments on an earlier draft of this paper and to Erik Castello for editing my English. The usual disclaimers apply.

References A. Corpus Oowekyala (Hilton ( and Rath 1982): w [Prim]: g əmkasʔux ƛ , The One Called g əmkasʔu [Nuw]: núˑwaqawaxƛ q , The One Called núˑwaqawa. w [Salm]: ǧ əniskaʡəw , The Dog Salmon Oneida (Abbott 1980, Michelson 1981): [Race]: The Race of Life (Abbott 1980) [Intr]: Introduction (Michelson 1981)

Alessio Muro

55

Korean (Hwang 1987): [SCh]: hyonye sim-cheng, The Story of Shim Chung

B. General Abbott, Clifford. 1980. “Two Oneida Texts.” In Northern Iroquoian Texts. International Journal of American Linguistics, Native American Texts Series, Monograph No. 4, ed. Marianne Mithun and Hanni Woodbury, 67-76. Chicago, IL: The University of Chicago Press. ―. 2002. Oneida. Languages of the World/Materials 301. Munich: Lincom Europa. ―. 2006. Oneida Teaching Grammar. Green Bay, WI: University of Wisconsin. https://www.uwgb.edu/oneida/filesToDownload/teaching%20grammar %20revised4.pdf Anderson, Stephen R. 1992. A-Morphous Morphology. Cambridge Studies in Linguistics 62. Cambridge: Cambridge University Press. Bach, Emmon. 2006. “Paradigm Regained: Deixis in Northern Wakashan.” SOAS Working Papers in Linguistics 14: 267-81. Accessed September 12, 2010. https://www.soas.ac.uk/linguistics/research/workingpapers/volume14/file37827.pdf Beek, Wouter. 2006. “Linguistic Relativism: Variants and Misconceptions.” Ms., Univ. of Amsterdam. Accessed May 2, 2014. http://staff.science.uva.nl/~bredeweg/pdf/BSc/20052006/Beek.pdf Berthelsen, Christian, Birgitte Jacobsen, Robert Petersen, Inge Kleivan and Jørgen Rieschel. 1997. Oqaatsit Kalaallisuumiit-Qallunaatuumut/ Grønlandsk-Dansk Ordbog. Nuuk: Ilinniusiorfik. Bittner, Maria. 2014. Temporality: Universals and Variation. Hoboken, NJ: Wiley-Blackwell. Boas, Franz. 1909. “Notes on the Iroquois Language.” In Putnam Anniversary Volume, ed. Franz Boas, 427-60. New York: Stechert. ―. 1911. “Introduction.” In Handbook of American Indian Languages, 1, ed. Franz Boas, 1-83. Washington, D.C.: Government Printing Office. ―. 1928. Bella Bella Texts. New York: Columbia University Press. Boas, Franz, Helene Boas Yampolsky and Zellig Harris. 1947. “Kwakiutl Grammar with a Glossary of the Suffixes.” Transactions of the American Philosophical Society 37.3: 202-377. Brown, Roger L. 1967. Wilhelm von Humboldt’s Conception of Linguistic Relativity. The Hague and Paris: Mouton.

56

“Lost in Translation” between Typologically Different Grammars

Chafe, Wallace. 1998. “Do Speakers of Different Languages Think Differently?” Languages of the North Pacific Rim 4: 1-16. Chung, Sandra and William Ladusaw. 2006. Restriction and Saturation. Cambridge, MA: MIT Press. Eschenberg, Ardis. 2005. The Article System of Umonhon (Omaha). Ph.D. diss., State Univ. of New York at Buffalo. http://linguistics.buffalo.edu/people/faculty/vanvalin/rrg/The%20Articl e%20System%20of%20Umonhon%20Diss.pdf Evans, Nicholas. 2007. “Insubordination and its Uses.” In Finiteness. Theoretical and Empirical Foundations, ed. Irina Nikolaeva, 366-431. New York: Oxford University Press. Flynn, Darin and Hilda Smith†. (in progress). Oowekyala-English Dictionary. Ms., Univ. of Calgary. Fortescue, Michael. 1984. West Greenlandic. Descriptive Grammars. London: Croom Helm. Greenberg, Joseph H. and Merritt Ruhlen. 2007. An Amerind Etymological Dictionary. Stanford: Department of Anthropological Sciences. Hilton, Susanne and John Rath, eds. 1982. Oowekeeno oral traditions, as told by the Late Chief Simon Walkus Sr. Ottawa, National Museums of Canada. Howe, Darin M. 2000. Oowekyala Segmental Phonology. Ph. Diss., University of British Columbia. http://depts.washington.edu/wll2/files/howe_00_diss.pdf Hwang, Shin Ja Joo. 1987. Discourse Features of Korean Narration. SIL Publications in Linguistics 77. Dallas, TX: SIL. http://www.sil.org/system/files/reapdata/11/03/47/1103474652128081 24676583518041085742461/21326.pdf König, Tina and Christian Lehmann. 2014. “Reference: communicative functions and operations.” Ms., University of Erfurt. Accessed May 2, 2014. http://www2.uni-erfurt.de/sprachwissenschaft/Vgl_SW/ referentiality_2014/cl_tk.pdf Lim, Dong Sik. 2010. Evidentials and Interrogatives: a Case Study from Korean. Ph.D. diss., University of Southern California. http://digitallibrary.usc.edu/cdm/compoundobject/collection/p15799col l127/id/396123/rec/1 McWhorter, John H. 2014. The Language Hoax: Why the World Looks the Same in Any Language. New York: Oxford University Press. Michelson, Karin, ed. 1981. Three Stories in Oneida. Ottawa: National Museums of Canada. Michelson, Karin and Mercy Doxtator. 2002. Oneida-English/EnglishOneida Dictionary. Toronto: University of Toronto Press.

Alessio Muro

57

Mithun, Marianne. 1984. “The Evolution of Noun Incorporation.” Language 60 (4): 847-94. Muro, Alessio. 2009. Noun Incorporation: A New Theoretical Perspective. Ph. Diss., Univerity of Padua. http://paduaresearch.cab.unipd.it/2113/1/Muro_NI.pdf Shaer, Benjamin. 2003. “Toward the tenseless analysis of a tenseless language.” In The Proceedings of SULA 2, Vancouver, BC, ed. Jan Anderssen, Paula Menéndez-Benito and Adam Werle, 139-56. Amherst, MA: GLSA, University of Massachusetts. Accessed June 25, 2014. http://www.umass.edu/linguist/events/SULA/SULA_2003_cd/files/sha er.pdf Swift, Mary D. 2004. Time in Child Inuktitut: A Developmental Study of an Eskimo-Aleut Language. Berlin and New York: Mouton de Gruyter.

METALINGUISTIC REFERENCES IN CROSS-MODAL TRANSLATIONS: SIGN LANGUAGE INTERPRETING AND ITS ISSUES GIULIA PETITTA ISTC-CNR, ROME

This paper aims to discuss some critical aspects related to the translation of metalinguistic references, focusing on a particular cross-modal translation: simultaneous interpretation from spoken to signed languages and vice versa. Reflexive metalinguistic uses represent a challenge in translation studies, since metalinguistic function requires special strategies to be properly identified and translated (or adapted). For example, in spoken languages, autonymy must be reported by leaving the quoted word as it is, without actually translating the word itself. Therefore, any modification requires a special contextualization. However, in sign language interpretation the word (or the sign) cannot be left as it is: sign languages are conveyed by a multidimensional medium, and spoken-signed translation has to take into account not only language differences, but also (and mainly) modality differences between speaking and signing (auditoryvocal modality vs. visual-gestural modality). Thus, the interpreter must employ appropriate strategies to express autonymy and other metalinguistic functions using a different medium.

1. Introduction Reflexivity, the possibility of language to be used to talk about itself (Rey-Debove 1978; Lucy 1993), is one of the properties of human language that has challenging implications in translation studies, since it involves the issue of the untranslatability and allows to discuss the concept of translation itself (Ramat 2007). Metalinguistic functions require indeed special strategies to be properly identified and translated (or adapted) (Rey-Debove 1978; Authier-Revuz 1995; De Meo 2010; Roibu 2010).

Metalinguistic References in Cross-modal Translations

60

In particular, autonymy, or self-reference, the quotation of words referring to the word itself (Rey-Debove 1978; Authier-Revuz 1995), represents an extreme in the metalinguistic references continuum, since the translator should leave the word as it is. When a word is quoted in an autonymous way, it bears no denotative meaning in the extra-linguistic reality, but simply refers to itself. Two examples are provided below with the autonymous word in italics: (1) (2)

Cat is an English word. In German you can Hänchenbruststreifen.

find

many

compounds,

such

as

An Italian translation of (1) could be “Gatto è una parola inglese”, but it would sound rather unusual, due to the fact that gatto is actually an Italian, not and English word, and an effective translation would be consequently “Cat è una parola inglese”. These references could obviously occur in metalinguistic discourse, such as lectures, seminars, conferences and discussions, aimed to describe, characterize, evaluate, consider and reflect on different aspects of language. However, reflexivity is a property of all languages and can occur in daily conversations not related to the language itself in different forms of quotation, formulation and repair strategies, with a crucial role in comprehension processes (De Mauro 1994). These interactions include sentences like “What does it mean?”, or “Did you say ‘high’ or ‘hi’?” (Taylor 2000; Duncker 2012), as well as name and word quotation at a further reflexive level in a non-specialized environment, and belong to the basic competence of all speakers since their infancy (Ferreri 2012). Thus, autonymy is not only part of metalinguistic discourse focusing on formal properties of language, but can also occur in daily nonspecialized interactions. Another famous example taken by the J. K. Rowling’s bestseller Harry Potter and the Philosopher’s stone (chapter 5) can be recognized as a typical metalinguistic interaction between two speakers with no academic purposes: (3)

“I never know,” Harry called to Hagrid over the noise of the cart, “what’s the difference between a stalagmite and a stalactite?”. “Stalagmite’s got an ‘m’ in it,” said Hagrid.

Without discussing the several theoretical implications related to the definition of metalinguistic references, it is possible to consider the following macro-levels of reflexivity:

Giulia Petitta

-

-

61

reference to language itself (e.g. autonymy: mere quotation of words); reference to a previous linguistic utterance or situation (reflexivity, formulation, etc. by the use of metadiscursive forms of expression for talking about language); metalanguage and reflexive vocabulary (more or less technical terms related to language)

Metalinguistic references are particularly interesting due to their status and effects on simultaneous interpretation, which has peculiar time constraints. Furthermore, as a crucial issue in translation practice and theory, reflexivity must be taken into account in order to achieve a deeper understanding of simultaneous interpreting processes, among which, sign language interpreting presents some particular features. The interpreter should indeed transfer the sense from the auditory-vocal modality to the visual-gestural modality and vice versa, with some theoretical and practical implications.

2. Metalinguistic references and sign language simultaneous interpreting The issue of metalinguistic references in sign language interpreting seems to be particularly challenging for two reasons. The first is related to the structure of sign languages: sign languages are conveyed through a visual-gestural medium, thus spoken-signed translation has to take into account not only language differences, but also (and mainly) modality differences between speaking and signing (auditory-vocal modality vs. visual-gestural modality). The second reason is related to social-historical factors: many signed and spoken texts produced among the signing community are related to the language itself. Thus, metalinguistic discourse is very often held in interpreter-mediated interactions between deaf and hearing people. Due to the crucial role played by research in the deaf communities and its impact on sign language usage and perception, in the last decades a lot of discussions and reflections at a more or less specialized metalinguistic level rose.1

 1

Sign Language research started in USA in late Fifties of the Twentieth century, and in Italy about twenty years later (Fontana et al. in press). In this paper I mainly refer to the Italian situation, but the challenges and the problems discussed here could be easily adapted to other signed languages, due to the similarities grounded

62

Metalinguistic References in Cross-modal Translations

Of course, reflexive use is more frequent in metalinguistic or epilinguistic discourses, with some problems related to the simultaneous presence of metalanguage, of general reflexive remarks and of autonymy (De Meo 2010). Consequently, it must be considered how metalinguistic issues interact with the factors affecting cross-modal simultaneous interpreting and the strategies an interpreter can adopt in these situations. Furthermore, metalinguistic signed discourse seems to have special features, with consequences on register variation and “processes of phatic control”, such as nodding, mouthing, repetition and accuracy in articulation (Meurant and Garcia 2010; Paligot and Meurant 2013). Moreover, many metalinguistic discourses are obviously held in an academic context, where the interpreter should take into account a considerable amount of variables (Napier 2004; Källkvist 2013). The direction of interpretation also has a crucial role. Autonymy can occur in two different ways: the first one is related to the language used in a discourse, e.g., quoting an English word during a discourse held in English. On the other hand, it is possible to quote a foreign word during a discourse held in a given language, e.g., quoting an Italian word during a discourse held in English. Consequently, in sign language interpreting, autonymy can also occur in different directions, and the interpreter should manage the following possibilities: a) word quotation while speaking (and being translated into signs) b) word quotation while signing (and being translated into words) c) sign quotation while signing (and being translated into words) d) sign quotation while speaking (and being translated into signs) On the other hand, in sign language interpreting settings, the visibility of the interpreter makes the participants aware that they are being translated: both in conferences and in community-based settings, the signing interpreter is in motion and visible for the participants to see.2 Thus, reflexive practices and remarks related to a meta-discourse interaction, such as the examples shown in (4), occur very often in sign language interpreting:

 in more general features that are linked to the visual-gestural modality of language expression. 2 In some spoken interpreting settings the interpreter is visible too, but it is important to note that in sign language interpreting deaf people involved in the interaction must have visual access to the interpreter, who is consequently always, in some way, involved in the interaction (Metzger 1999).

Giulia Petitta (4)

63

a) “This word is very difficult.” b) “Sorry for the technical terms.” c) “I am going to use a technical word.”

Thus, it is important to evaluate the interpreting strategies of metalinguistic interventions and glosses, related both to the discourse itself and to the translation process and results, not only to discuss their status in the language use, but also to consider their implications in translation studies, referring to the signed interpreter-mediated interaction.

2.1. Cross-modal simultaneous interpreting: semiotic and pragmatic issues According to Jakobson (1959), an intersemiotic translation “is an interpretation of verbal signs by means of signs of nonverbal sign systems” (1959: 233). Some researchers have discussed the modality-dependent factors affecting sign language interpreting in these terms, and considered the difference of modality between spoken and signed languages as a different way of convey meanings, and consequently have defined sign language interpreting as an intersemiotic or a bimodal translation (see among others Leeson 2005; Nicodemus and Swabey 2011; Petitta and Del Vecchio 2011; Fontana 2013). However, due to the word nonverbal used by Jakobson, the intersemiotic translation is more often referred to, or at least has been often understood, as an interpretation of linguistic signs by a non-linguistic material, e.g. a book becoming a movie, a poem becoming a video, etc. Especially recently, when translation has involved different levels of materials at the same time (subtitling, video recording, speech, writing, etc.), also the term multimodal has begun to be used to refer mainly to the integration of linguistic and non-linguistic materials. Consequently, the vast majority of scholars define multimodality as an integration of verbal (linguistic?) and non-verbal (non-linguistic?) information. However, according to some researchers (McNeill 1992; 2005; Kendon 2004), language can be defined as a gesture-speech integrated system, and furthermore gestures and signs can be considered as parts of a ‘continuum’ (McNeill 2012). In this framework, a crucial relationship between gestures and signs can be identified, and the iconization process that represents the ground for emergency of sign languages can be interpreted as essential in considering the characteristics of human language and communication. Nevertheless, further research is needed to discuss the continuum “Action-Gestures-Signs/Speech”, in order to develop an approach to language as a form of action (see also Capirci and Volterra 2008; McNeill 2012).

64

Metalinguistic References in Cross-modal Translations

Though the term multimodality currently mainly refers to the interaction among different channels of communication, not necessarily at a linguistic level (subtitling, filmed actions, movies, writing, human language, etc.), it is important to note that in sign language studies multimodality is primarily used to denote the use of different articulators in producing sign language (Pizzuto 2003, Vermeerbergen et al. 2007). In sign language research, Jakobson’s intersemiotic translation as been interpreted as the transfer of a linguistic message from a given modality to another. Consequently, I prefer to use the term cross-modal, according to a long-term tradition of studies related to multimodality in a linguistic (verbal and non-verbal) perspective. Furthermore, it is important to note that, while multimodality and the simultaneous use of visual-gestural articulators are pervasive in sign languages, the multimodality of spoken languages should not be underestimated. This has a crucial implication in sign language interpreting, since the signing interpreter must be visible to the deaf audience during the spoken-to-signed interpretation, and is often not visible to the non-deaf audience during the signed-to-spoken interpretation, especially in conference settings. For the same reasons, the interpreter often doesn’t have access to the visual information conveyed during a talk to be translated in sign language. Therefore, cross-modal interpreting involves some modality-dependent elements that can affect the interpreting process.

2.2. The peculiarity of the information conveyed by signs: the “saying-by-showing” modality of expression and its implications for metalinguistic references One of the main differences between spoken and sign languages has been defined as a “semiological-functional possibility of expression”, since the visual-gestural channel affords signers the possibility to provide very detailed visual information. According to Cuxac (2000), this semiological difference can have some crucial consequences: signers have the possibility to use two different modes of expressions: they can provide information by “saying”, i.e. using lexical units, or by the so-called “saying-by-showing” mode, using particular iconic structures with a depictional and descriptive intent, and providing visual details on the events, processes, shapes, sizes and spatial relations related to the referents they are talking about (see also Cuxac and Antinoro Pizzuto 2010). This dichotomy between lexical items and descriptive/depictional structures has been noted since the very beginning of sign language

Giulia Petitta

65

research, and has been interpreted in several ways, from both a theoretical and a terminological point of view (among others, see Friedman 1977; Klima and Bellugi 1979; Johnston and Schembri 1999; Emmorey 2003, Vermeerbergen 2006). The saying-without-showing mode includes units that have been defined as “frozen signs”, “lexical signs”, “lexical items”, or “conventional signs”, while the saying-by-showing expressions cover units defined as “classifiers”, “depicting signs”, “productive morphemes”, “property markers” and all the categories related to “productive signs”, “constructed action”, and “role playing” and “impersonation”. The semiological model (Cuxac 2000; Cuxac and Antinoro Pizzuto 2010) identifies some crucial aspects related to the communicative intent of the signer, and seems to be very suitable to improve our understanding of the decision making process during the interpretation (see also Petitta and Del Vecchio 2011). The saying-by-showing mode represents a challenge in simultaneous interpretation because it conveys a very detailed information that cannot be provided by a single word. Depictional and highly iconic signs can be also partially assimilated to gestural units, and their translation using the vocal channel alone could be difficult, due to the visual information provided by the signer (Leeson, 2005; Petitta and Del Vecchio 2011). Taking into account these elements, McNeill’s “Gesture continuum” (McNeill 2012) could be relevant not only to discuss the nature of signed units, but also to improve our understanding of the spoken rendering of iconic signed structures. The pragmatic function of autonymy is conveyed by this latter semiotic expression mode, and shares with it some formal elements related to articulation, with special reference to the eye-gaze direction (Meurant and Garcia 2010). Thus, it is possible to consider autonymy as a ‘showing words’ (or ‘showing signs’) operation. However, though autonymy can be defined as a word (or a sign) quotation in order “to mention the expression rather than to use it in its normal way” (Saka 2013: 938), it is important to note that words (and signs) are quoted with different purposes. If the aim is to “exhibit a word or expression in order to speak of its linguistic properties” (Saka 2013: 937), the information included in the original signed or spoken form is necessary, and should be rendered accessible to the audience. On the other hand, if the word (or the sign) is quoted not to discuss its form at a metalinguistic level, it is possible to translate it, or to quote it in the target language, using different strategies of “literal” translation.

66

Metalinguistic References in Cross-modal Translations

3. Cross-modal Metalinguistic Interpretation: processes and strategies Since sign languages are conveyed through the visual-gestural medium, the interpretation of reflexivity pose some problems related to the use of language in a cross-modal perspective. It is important to note that all signers are bimodal bilingual individuals: they manage at least a signed language and a spoken language to some degree, and can refer to their own metalinguistic awareness to make crossmodal metalinguistic references. Sign language interpreters have to take cross-modal issues into account, though they cannot refer to same medium used by those languages. For example, in the case of word quotation, the interpreter cannot leave the word (or the sign) as it is (in the original auditory or visual channel): he/she must employ appropriate strategies to express autonymy and other metalinguistic senses using the other communication channel. Therefore, many of the strategies that can be adopted by interpreters can be individuated in the reflexive uses at a cross-modal level, i.e. when signers quote a spoken/written word or when speakers quote a sign.

3.1. Word quotation: Fingerspelling as metalinguistic reference Signed languages can express the forms of written (and spoken) languages through the use of fingerspelling, a means of communication that allows the signer to replicate the alphabetic characters using a set of handshapes. Fingerspelling, also referred to as the manual alphabet, has been largely studied in a sociolinguistic perspective as a borrowing mechanism, often used when a corresponding sign is not available (Battison, 1978; Sutton-Spence 1998; Mulrooney 2002; Adam 2012; Quinto-Pozos and Adam 2013), though its metalinguistic potentialities had been already individuated in relationship with written language and reading abilities (see among others Hirsh-Pasek 1987; Puente et al. 2006). However, without making any explicit reference to the metalinguistic function, many researchers define fingerspelling as a ‘representational’ tool for spoken/written words, and point out that it is often used to convey meanings even when a signed equivalent of the word exists (SuttonSpence and Woll 1993; Brennan 2001; Padden and Gunsauls 2003; Schembri and Johnston 2007). Furthermore, the “contrastive function” individuated by Padden and Gunsauls (2003) seems to be related to a general metalinguistic and metapragmatic use, since “fingerspelling goes

Giulia Petitta

67

beyond simply representing to actually signifying […] to assign contrastive meanings to both native signs and fingerspelled words” (2003, 26). Some metalinguistic functions have been identified in the fingerspelling use in Italian Sign Language (Petitta and Di Renzo 2013): a) Reference to the meaning of a sign by using a signifier belonging to a spoken language, in order to clarify the meaning of a sign previously articulated; b) Reference to a given meaning to explain or ask for an explanation about a concept without using signs; c) Reference to a word (or a part of it) of a vocal language, despite the presence of a corresponding sign; d) Reference to fingerspelling itself.3 Functions a) and b), though strictly related, are different: the first one is often used by interpreters during translations to clarify the use of signs that might not be well-known to the audience. This strategy can be used both by interpreters and by signers, in spite of the level of proficiency. On the other hand, function b) refers not to a sign but to a concept or a meaning, and it is often used by non-expert signers to ask for the form of a sign. Fingerspelling in a) always occurs after or before a sign with approximately the same meaning, while in b) the fingerspelled word is referred to a concept, and never occurs before or after a sign, unless a sign is used with a metalinguistic function (i.e. an autonym), and is signed by the other interlocutor.4 A similar strategy is often used by teachers to convey technical terms, scientific vocabulary and new concepts with educative purposes. Fingerspelling has often been identified in bilingual education programs and deaf schools as an educational technique to demonstrate lexical equivalence between signs and words (Kelly 1995; Humphries and MacDougall 2000; Padden and Gunsauls 2003). This use of fingerspelling can be also considered as the c) function, used to refer to a signifier of a spoken/written language. It is often used during metalinguistic discourses at a more or less specialized level, e.g.

 3

Since metalinguistic reference is a universal property of languages, a signer can of course use fingerspelling with a metalinguistic function, i.e. referring to the manual alphabet itself, for example during a lecture or a talk related to sign language and fingerspelling. 4 For example: a student can ask to a teacher the sign for ‘good’ using fingerspelling, and the teacher can reply showing the sign.

68

Metalinguistic References in Cross-modal Translations

when a signer wants to talk about some phonological or morphological features of a spoken/written language. Interpreters often use fingerspelling for metalinguistic reference to words, both in academic and in community-based settings. Yet, the manual alphabet is not the only means of expression related to spoken/written words. It is important to note that autonymy and word quotation can be made at different levels, and often can be adapted and translated without leaving words in their original form. For example, when an autonymous reference does not discuss specific features of a language, modifications and adaptations can be added to a sign-to-word literal translation. This strategy, known as transcoding, may be easily adopted in some special contexts, taking into account the bilingualism of deaf people and the needs of deaf audience (Nicodemus and Emmorey 2013).

3.2. Sign quotation: showing signs The quotation of a sign poses the same issues posed by word quotation in the interpreting process, but in this case spoken language has no means of communication to express visual information, while signers can use fingerspelling to convey spoken/written-related information while signing. Furthermore, in sign language interpreter-mediated interactions, all signers are also spoken/written language users, and are aware of what a spoken/written word is, but not all speakers (in other words: not the whole non-deaf audience) are signers and are aware of what sign language is and what a sign is. This asymmetry of the audience must be taken into account, since in a metalinguistic context people unaware of sign language receive information about signs. Therefore, the saying-by-showing mode represents a problem in simultaneous interpretation, since the visual information has a crucial role and can convey details (Petitta and Del Vecchio 2011). In these cases, it is often possible that the interpreter would not be visible for the non-deaf audience, and must use the vocal channel alone to convey meanings and senses referring to visual information. Sign quotation as a showing operation is strictly related to the visual information conveyed by signs and by the saying-by-showing mode: when working between a spoken source language and a signed target language, interpreters are also forced to make decisions regarding the relative locative size and shapes of items. Information about these features may be inferred in an English source language text, but are not explicitly coded in the language to the extent that they are in a signed language. Similarly, the signed structure of numerous verb complexes, especially

Giulia Petitta

69

classifier-verb predicates, may require additions in the target English text. (Leeson 2005, 60)

In the case of metalinguistic references related to the quotation of signs, or to the discussion of their articulatory and formal features, the visual information provided is obviously crucial and should be rendered in order to provide an effective interpretation. In autonymous quotation during a signed discourse the interpreter cannot assume that the information related to the form of the sign will be understood by the audience. In order to transmit this information the interpreter may adopt two strategies: 1) he/she can stop the signer in order to allow the nonsigning audience to look at the sign showed during the discourse; 2) he/she can try to convey visual information through the vocal channel, for example using metalanguage or describing the sign shown in the discourse. In the first case, if the signer is illustrating some characteristics related to a sign or a part of it by showing the sign itself, the interpreter will interact with the signer, by saying something like “This sign is…”, but of course the word this cannot make sense without a reference to what the signer is showing. Though the audience can have access to visual information, the lag time can interfere, since the spoken information provided by the interpreter comes necessarily later than the visual information (and the sign quotation itself) provided by the signer. Thus, the signer could show a sign while the interpreter is still introducing it, and he/she must be interrupted in order to allow the interpreter to show the sign to the audience using deictic elements such as “This sign”, “This handshape” etc. On the other hand, in the second case, the interpreter can avoid the deictic reference and makes use of periphrases (e.g. “the sign articulated with the pinky finger extension”, etc.), often containing further metalinguistic references (e.g. “the handshape used in signs such as “mother”, “work”, “car” etc.). To use such a strategy, the interpreter must not only know the metalanguage required to translate these contents, but also must take into account the time-dependent constraints that affect the spoken interpretation of signs. Simultaneity of sign language can indeed affect the timing for the reasons related to the gap between visual and spoken information discussed above. In sign quotation made by a speaker, the same lag time constraints apply. The interpreter must often interact directly with the speaker in order to have access to the visual information provided by him/her, and it is important to note that the interpreter cannot often see speakers, who are usually positioned in a place visible for the audience but not necessarily

70

Metalinguistic References in Cross-modal Translations

for the signing interpreter, especially in conference settings. For example, the interpreter may use the strategy of asking the speaker to wait so that the deaf people can turn their gaze from the interpreter to the speaker who is showing a sign, in order to receive the visual information directly from him/her. Thus, metalinguistic reference to signs can be effectively interpreted not only taking into account all constraints, available strategies and decision making processes related to reflexivity. Interpretation is indeed also affected by extralinguistic factors such as the opportunity of interrupting the signer/speaker; the visibility of the signer; the metalinguistic knowledge (in terms of metalanguage and language theory) of the interpreter; time constraints related to the lag time and to the gap between signs and words in signed-to-spoken language interpretation.

3.3. The extralinguistic paradox: pointing, illustration and interactions Saying-by-showing is a mode of expression typical of sign languages, but is also a strategy largely adopted by speakers during oral conversations. The multimodality of spoken languages can depend on the interaction of different semiotic expressions (voice, gestures, facial expression, body movements) and materials (pictures, writing, videos, etc.). Especially in academic and formal contexts, speaker and signers make multimodal references during their discourse, and different sources of information are provided simultaneously (e.g. video, pictures and language). Such simultaneous co-references and cross-references to different materials cannot be accomplished in a signed interpretedmediated interaction, since deaf people cannot look at different places simultaneously. On the other hand, hearing people are used to receive audio-visual information simultaneously. Thus, if the speaker is talking during a video-projection, or is talking while focusing at a particular point of a slide and is showing it, the interpreter must be aware that deaf people cannot receive simultaneous visual linguistic and non-linguistic information, and must consequently act in order to provide all information. Therefore, when sign language interpreters interact with speaker/signers making multimodal references, taking into account the visual attention patterns of deaf people, they have several options: they can stop the speaker, point at the materials or establish other ways of interaction with the participants. These strategies are obviously effective also in case of reflexive uses and autonymy. If the quoted word (or sign) is available and

Giulia Petitta

71

visible in the interpreting setting, the interpreter can point at it in order to make it visible for the deaf audience. In this way, reflexivity, one of the main linguistic properties of human language, both in its spoken and signed form, could be rendered via a reference to a non-linguistic object such as a picture, a printed word or an iconic material. In this case, as in the discussion of sign quotation, aspects related to the setting (position, visibility, distance, etc.) become crucial for the interpreter to assess the possibilities to adopt different strategies. For this reason, such strategies are mainly used when the direction of interpretation is speech-to-sign and when the audience includes a limited number of deaf individuals. This strategy is largely used for a particular form of autonymy. To show technical terms the interpreter often refer to the written form of the word (available in a book, in a PowerPoint presentation provided during the lecture or in another accessible means of visual communication), while rendering the meaning with a sign.5 If a technical term occurs in a lecture about economics, the interpreter has of course to translate it, but has also to show it to the deaf participant, in order to let him/her recognize the word as it is in further quotations. This autonymous use is peculiar, since it doesn’t occur in the original text, where the word is quoted not necessarily with a metalinguistic purpose. On the contrary, this purpose is present in the interpreter-mediated interaction. Such extra-linguistic strategies are very frequently used in sign language interpreting in academic settings, and can be adopted for autonymy and for other non-metalinguistic showing strategies, but of course cannot be used when the metalinguistic reference involves the translation itself or the reflexive use of language at different levels.

3.4. Metalinguistic glosses and interventions As mentioned before (see section 2), in signed interpreter-mediated interactions the participants are often—though not always—aware of being translated, and the interpreter is frequently perceived as a participant him/herself. In these cases, the speaker/signer is aware of how difficult could it be for the interpreter to translate, and he/she often apologises for the complexity of the lexicon or for other linguistic features, adding

 5

Many of the strategies related to technical terms and metalanguage depend also on sign conventions established between deaf and interpreters during the interpretation.

72

Metalinguistic References in Cross-modal Translations

metalinguistic glosses such as “Sorry for the term”; “Am I talking too fast”, “Is my interpreter ok with it?”, “I can’t wait to see how the interpreter signs this” etc. To clarify the constraints and possibilities related to these reflexive remarks, a brief example extracted from an interpreter-mediated signed interaction will be discussed. During a meeting held in spoken Italian and translated in Italian Sign Language (LIS) by an interpreter, a speaker used a metaphor including an Italian idiom, and then apologised for having used it: (5)

Speaker: “[…] We have now to take the bull by the horns. Oh, sorry: I am always using these idioms and metaphors: I don’t know how the interpreter translated it”.6

The interpreter had just translated the idiom making the choice to omit the reference to the idiom and to interpret it with an equivalent nonfigurative expression. However, the metalinguistic comment had been translated: (6)

Interpreter: “ […] We have to face the situation openly and directly. Oh, sorry: I always use these idioms and metaphors: I don’t know how the interpreter translated it”.7

Obviously, the deaf audience immediately asked what idiom was the speaker referring to, since they didn’t receive the same message as the non-deaf audience. In this case, the interpreter omitted a metalinguistic and pragmatic information related to the linguistic choices of the speaker, but didn’t omitted the further metalinguistic comment, and the interaction had to be interrupted in order to clarify what can be defined, in Cokely’s terms, a miscue (Cokely 1992).8

 6

As already pointed out, this text is an extract of a meeting held in spoken Italian and translated in LIS. The sentence has been translated in English in order to clarify the aspects under discussion. The original idiom in Italian has by chance a literal correspondence with the English one: prendere il toro per le corna. 7 Here the LIS signed interpretation has been translated in English in order to make the content accessible to the readers. The LIS signs have not been transcribed because the form of the signs are not necessary, in this case, for the discussion of the pragmatic implications of the decision making process. 8 Cokely (1992) identified several categories of miscues, or errors in a target language production: additions, omissions, substitutions, intrusions and anomalies. Among these, “Anomalies”, are related to “instances in which the tL [target

Giulia Petitta

73

Metalinguistic remarks, especially when referring to a previous quotation (De Brabanter 2010) require special attention by interpreters, who must be aware of omissions and additions made during the interpretation process (Napier 2004; Leeson 2005). Many of these miscues can be avoided by previous conventions established between interpreters and deaf people. However, it is important to note that, especially in contexts such as university lectures and international conferences, the decision making process is often independent from the audience needs and preferences, since metalinguistic references are unpredictable. At a more general level, a great deal of metalinguistic information is provided during an interpreter-mediated conversation (in both conference and community-based settings). For example, when a speaker uses a technical term, every addressee is able to realize, irrespective of the ability to process the meaning, the technical level of the word. The metalinguistic information provided by the form of the word is rather inaccessible for deaf people. The same happens of course for signs used during a signed discourse. Moreover, word quotation in a metalinguistic context often involves foreign words (or signs), that cannot be processed by the interpreter (they can belong to a language that is not among the working languages of the interpreter). In these cases, interpreters often communicate that a foreign word has been pronounced, and if they realize what language it is, they can provide this information. On the other hand, when foreign signs are used in an autonymous way, the interpreter (who is speaking) can direct the audience to see the sign shown by the signer (if not prevented by problems related to lag-time, positioning and contexts as mentioned above). It is important to note that signs are always visible, while words are not audible for deaf people. Therefore, it is important that the deaf audience would receive information about the formal aspects of language. For this reason, a lot of interpreters adopt a strategy that involves meta-translation glosses, such as “this sentence is in German”, “this word is in English”, “this is an idiom” etc. Thus, another crucial issue on translation provided through languages conveyed in different modalities (auditory-vocal vs. visual-gestural) correlates to the general metalinguistic information involving the channel of communication used in the source language (rhetorical and formal features, intonational aspects, emphasis added etc.), according to the role

 Language] message is meaningless or confused and cannot be reasonably accounted for or explained by another miscue type” (Cokely 1992: 88).

74

Metalinguistic References in Cross-modal Translations

that these characteristics have in the pragmatic intent of the signer/speaker, often explicating, elucidating or omitting it.

4. Conclusions Cross-modal metalinguistic translation—with special reference to simultaneous sign language interpreting—represents a challenging step in considering translatability as a universal property of the human language, despite the modality in which it is conveyed. Though the translation of metalinguistic references requires special adaptation and evaluation, crossmodality strategies in signed interpreted-mediated interactions are often effective in making information accessible in another medium such as the visual-gestural modality. The particular status of sign language interpreting as a cross-modal translation requires a deep discussion of theoretical issues related to sign language structures and units of analysis. In fact, a better understanding of how different pragmatic and discursive functions and information are conveyed using sign language can allow to discuss the effectiveness of translation in a cross-modal perspective. Furthermore, many interpreting strategies are modelled on sign language discourse strategies. Thus, it is possible that sign language studies can provide crucial information on translation of reflexivity and on other properties of human language through a deeper investigation of signed discourse and its features. Deaf communities are nowadays completely aware of the linguistic status of sign languages, and due to video-recording and video-sharing technologies, they can share information about their languages in a metalinguistic perspective (Fontana et al. in press). Thus, a better understanding of metalinguistic references and of reflexive remarks in signed discourse could also improve our understandings of metalinguistic awareness processes in signing community. Interpreters also have an essential role in these processes, and their work should be considered in different perspectives in order to identify the modality-dependent factors. Finally, further research is needed to clarify the definition of metalinguistic references in sign languages as saying-by-showing operations, already identified as a highly salient feature of the signed discourse and as a challenge in sign language interpreting (Petitta and Del Vecchio 2011). In sign languages, the description of autonymy as a displaying strategy in terms of word-showing and sign-showing is strictly related with a functional and semiological view of languages, and with a definition of multimodality as a pervasive feature of language. However, in interpreter-mediated interactions multimodality is conditioned because

Giulia Petitta

75

multimodal references, despite their linguistic or non-linguistic nature, are not directly conveyed through participants, and the interpreter should make choices in order to provide the information originally conveyed in a multimodal way. The necessity of mediating cross-modal information through a multimodal medium as sign language is undoubtedly a further challenge in interpreting studies. In this framework, taking into account the relationships among visualgestural modality, simultaneous interpreting, and metalinguistic uses in a cross-modal perspective can allow to discuss the nature of language brokering, and how it can be provided despite the strong differences that languages present at a linguistic, social, historical, and pragmatic level. This study will help to define limits, constraints and possibilities of translation.

Acknowledgements I wish to express my gratitude to Isabella Chiari and Olga Capirci for their comments on earlier drafts of this paper; to Silvia Del Vecchio, Alessio Di Renzo, and Virginia Volterra, who provided suggestions and criticisms at various stages of this work, and to Stephanie Feyne and Brenda Nicodemus for many insightful and helpful observations.

References Adam, Robert. 2012. “Language contact and borrowing.” In Sign Language: An International Handbook, eds. Roland Pfau, Markus Steinbach, Bencie Woll, 841-62. Berlin: De Gruyter Mouton. Authier-Revuz, Jaqueline. 1995. Ces mots qui ne vont pas de soi. Boucles réfléxives et non coïncidences du dire. Paris: Larousse. Battison, Robbin. 1978. Lexical borrowing in American Sign Language. Silver Spring, MD: Linstok Press. Brennan, Mary. 2001. “Encoding and capturing productive morphology.” Sign Language and Linguistics 4 (1/2): 47-62. Capirci, Olga and Virginia Volterra. 2008. “Gesture and speech. The emergence and development of a strong and changing partnership.” Gesture 8 (1): 22-44. Cokely, Dennis. 1992. Interpretation: a sociolinguistic model. Burtonsville, MD: Linstok Press. Cuxac, Christian. 2000. La langue des signes française (LSF): les voies de l’iconicité. Faits de langues 15/16. Special Issue.

76

Metalinguistic References in Cross-modal Translations

Cuxac, Christian, and Elena Antinoro Pizzuto. 2010. “Émergence, norme et variation dans les langues des signes: vers une redéfinition notionnelle.” Langage et société 131: 37-53. De Brabanter, Philippe. 2010. “Constraints in metalinguistic anaphora.” In Constraints in discourse 2, eds. Peter Kühnlein, Anton Benz, and Candace L. Sidner, 141-62. Amsterdam: Benjamins. De Mauro, Tullio. 1994. Capire le parole. Bari: Laterza. De Meo, Anna. 2010. “Tradurre gli autonimi con funzione metalinguistica: gli esempi nei testi di linguistica.” In Ginevra-Napoli Naples-Genève. Quaderno di letteratura, lingua, cultura, ed. Giovannella Fusco Girard, 83-93. Napoli: Università “L’Orientale”. Duncker, Dorthe. 2012. “‘What is called?’ – Conventionalization, glossing practices, and linguistic (in)determinacy.” Language and Communication 32: 400-19. Emmorey, Karen, ed. 2003. Perspectives on Classifiers Constructions in Sign Languages. Mahwah, NJ: Erlbaum. Ferreri, Silvana. 2012. “Metalinguisticità riflessiva: statuto teorico e potenzialità d’uso.” In Per Tullio De Mauro, eds. Anna M. Thornton and Miriam Voghera, 107-27. Roma: Aracne. Fontana, Sabina. 2013. Tradurre lingue dei segni. Modena: Mucchi. Fontana, Sabina, Serena Corazza, Penny Boyes-Braem, and Virginia Volterra. In press. “Language research and language community change: Italian Sign Language 1981-2013.” International Journal of Sociology of Language. Friedman, Lyn, ed. 1977. On the other hand: New Perspectives on American Sign Language. New York: Academic Press. Hirsh-Pasek, Kathy. 1987. “The Metalinguistics of Fingerspelling: An alternate way to increase reading vocabulary in congenitally deaf readers.” Reading Research Quaterly, 22 (4): 455-74. Humphries, Tom, and Francine MacDougall. 2000. “‘Chaining’ and Other Links: Making Connections between American Sign Language and English in Two Types of School Settings.” Visual Anthropology Review 15 (2): 84-94. Jakobson, Roman. 1959. “On Linguistic Aspects of Translation.” In On Translation, ed. Reuben A. Bower, 232-9. Cambridge, MA: Harvard University Press. Johnston, Trevor, and Adam Schembri. 1999. “On defining lexeme in a sign language.” Sign Language and Linguistics 2 (1): 115-85. Källkvist, Marie. 2013. “Languaging in translation task used in a university setting: particular potential for student agency.” The Modern Language Journal 97 (1): 217-38.

Giulia Petitta

77

Kelly, Arlene B. 1995. “Fingerspelling Interaction: A Set of Deaf Parents and Their Deaf Daughter.” In Sociolinguistics in Deaf Communities, ed. Ceil Lucas, 62-73. Washington, DC: Gallaudet University Press. Kendon, Adam. 2004. Gesture. Visible action as utterance. Cambridge: Cambridge University Press. Klima, Edward, and Ursula Bellugi, eds. 1979. The Signs of Language. Cambridge, MA-London: Harvard University Press. Leeson, Lorraine. 2005. “Making the effort in simultaneous interpreting: Some considerations for signed language interpreters.” In Topics in Signed Language Interpretation, ed. Terry Janzen, 51-68. Amsterdam: John Benjamins. Lucy, John, ed. 1993. Reflexive Language. Reported Speech and Metapragmatics. Cambridge: Cambridge University Press. McNeill, David. 1992. Hand and mind: What gestures reveal about thought. Chicago: University of Chicago Press. —. 2005. Gesture and Thought. Chicago: University of Chicago Press. —. 2012. How language began: Gesture and speech in human evolution. Cambridge: Cambridge University Press. Metzger, Melanie. 1999. Sign language interpreting: Deconstructing the mith of neutrality. Washington, DC: Gallaudet University Press. Meurant, Laurence, and Brigitte Garcia. 2010. “Signing about signing. Sign metalanguage in LSF and LSFB.” Paper presented at Theoretical Issues in Sign Language Research 10. West Lafayette, Indiana, USA, September 30-October 2, 2010. Mulrooney, Kristin J. 2002. “Variation in ASL Fingerspelling.” In Turntaking, Fingerspelling, and Contact in Signed Languages, ed. Ceil Lucas, 3-23. Washington, DC: Gallaudet University Press. Napier, Jemina. 2004. “Interpreting omissions: A new perspective.” Interpreting 6 (2): 117-42. Nicodemus, Brenda, and Karen Emmorey. 2013. “Direction asymmetries in spoken and signed language interpreting.” Bilingualism: Language and Cognition 16 (3): 624-36. Nicodemus, Brenda, and Laurie Swabey, eds. 2011. Advances in Interpreting Research. Amsterdam: Benjamins. Padden, Carol and Darline Clark Gunsauls. 2003. “How the alphabet came to be used in a Sign Language.” Sign Language Studies 4 (1): 10-33. Paligot, Aurore, and Laurence Meurant. 2013. “Register variation in LSFB: The influence of the metalinguistic function.” Paper presented at Theoretical Issues in Sign Language Research 11. London, July, 1013 2013.

78

Metalinguistic References in Cross-modal Translations

Petitta, Giulia, and Alessio Di Renzo. 2013. “Fingerspelling and metalinguistic reference: New data from Italian Sign Language (LIS) discourse.” Paper presented at Theoretical Issues in Sign Language Research 11. London, July, 10-13, 2013. Petitta, Giulia, and Silvia Del Vecchio. 2011. “Implicazioni linguistiche e strategie traduttive nell’interpretazione dalla LIS all’italiano e viceversa: una prospettiva semiotica.” Rivista di psicolinguistica applicata 11 (3) 2011: 47-65. Pizzuto, Elena. 2003. “Coarticolazione e multimodalità nelle lingue dei segni: dati e prospettive di ricerca nello studio della lingua dei segni italiana (LIS).” In La Coarticolazione. Atti delle XIII Giornate di Studio del Gruppo di Fonetica Sperimentale (A.I.A.), eds. Giovanna Marotta and Nadia Nocchi, 59-77. Pisa: ETS. Puente, Anibal, Jesus M. Alvarado, and Valeria Herrera. 2006. “Fingerspelling and Sign Language as Alternative Codes for Reading and Writing words for Chilean Deaf Signers.” American Annals of the Deaf, 151 (3): 299-310. Quinto-Pozos, David, and Robert Adam. 2013. “Sign Language Contact.” In The Oxford Handbook of Sociolinguistics, eds. Robert Bayley, Richard Cameron, and Ceil Lucas, 379-400. Oxford-New York: Oxford University Press. Ramat, Paolo. 2007. “The Problem of (Un)Translatability.” R.I.L.A., Rivista Italiana di Linguistica Applicata 39 (1-2): 21-32. Rey-Debove, Josette. 1978. Le métalangage. Etude linguistique du discours sur le langage. Paris: Le Robert. Roibu, Melania. 2010. “On Autonymy (with reference to Romanian).” Revue Roumaine de Linguistique 55 (1): 69-85. Saka, Paul. 2013. “Quotation.” Philosophy Compass 8 (10): 935-49. Schembri, Adam and Trevor Johnston. 2007. “Sociolinguistic Variation in the Use of Fingerspelling in Australian Sign Language: A Pilot Study.” Sign Language Studies 7 (3): 319-47. Sutton-Spence, Rachel. 1998. “Grammatical Constraints on Fingerspelled English Verb Loans in BSL.” In Pinky Extension and Eye Gaze: Language Use and in Deaf Communities, ed. Ceil Lucas, 41-58. Washington, DC: Gallaudet University Press. Sutton-Spence, Rachel, and Bencie Woll. 1993. “The Status and Functional Role of Fingerspelling in BSL.” In Psychological Perspectives on Deafness, vol. 1, eds. Mark Marschark and M. Diane Clark, 185-208. Hillsdale, NJ: Erlbaum. Taylor, Talbot J. 2000. “Language constructing language: the implications of reflexivity for linguistic theory.” Language Sciences 22: 483-99.

Giulia Petitta

79

Vermeerbergen, Myriam. 2006. “Past and current trends in sign language research.” Language and Communication 26: 168-92. Vermeerbergen, Myriam, Lorraine Leeson, and Onno Crasborn, eds. 2007. Simultaneity in Signed Languages: Form and Function. Amsterdam: Benjamins.

PART II: THEORIES OF TRANSLATION

TRANSLATING METAPHORS MICHELE PRANDI UNIVERSITY OF GENOA

There are many different kinds of metaphor, with different grammatical, conceptual and semantic properties. Each of them raises specific problems to the translator. The purpose of this article is to draw attention to these different kinds of metaphor and to their properties, and to underline those that could be a challenge for translation.

When I taught at the High School for Translation and Interpreting of the University of Bologna at Forlì, I happened to hear one colleague, a teacher of simultaneous interpreting, give his students the following instruction: “Kill the metaphor”. Owing to the pervasiveness and functional engagement of metaphors in linguistic expression, which end up by undermining the (metaphorical) instruction itself, I think that this idea is not only absolutely wrong but also out of reach. Contrary to hasty conclusions, however, the main reason for my disagreement does not reside in the cruel verb to kill, but rather in the innocent singular form of the direct object, metaphor. The use of the singular form strongly suggests that it should be a unitary object. By contrast, there are many different kinds of metaphor, with different grammatical, conceptual and semantic properties. Each of them creates specific problems to the translator. I am not a translator, and my contribution to the discussion does not lie in explaining how to translate or not to translate metaphors. My purpose is simply to draw the attention to the different kinds of metaphor, to bring into focus their different grammatical, conceptual and semantic properties, and to underline those that could be a challenge for translation. My idea is that linguistic description can help translation at the level of a fine-grained analysis of the structure and content of the expressions that form the source text—at the level of what I name the “works preliminary to translation” (Prandi 2007).

84

Translating Metaphors

1. Metaphor: a general definition What is absolutely amazing about metaphor is the great number of different and even opposite theories that have been put forward during the centuries. In the history of thought, metaphor has been defined both as a transfer of a word into an alien conceptual domain (Aristotle, Poetics) and as an extension of word meaning (Dumarsais 1730); as a strange substitute for a proper word (Fontanier 1968; Genette 1968; Groupe ȝ 1970; Todorov 1970); as a way of putting strange concepts into interaction (Richards 1936; Black 1954; Coseriu 1967); as a system of shared and non-dispensable concepts at the service of consistent thought1 (Blumenberg 1960; Weinrich 1958, 1964; Lakoff and Johnson 1980; Lakoff and Turner 1989; Gibbs 1994) and as a textual interpretation of a conflictual complex meaning that challenges the consistent thought (Weinrich 1963, 1967; Prandi 1992, 2004). None of these ideas of metaphor is completely false, and none is completely true. Each of them is supported by some relevant data; none of them is compatible with all data. This proliferation of definitions is the consequence of an inadequate perspective on metaphor. The metaphorical process has a unique spring, that is, conceptual transfer and interaction, but is open to many different, and even opposite, outcomes. On these grounds, each theory that focuses on one outcome is condemned to be partial, and therefore to be torn between inadequacy and inconsistency. For instance, the idea that metaphors are textual interpretations of conceptual conflicts is true when applied to living metaphors, that is, to non-conventional, creative metaphors to be interpreted at text level, and false if extended to conventional metaphors whose content is consistent and coded in lexical structures. Therefore, it cannot be generalised into an adequate theory of metaphor. In order to build up a unitary and unifying theory, one has to move the focus from the heterogeneous, even opposite outcomes, onto the unique spring. On these grounds, the different issues are easy to account for. 1

Consistency is a negative property of a complex meaningful expression— typically, of a sentence. It may be defined as the absence of conflict between the concepts involved in a given relation. Mary smiles, for instance, has a consistent meaning; The moon smiles (Blake), by contrast, has an inconsistent meaning, for the moon is a kind of being that is not admitted to occupy a role of experiencer within a process such as smiling. Conventional metaphorical extensions of meanings are consistent in an obvious way: since embrace has an extended meaning applied to abstract entities, for instance, to embrace an idea is not a conflictual combination.

Michele Prandi

85

The origin is one and the same for all kinds of metaphors, and the best way to illustrate this point is to quote a well-known metaphorical definition. According to Geoffroy de Vinsauf (12th-13th century), a metaphor is a sheep that has jumped over the fence into an alien meadow: “propria ovis in rure alieno”. The adventure has one beginning—the jump over the fence—but is open to many developments. It is possible that the intruder, worried about the consequences of its act, will jump back to its own territory. If it decides to remain and face the indigenous beasts, it may submit and negotiate a peaceful hospitality, or fight to impose its conditions, to a lower or higher degree. At the basis of any metaphor, there is the transfer of a concept into a strange conceptual area, which necessarily ends in conceptual interaction between strange concepts. Whereas the transfer is one and the same, conceptual interaction is open to different, even opposite outcomes.

2. Conceptual transfer and interaction Metaphor is the only figure that transfers a concept into an alien conceptual area. Contrary to current assumptions (see for instance Bonhomme 1987, 2006; Barcelona 2000, Kövecses 2002), metonymy does not involve conceptual transfer. This point can be illustrated by analysing an ambivalent sentence such as And the mountain-peaks are asleep (Alcman),2 which is open to both a metonymic and a metaphorical interpretation. The sentence contains a conflictual meaning because the verb sleep has been displaced from the realm of living creatures into the alien conceptual domain of inanimate things. The transfer of a word, however, is far from implying that a concept is transferred—namely, that the process of sleeping is transferred onto the mountains. If a metonymic interpretation is chosen, sleeping is not applied to the mountains, but to the animate beings that live on them. Consequently, the concept of sleeping remains firmly anchored in the realm of living creatures, while the mountains remain inanimate things. Once the two heterogeneous and potentially conflicting concepts—the living beings and the inanimate mountains—are connected by a relationship, the conflict is dismantled and the transfer is blocked: once it is applied to the animate beings that live on mountains, in 2

Alcman, Frgm. Diehl 58, in Greek Lyric, Vol. II. Actually, the initial And does not exist in Alcman’s text which begins with İ੢įȠȣıȚ į’ੑȡȑȦȞ țȠȡȣijĮí IJİ țĮ੿ ijȐȡĮȖȖİȢ, but I stick to David A. Campbell’s translation in the Loeb Classical Library (London: William Heinemann LTD – Cambridge/Mass.: Harvard University Press, 1973: p. 454-455).

86

Translating Metaphors

particular, sleeping can no longer be predicated of mountains. By contrast, if Alcman’s line is interpreted as a metaphor, a concept is really transferred into a strange domain: in the quoted example, sleeping receives an inanimate subject—namely, the mountains: therefore it is really transferred onto inanimate nature. Owing to the transfer, heterogeneous concepts interact: a given state of the mountains is seen as if it were a kind of sleeping, and mountains are seen as if they were living beings.3

2.1. Conceptual interaction as an algebraic magnitude: reduction, substitution, projection The best way of approaching conceptual interaction is the observation of conflictual figures. An example of conceptual conflict is provided by the following lines of a poem by the Italian poet Giovanni Pascoli (‘Il giorno dei morti’, Myricae): A ogni croce roggia pende come abbracciata una ghirlanda donde gocciano lagrime di pioggia

From every rust-red cross Hangs, as if embracing it, a garland From which drop tears of rain

Since the expression tears of rain is used to refer to raindrops, the concept of tear and the concept of drop are in competition for determining those aggregates of water falling from sky when it rains. The concept of drop is the tenor (Richards 1936), it has the privilege to be both the relevant textual topic and the shared guardian of the conceptual identity of the thing. The concept of tear is the subsidiary subject (Black 1954),4 that is, the intruder that challenges the identity of the tenor. When a competition 3

The fact that one and the same meaningful sentence can be interpreted either as a metaphor or as a metonymy shows that the figure is not encapsulated in the meaning of the sentence but is the outcome of an act of textual interpretation inspired by textual relevance criteria. 4 A brief terminological clarification is needed. Frame and focus (Black 1954) identify the overt constituents of the conflicting expression, whereas tenor (Richards 1936) and subsidiary subject (Black 1954) refer to the concepts that interact in metaphor. The two couples of concepts do not overlap because both tenor and subsidiary subject may be covert as well as overt constituents of the expression. Our choice to use the couple tenor vs. subsidiary subject, which picks one term from Richard’s couple (vehicle vs. tenor) and the other from Black’s one (primary subject vs. subsidiary subject) is motivated by the same reason: both primary subject and vehicle strongly suggest that these concepts are overt constituents of the conflicting expression, which is not necessarily the case.

Michele Prandi

87

takes place, we can imagine two opposite issues and a null balance. In the same way, the balance of the interaction between a subsidiary subject and a tenor is an algebraic magnitude, which is open towards a negative, a null and a positive balance. The positive balance of interaction is projection: the conceptual profile of the tenor is both challenged and reshaped under the pressure of the subsidiary subject. As an empirical datum, projection is a graded magnitude. Projection can stop after the first step, as in the case of a stereotyped interpretation, as well as run ahead, towards the most unpredictable outcomes. When faced with the conflictual concept of tears of rain, for instance, one can simply conclude that raindrops and tears are both made of water and look very similar—and stop there. Recovering a trivial analogy is the most direct way of stopping projection immediately after its start. However, one can also go on. If there are tears, one may think, someone is crying. Who? Probably, Nature. If Nature is crying, there must be some reason. The poem’s title, Il giorno dei morti, reminds us of All Souls Day. Maybe Nature is sympathetic to the suffering of human beings. Then, Nature is no longer the cruel stepmother described by another Italian poet, Giacomo Leopardi, but a sympathetic mother, who shares her children’s pain—and the reader might go further along this path. The trivial analogy between tears and raindrops we started from looks very far away.5 Simple substitution can be seen as a null balance of interaction: in the presence of a substitutive option, interaction is blocked before it starts. If, after realising that the expression tears of rain refers to raindrops, the reader simply restores the tenor, namely, raindrops, at the expense of the subsidiary subject, namely, tears, no interaction is triggered. The negative balance of interaction gives birth to reduction, documented by lexical catachresis.6 The subsidiary subject is wholly 5

The actual interpretation of a metaphor in a given text is a pragmatic process governed by a criterion of relevance that is the same that motivates the interpretation of any utterance (Sperber and Wilson 1986ab; 2008; Prandi 2004: Ch. 1). Contrary to Sperber and Wilson’s claim, however, this does not imply that metaphorical projection does not have a conceptual structure of its own. Relevance is a contingent criterion, bound to a given text or situation; as a conceptual structure, projection has a long-lasting structure. As far as consistent metaphorical concepts are concerned, this structure is documented by a shared repository of lexical values; in the presence of conflictual metaphors, the conceptual structure of projection is circumscribed by the consistent conceptual environment of the subsidiary subject. 6 The influent French rhetorician Fontanier (1968: 213) names catachresis (catachrèse) a metaphorical or metonymic meaning extension leading to a new

88

Translating Metaphors

adapted to the conceptual identity of the tenor. Instead of enriching the tenor under the pressure of the subsidiary subject, a lexical catachresis prunes the subsidiary subject, so to speak, to the point of full consistency with the tenor. An instance is the concept of wing of a building. Instead of launching the building up into the sky, its metaphorical wings lose the aptitude for flight.

2.2. Consistent and conflictual metaphors A distinction that cuts across our first typology based on the balance of interaction has to do with an essential property of the metaphorical content—namely, conceptual consistency: there are metaphors that have a consistent content (Weinrich 1958; Lakoff and Johnson 1980), and there are metaphors that challenge our most deeply rooted conceptual structures (Weinrich 1967; Prandi 2012). Consistent metaphors are well integrated in our shared ways of speaking and thinking. For instance, when we speak of the wing of a castle or of a wish that is cherished, we express ourselves in the most natural way. We are hardly aware of using metaphors. Conflictual metaphors are textual interpretations of complex meanings that contain a conceptual conflict: for instance And Winter pours its grief in snow / When Autumn’s leaves are lying (Emily Brontë, ‘I know tonight the wind is sighing’). Within our shared natural ontology, winter is not conceived of as a human being that can feel grief; besides, grief is not a concrete liquid substance to be poured. In metaphors of this kind, traditionally named living metaphors (Ricœur 1975) the conceptual conflict is overt—one is fully aware of it. However, the conflict is not felt as a fault, but as a resource and an occasion—the proof that the structure of the expression contains the seeds of conceptual creation.

meaning of an old word: “Catachresis is the affectation of an already significant word to a new idea devoid of its own expression in the considered language”. I find this concept very useful because it allows us to draw a clear distinction between regressive and projective lexical extensions. As we shall see in the following lines, catachreses remain isolated and non-productive, whereas extensions depending on living metaphorical concepts are ready to form rich interconnected networks of new uses of old words.

Michele Prandi

89

2.3. Kinds of consistent metaphors .

Beside regressive catachreses, the heritage of consistent metaphors at the disposal of a linguistic community contains an impressive wealth of metaphors that are both consistent and projective. Moreover, a good amount of consistent concepts that circulate in a linguistic community are not anonymous products of a kind of spontaneous generation, but spring out of individual acts of creation based on the linguistic construction of conceptual conflicts. Catachresis—for instance, the noun wing applied to a building—is simply a way of expressing a familiar concept using an alien word. In the presence of catachresis, the tenor is well known independently of its metaphorical label, so that its independent conceptual identity governs the path of metaphorical extension. The absence of any active conceptual pressure on the tenor blocks projection, and therefore condemns the extension to isolation. Isolation and non-productivity, in turn, are a precondition of death. The most significant instances of consistent metaphors are not provided by isolated catachreses, but by shared metaphorical concepts of the kind focused on by Lakoff and Johnson (1980). Unlike catachreses, shared metaphorical concepts are grounded in active and productive schemes of thought, and tend to form active and productive relational networks. As Gibbs (1994: 277) points out, metaphorical concepts prove that “what is conventional and fixed need not be dead”. As in the presence of conflictual metaphors, projection is not an isolated transfer, but has the structure of a complex conceptual swarm: a whole system of interconnected concepts is ready to provide a conceptual model for categorising a whole target domain. A telling example is provided by Vendler (1970: 91): “Take belief. It is like a child: it is conceived, adopted, or embraced; it is nurtured, held, cherished and entertained; finally, if it appears misbegotten, it is abandoned or given up. The same is true, with lesser variety, of thoughts, suspicions, intentions and the like”. Consistent and conflictual metaphors do not belong to completely separated domains, but are bridged by a territory of concepts that, though consistent and shared, spring out of acts of individual creation based on conflict. These concepts do not form part of the anonymous repository of the common language, but belong to specialised areas of expression, to philosophy and science in the first place, but also to more down-to-earth terminological repertories. A good example of consistent concept born out of a conceptual conflict is Kuhn’s idea of scientific revolution (Kuhn 1962), belonging to the technical lexicon of epistemology. This metaphor challenges the traditional assumption of a linear progress of scientific

90

Translating Metaphors

research by evocating the conflicting and traumatic idea of political revolution, and encourages us to apply the most qualifying properties of political revolutions to the way sciences develop in history. Nowadays, the insights encouraged by conflict and projection have been integrated into a firm and consistent concept accepted by the scientific community, and the metaphor has become one of the conceptual tools of epistemology as a distinct and technical meaning of the word revolution.

2.4. Linguistic meanings and textual interpretations A catachresis is one sense7 of a polysemous word: for instance, wing as side appendix of a building. A metaphorical concept is a structure of thought documented by more or less rich swarms of interconnected extended meanings of words. The use of cherish with hope as direct object, for instance, is a distinct sense of a polysemous word, which documents the metaphorical concept EMOTIONS ARE CHILDREN, OR DEAR PERSONS. Metaphorical concepts also justify the use of complex expressions as idioms: the idea that money is liquid, for instance, motivates both the extended sense of pour, cf. The US continued to pour money into the South to keep it in power (British National Corpus); and the idiomatic use of the expression throw down the drain with money: I’m worried about the security situation in Afghanistan and might be money just has been thrown down the drain (EPIC).8 The contents of consistent metaphors (both catachreses and lexical extensions motivated by metaphorical concepts) are coded and shared meanings of words or phrases, registered as such in dictionaries. Metaphor is encapsulated in these meanings. The content of a conflictual metaphor is not the meaning of the conflictual expression, but one contingent and reversible message that can be interpreted from this meaning within a text or a communicative situation. This point is well illustrated, once again, by Alcman’s line And the mountain-peaks are asleep. The meaning of the expression is conflictual but in no way ambiguous or indeterminate: it univocally attributes the state of sleeping to mountains. 7

The term sense is used in structural semantics to denote each distinct meaning of a polysemous lexeme, and occurs in the phrase sense relations (Lyons 1977; Palmer 1976; Cruse 1986, who also speaks of sense spectrum). 8 The multi-lingual corpus EPIC was created by a multidisciplinary research group coordinated by Mariachiara Russo at the Department of Interdisciplinary Studies on Translation, Language and Culture (SITLeC) of the University of Bologna at Forlì (Spinolo and Garwood 2010).

Michele Prandi

91

At one and the same time, it opens towards many different interpretative paths. Some of them are figurative and some are not. In the first place, the expression admits a literal interpretation, referring to an alien world crowded with living mountains. Such is, for instance, the standard interpretation of Phaedrus’ fables, where animals and even trees do actually speak. For a figurative interpretation to be activated, the expression must apply to our shared world, where, according to our silently accepted natural ontology, mountains do not sleep. Among figurative options, metaphor is in competition with metonymy, as we have remarked above.9 If one meaning is compatible with literal, metaphorical and metonymic interpretations, it does not make sense to read the figure in the meaning of the expression itself. When one speaks of metaphorical meaning, and more generally of literal, non-literal and figurative meaning, one simply confuses the meaning of the expression with one of its allowed contingent interpretations.

3. Implications for translation We are now approaching the main question of translation: what is actually translated during the translation process? A meaning, that is, a conceptual structure shaped in a specific way by a linguistic expression, or an interpretation, that is, the outcome of an inferential process triggered by this meaning within a contingent use and made by a person? Our more specific question—the translation of metaphors—has to do precisely with this point: some metaphors surface in texts as distinct lexical meanings of polysemous words or idioms, whereas others are interpretations of complex meanings of whole sentences. Thus, when dealing with metaphor translation, we have to face a double opposition: between meanings and interpretations on the plane of the content, and between words and sentences on the plane of expression.

9

It is interesting to compare the virtual set of options available for the isolated sentence and their relevance within the whole text. If the sentence And the mountain-peaks are asleep is located within the co-text of the whole fragment, in particular, the metonymic interpretation is no longer relevant, for sleeping is explicitly attributed to living beings themselves, and in particular to “the wild animals of the mountains”: And the mountain-peaks are asleep and the ravines, the headlands and the torrent-beds, all the creeping tribes that the black earth nourishes, the wild animals of the mountains, the race of bees and the monsters in the depths of the surging sea; and the tribes of long-winged birds are asleep.

92

Translating Metaphors

It is at this point that the distinction between conflictual and consistent metaphors, which is connected with the distinction between meanings and interpretations, bears an immediate consequence on translation. When one speaks of translation of metaphors, the presupposition is that the object of the translation process is always a metaphor. Now, this presupposition has to be called into question: since some metaphors are meanings (of words) while others are interpretations (of sentences), it is clear that either we do not always translate meanings or we do not always translate metaphors. When a polysemous word or an idiom motivated by a metaphorical extension is translated, what has to be translated is both a metaphor and a meaning, that is, a metaphorical meaning. Now, whether a given metaphorical meaning has an exact equivalent in the target language is an empirical question. It may happen that the equivalent expression in the target language is the same metaphor, or a different metaphor, but it may also happen that the metaphor gets lost in translation. In other words, it may happen that the metaphor is killed, so to speak. The killing, however, is not a choice made by the translator, but depends on the lexical structure of the target language. Let us consider some significant examples. The French expression porte condamnée has an exact metaphorical equivalent in English. Owing to this, the translation word by word— condemned door—both preserves the source metaphor and meets a shared and conventional English lexical structure. Of course, the translation word by word is not a choice made by the translator, but the outcome of a perfect match between the lexical structures of the involved languages. The English idiom to throw money down the drain has a metaphorical equivalent in Italian, but the metaphor is different: in Italian, money is thrown dalla finestra—out of the window. The metaphor is not killed in translation, but simply changed. Once again, this is not a choice made by the translator, but the outcome of a mismatch between the lexical structures of the involved languages. In Italian, the French expression porte condamnée and its English equivalent condemned door have no metaphorical equivalent. Owing to this, a translation word by word—porta condannata—would be misinterpreted as a living, creative metaphor open to free projection: the Italian addressee would see the door as a human being who has been convicted, judged and condemned for some crime. This is the reason why the French and English metaphors are killed, so to speak, by an Italian standard translation: for instance, porta sbarrata, ‘barred door’. Once again, the killing is not a choice made by the translator, but the inevitable outcome of a mismatch between the lexical structures of the involved languages. For the same reasons, the French combination ténaillé par le

Michele Prandi

93

désir cannot be translated literally into English as clasped with pincers by desire, but, for instance, as devoured by desire. The same happens with metonymies: the French metonymic idiom casser la pipe cannot be translated as break the pipe, but, for instance, as kick the bucket—which is in turn a metonymy—or simply as die. The difficulties in translating lexicalised consistent metaphors depend on the different ways lexical structures are organised in different languages, that is, on anisomorphism, a phenomenon highlighted by structural semanticists (Trier 1932; Hjelmslev 1943). Language-specific lexical structures impose a language-specific shape on shared metaphorical concepts just as they impose a language-specific shape on any kind of shared concept. As Lyons (1963: 37) points out, “each language must be thought of as having its own semantic structure, just as it has its own phonological and grammatical structure”. The fact that “languages are anisomorphic metaphorically” (Dagut 1976: 32) is simply a particular case of a general property of natural language. The reason why the French verb ténailler has no English equivalent co-occurring with desire is the same why the English noun river has two French equivalents, that is, fleuve and rivière. This in turn is the reason why the paths of metaphorical motivation are accessible a posteriori, in the presence of documented data, but cannot be predicted. Owing to anisomorphism, the distribution of metaphorical meanings even among the languages that share the same metaphorical concepts is hardly the same. This is the reason why consistent metaphors, as a rule, cannot be translated word by word. In the presence of a living, conflictual metaphor, things are different. The metaphor is not in the meaning of the conflictual expression, but is the outcome of a contingent process of consistent interpretation. This implies that meaning and metaphor have different destinies in translation: the object of translation is not the metaphor, that is, the content of a contingent interpretation, but the conflictual complex meaning that provides its conceptual input. As a rule, the conceptual conflict documented in the source expression can be immediately transferred into the target language by a word by word translation. When facing the French conflicting expression La lune rêve (Charles Baudelaire, ‘Tristesse de la lune’, Les fleurs du mal), for instance, one can simply translate it word by word: The moon dreams. The reason is clear: we do not translate the content of the metaphorical projection, which is the outcome of a chain of inferences made by the interpreter, but the meaning of the conflictual expression itself. Since its conceptual purport is shared far beyond the boundaries of the community speaking the source language, the conflict does not get lost in translation and is ready to trigger the same projections as the source

94

Translating Metaphors

expression. The natural ontology according to which the moon is not a human being allowed to dream rules the practical behaviour of Frenchmen as well as of Englishmen: neither would really ask a question, give orders or ask a piece of information to the moon. For the same reason, the addressee will be able to interpret the metaphor, drawing the relevant inferences, no matter which language one speaks. On the other hand, an individual sharing a different ontology would interpret the sentence as a literal statement independent of the language that expresses it. Let us suppose, for instance, that Athabaskans see the moon as a living being; a native Athabaskan would interpret Baudelaire’s line depicting a dreaming moon as a literal statement—as a faithful description of how things go— independently of the language that expresses it. An Athabaskan who speaks French, for instance, would interpret the original version of Baudelaire’s line as literally as its Athabascan translation. Neither consistency nor conflict are a matter of language; they are a matter of thought, and this is the reason why they are not affected by translation. Creative metaphorical concepts of philosophy and science are based on creative conflictual meanings of individual expressions. According to this premise, it is easy to predict that they will be translated on the same conditions as the conflictual complex expressions that feed poetic metaphors. The English expression scientific revolution, for instance, is translated into French as révolution scientifique, into German as wissenschaftliche Revolution, into Italian as rivoluzione scientifica, into Spanish as revolución científica, Russian nauþnoje revoliùþja, and so on. The conclusion of the previous reflection is that what is translated is always a meaning. In the case of conventional metaphors, we also translate a metaphor, that is, a coded metaphorical meaning of either a polysemous word or an idiom, which is not likely to be entrusted to the same metaphor, and even to a metaphor, across different languages. In the case of poetic metaphors, we also translate a meaning. This meaning, however, does not contain the metaphor, which is the outcome of a contingent act of interpretation. Moreover, the meaning one translates is not the lexical meaning of a word or of an idiom, but the complex and conflictual meaning of a whole sentence. Since it is a complex conceptual structure whose consistency is measured against a system of conceptual criteria shared far beyond the source language, this meaning can generally be translated word by word without losing its conceptual properties. As such, it is ready to trigger the same network of projection as the original one. These are the reasons why, contrary to a widespread opinion, the most difficult metaphors to be translated are not creative, conflictual metaphors, which are rich in content and typically designed for open-ended

Michele Prandi

95

interpretation, but consistent, conventional metaphors, documented by extended uses of polysemous words and by idiomatic uses of complex expressions. As Van den Broeck (1981: 82) points out, “There is even no exaggeration in saying that this type [that is, lexicalised metaphors, M. P.], precisely in that it is institutionalized and thus belongs to the particular system (linguistic and cultural), forms the main challenge for the translator of certain texts”. This does not imply that the translation of conflicting metaphors is exempt of risks. These risks, however, are of a completely different nature.

4. Some significant cases After examining the main point connected to the translation of metaphors, that is, the distinction between conventional and living metaphors, we shall now focus on some interesting cases. Since they are isolated extensions, catachreses are naturally expected to have an unpredictable distribution across languages. One example is the French use of the adjectival participle condamné with doors, which has an exact equivalent in English—condemned door—but not in Italian. The case of consistent metaphorical concepts is by far more interesting, for in this case the conceptual purport is generally common to different languages. As structures of thought, consistent metaphorical concepts are widely shared across languages, a fortiori if these languages belong to the same cultural tradition. In spite of this, the lexical values motivated by these metaphors are not isomorphic across languages and, as a rule, unpredictable. Even metaphorical motivation, in a sense, acts in an arbitrary way. In order to discuss this point, let us consider some metaphorical verbs that co-occur with feelings, and in particular with wish and desire, in English, French and Italian (Prandi and Caligiana 2007). The examined languages belong to a broader cultural area, which is also a “community of metaphors” (Weinrich 1958). The multifactorial metaphorical categorisation of desire, in particular, is shared by the three languages. Desire appears as a Janus-like concept: it may refer to a beloved child or person to nourish, nurse and cherish, but it can be seen also as a burning flame and as a savage beast ready to seize and devour its victim. To English nourish, nurse and cherish correspond nourrir and caresser in French and nutrire and accarezzare in Italian. To English burn, seize, enslave, entangle, govern, shatter, overcome, enthrall correspond brûler, enflammer, consumer, happer, posséder, habiter, tirailler, emporter and dévorer in French, and bruciare, ardere, infiammare, accendere, assalire, incalzare, assillare, tormentare,

96

Translating Metaphors

torturare, impadronirsi, divorare in Italian. Although the source domains are shared, the semantic networks are not exactly isomorphic. For such English verbs as gratify, obey and espouse there is no equivalent in either Italian or French; French embraser and English burn to a cinder have no equivalent in Italian; Italian has struggersi (‘melt’), an archaic verb hardly used outside the domain of feelings. Only in Italian the person devoured by desire can scalpitare (‘paw’) as an impatient horse. The idea of liquid desire is shared by English and French, but the former has pour while the latter has submerger (‘drown’). Outside these core domains, we find some idiosyncratic metaphorical concepts. In English, for instance, desire is a boat one can harbour as to protect it from storms, and can fuel action, two ideas unknown in both French and Italian. French tenailler (‘clasp with pincers’), whose Italian equivalent, attanagliare, is used with paura (‘fear’), has no equivalent in English; French tarauder is a technical term which means ‘thread’, and is said of a screw as direct object; it has no equivalent in either Italian or English. Of course, the underlying metaphorical concepts are ready to inspire the creation of new metaphorical expressions that cross the boundaries of the heritage of lexicalised meanings. If one happened to read a sentence such as I’m clasped with pincers by desire in an English text, one would consider it a living metaphor to be interpreted in a text; desire would be seen as a cruel human being armed with pincers engaged in torturing a victim. The conceptual purport of this living metaphor could be translated word by word in any language except in those languages that, like French, lexicalise this use of the verb. For the same reason, the same sentence would in turn be a bad translation of the French sentence Je suis ténaillé par le desir, for it treats a conventional meaning as if it were an act of creation. Of course, here ‘bad’ has no axiologic implication: it simply means that whereas the French original expression is a standard use of a lexicalised metaphor, its word by word translation into English would be a creative figure. The behaviour of idioms is very similar. An idiom that is transparent for a speaker of the source language is likely to be transparent for the speakers of the target language too, for the conceptual humus that feeds them is shared. Once again, however, idioms are highly language-specific. In English, one throws money down the drain; in Italian, money is thrown dalla finestra—out of the window. When the difference is very slight, idioms are even more dangerous. The EPIC corpus documents a translation of Now we can sit back on our laurels as Ora possiamo forse magari sederci sugli allori (Spinolo and Garwood 2010). But in Italian the

Michele Prandi

97

coded form is dormire—sleep—sugli allori. The verb dormire alone is also used in an extended sense to mean the same, that is, ‘to remain inactive and miss an opportunity’, as testified by the proverb Chi dorme non piglia pesci: ‘Who’s sleeping doesn’t catch any fish’. When introducing our workshop, Paolo Ramat quoted a good example done by Bernard Pottier: “[Pottier] compared four German, English, Spanish and Italian prepositional phrases, namely von Kopf bis Fuß, from head to tail, de popa a proa and da cima a fondo. These metaphoric expressions have all the same functional meaning (namely ‘completely, totally’), i.e they share the same 'tertium comparationis' and it is easy to transfer the German phrase into English, Spanish or Italian, although they are linguistically very different, as they are drawn from different cultural backgrounds” (see Ramat 2007). An interesting point in this connection is underlined by Van den Broeck (1981: 75). Some metaphors belonging to a given literary tradition, although living and by no means lexicalised, have become stereotyped within this tradition as “shared poetic metaphors”. A telling example is provided by some Homeric formulae such as rhododáctylos Éos, ‘rosyfingered dawn’. Such metaphors “belong to the restricted area of literature and are only conventional within the period, school or generation to which they belong”. Owing to this, their translation into another language inevitably acquires a semantic weight and flavour that is totally lost and absent in the original expression. If they are not translated word by word, on the other hand, a significant mark of the original text gets lost. This is one case when a faithful translation is really out of reach.10 In principle, the translation of living metaphors does not meet this kind of difficulties except in two cases: when the living metaphor awakens a conventional metaphor and when the conflict that provides the semantic purport is due to the transgression of a language-specific combinatory restriction, of a lexical solidarity (Porzig 1934; Coseriu 1967). A conventional metaphor is awakened when the primitive meaning is made textually relevant. In the conventional meaning of the idiom to be under one’s feet, for instance, there is no room for feet. In Yeats’ lines11 Tread gently, tread most tenderly, / My life is under thy sad feet, by contrast, feet are made relevant at textual level, which triggers an open metaphorical interaction. Under such conditions, an awakened conventional metaphor can be translated on the condition that the target 10

For a fine analysis of the problems raised by the translation of Homeric formulae, see Mureddu 2013. 11 From ‘Your pathway’, in Under the moon: The Unpublished Early Poetry, New York 1995.

98

Translating Metaphors

language shares the same source concept. Now, Yeats’ lines are immediately translatable into Italian, where the idiom tenere qualcuno sotto i piedi is similar enough to to be under one’s feet. When there is no direct equivalent in the target language, by contrast, the figure gets lost in translation. McGough’s 2002 de-idiomatisation of You’ve made your bed: you’ll have to lie in it gets lost in Italian, where the same meaning is conveyed by a different idiom, drawn from a different source domain, that is, Hai voluto la bicicletta. Pedala! (‘You wanted a bike: you’ll have to paddle’): Life is a hospital ward, and the beds we are put in / are the ones we don’t want to be in [...] We didn’t make our beds, but we lie in them.12 A lexical solidarity is a language-specific semantic relation that imposes language-specific restrictions on the range of arguments allowed to saturate a relational term—typically, a verb or adjective: an example is the relation between bark and dogs in English. In the presence of lexical solidarities, anisomorphism is a fatal obstacle to metaphor translation. German, for instance, has two verbs for eating: essen for human subjects and fressen for animals. Owing to these solidarities, such an expression as Hans frißt is conflictual and metaphorical in German, but not in its word by word translation into English: John is eating. The only way to preserve the projective potential of the original conflict is a reformulation: for instance, John is eating like a beast. It should be stressed on that account that the violation of a lexical solidarity does not end in inconsistency. Unlike true inconsistencies, which break largely shared conceptual structures, lexical conflicts depict states of affairs that are perfectly consistent in conceptual terms: both animals and humans have access to the action of eating. What is inappropriate is simply the lexical choice, which is governed by languagespecific criteria: in German, the appropriate verb for expressing human eating is not fressen but essen. Since it is consistent, the depicted state of affairs is open to a non-conflictual expression on condition that the appropriate lexical choice is restored. Therefore, the conflict is by definition reversible, and the metaphor subtitutive: Hans ißt (Prandi 2004: 205-212). In this case, the metaphor disappears with the conflict. Even when no linguistic or conceptual obstacle threatens it, the translation of living metaphors is not without risks. However, the risks depend now on a choice made by the translator, and in particular on his/her choice to superpose his/her own interpretation on the source 12

The Italian translator Franco Nasi chooses a radical reformulation that drops the connection between hospital beds referred to in the first lines: La vita è un reparto d’ospedale, e i letti in cui ci mettono / sono quelli in cui non vorremmo stare […] L’anima / starebbe meglio in qualunque altro posto da quello in cui è.

Michele Prandi

99

meaning. This is a particular case of a general tendency of translators to over-interpretation and explicitation (Vinay and Darbelnet 1958; BlumKulka 1986; Klaudy 1998). A significant example of this attitude in the field of living metaphors is the tendency to reformulate metaphors as similes. Since Quintilian’s famous statement—“On the whole metaphor is a shorter form of simile” (Institutio Oratoria: VIII, 6, 8)—a “comparative conception of metaphor” (Black 1954) considers that simile is a structure that makes explicit the same comparison that is implicit in metaphor. As a matter of fact, simile and metaphor are very different conceptual and linguistic tools. Simile is an explicit predication of analogy, so that the figure is encapsulated in the meaning of the expression. To the extent that “everything is like everything and in endless ways” (Davidson 1978(1984: 254)), and “any two things are similar in some respect or other” (Searle 1979: 95), a predication of analogy is by definition consistent and therefore opposes no conceptual obstacle to literal interpretation. A living metaphor, by contrast, promotes interaction and projection thanks to the construction of an inconsistent complex meaning, which is barred to literal interpretation. The figure is not encapsulated in the meaning of the expression but is the outcome of a contingent act of non-literal interpretation. Even if a simile actually were a good interpretation of a metaphor, translating a metaphor into a simile would amount to framing in words the translator’s own interpretation instead of the conflictual complex meaning. Significant examples of over-interpretation are found in Bible translations. The Latin version of Proverbs 11, 22 faithfully translates the original conflictual expression that is designed to be interpreted as a metaphor: Circulus aureus in naribus suis mulier pulchra et fatua. The same expression is turned into a simile in all the major English translations, from King James’—As a jewel of gold in a swine’s snout, so is a fair woman which is without discretion—to the The New American Standard Bible: As a ring of gold in a swine’s snout So is a beautiful woman who lacks discretion.

5. Conclusions The translation of metaphors is not a specific problem, but a specific case of more general problems. The translation of consistent metaphors, both motivated by very general conceptual structures and isolated extensions, which are documented in the lexical structure of each language, is a specific case of

100

Translating Metaphors

the more general question of anisomorphism.13 Owing to anisomorphism, that is, to the fact that meanings have different forms in different languages, it is not always possible to directly compare a meaning in the source language and a meaning in the target language, and to find an immediate equivalent in the target language. When the source meaning is an extended metaphorical sense of a polysemous lexeme the terms of the question do not change. The translator has to be aware of such semantic obstacles and make an appropriate choice. The translation of living metaphors is a special instance of the more general problem dealt with in translation studies under the label of ‘explicitation’: the translator is morally engaged in asking himself at any step whether one is translating the contents of the expressions documented in the source text or some personal interpretation of them. The general criterion is also the same, that is, a deep respect for the structures and contents of each expression present in the source text. Both adequate lexical choices and respect for the source text, however, are condemned to remain void claims if they are not supported by an equally deep awareness of the structures and contents of both the source and target language and text. Awareness is attained by knowledge. This is the reason why I think that a better knowledge of the different kinds of metaphor and of their structural and semantic variety, which belong to the works preliminary to translation entrusted to the linguist, is an essential step to improve the translator’s performance.

Acknowledgments I am grateful to my colleague Ilaria Rizzato for revising the English expression.

References Aristotle. 1973. Poetica. Engl. transl. by W. Hamilton Fyfe: The Poetics. London-Cambridge/Mass.: The Loeb Classical Library. —. 1962. De interpretatione. Engl. transl. by H. P. Cook: The Categories; On Interpretation. London-Cambridge/Mass.: The Loeb Classical Library. Barcelona, Antonio, ed. 2000 (2003). Metaphor and Metonymy at the Crossroads. 2nd ed. Berlin-New York: Mouton de Gruyter.

13

On anisomorphism from the morphosyntactic point of view see Muro (this vol.).

Michele Prandi

101

Black, Max. 1954 (1962). “Metaphor.” Proceedings of the Aristotelian Society 55: 273-94. Repr. in Models and Metaphors, ed. Max Black, 25-47. Ithaca-London: Cornell University Press. Blum-Kulka, Shoshana. (1986). “Shifts in Cohesion and Coherence in Translation.” In Interlingual and Intercultural Communication: Discourse and Cognition in Translation and Second Language Aquisition Studies, eds. Juliane House et al., 17-35. Tübingen: Narr. Blumenberg, Hans. 1960. Paradigmen zu einer Metaphorologie. Bonn: Bouvier und Co. Bonhomme, Marc. 1987. Linguistique de la métonymie. Bern: Peter Lang. —. 2006. Le discours méthonymique. Bern: Peter Lang. Chomsky, Noel A. 1965. Aspects of the Theory of Syntax. Cambridge/Mass: The M.I.T. Press. Coseriu, E. (1967). “Lexikalische Solidaritäten.” Poetica 1, 293-303. Cruse, Alan D. 1986. Lexical Semantics. Cambridge: Cambridge University Press. Davidson, Donals. 1978 (1984). “What metaphors mean.” Critical Inquiry 5. Repr. in Inquiries into Truth and Interpretation, ed. Donald Davidson, 245-63. Oxford: Clarendon Press. Dumarsais, César Chesneau. 1730 (1988). Des tropes, ou des différents sens. Paris: Flammarion. Fontanier, Pierre. 1968. Les figures du discours. Paris: Flammarion. Genette, Gérard. 1968. “Introduction.” In Les figures du discours, ed. Pierre Fontanier, 5-17. Paris: Flammarion. Gibbs Raymond W. 1994. The Poetics of Mind. Cambridge: Cambridge University Press. Gibbs, Raymond W., ed. 2008. The Cambridge Handbook of Metaphor and Thought. Cambridge: Cambridge University Press. Groupe ȝ. 1970. Rhétorique générale. Paris: Larousse. Hjelmslev, Louis. 1943 (1961). Omkring sprogteoriens grundlaeggelse. Copenhagen. Engl. transl.: Prolegomena to a Theory of Language. Madison: The University of Wisconsin Press. Husserl, Edmund. 1901 (1970). Logische Untersuchungen. Band 1: Halle 1900; Band II: Halle 1901. Rev. edition: Band 1 (Prolegomena), Band II, 1° Teil (Res. I - V), Halle 1913; Band II, 2° Teil (Res. VI), Halle 1921. Critical edition: Husserliana, Vol. XVIII (1975), XIX/I-II (1984). The Hague: Nijoff,. Engl. transl.: Logical Investigations. London: Routledge & Kegan Paul. Jäkel, Olaf. 1999. “Kant, Blumenberg, Weinrich: Some forgotten contributions to the cognitive theory of metaphor.” In Metaphor in

102

Translating Metaphors

Cognitive Linguistics, ed. Raymond W. Gibbs and Gerard J. Steen, 928. Amsterdam-Philadelphia: John Benjamins. Jacobson, Roman. 1966. “A la recherche de l’essence du langage.” Diogène 51: 22-38. Kant, Immanuel. 1763 (1992). Versuch den Begriff der negativen Größen in die Weltweisheit zu führen, Königsberg. Repr. in Immanuel Kants Werke, Band II, Berlin 1912. Engl. transl.: “Attempt to introduce the concept of negative magnitudes into philosophy.” In The Cambridge Edition of the Works of Immauel Kant, Theoretical Philosophy, 17551770. Cambridge: Cambridge University Press. Karttunen, Lauri. 1969. “Discourse Referents.” Working Paper, International Conference of Computational Linguistics, Stockholm. Klaudy, Kinga. 1998. “Explicitation.” In Routledge encyclopedia of translation studies, ed. Mona Baker, 80-4. London-New York: Routledge. Kövecses, Zoltán. 2000. Metaphor and Emotion: Language, Culture and Body in Human Feeling. Cambridge: Cambridge University Press. Kövecses, Zoltán. 2002. Metaphor. A Practical Introduction. Oxford: Oxford University Press. Kuhn, Thomas Samuel. 1962. The Structure of Scientific Revolutions. Chicago-London: The University of Chicago Press. Lakoff, George and Mark Johnson. 1980. Metaphors we Live by. ChicagoLondon: The University of Chicago Press. Lakoff, George and Mark Turner. 1989. More than Cool Reason. ChicagoLondon: The University of Chicago Press. Landheer, Ronald. 2002. “La métaphore, une question de vie ou de mort?” Semen 15: 25-40. Leech, Geoffrey Neil. 1969. A Linguistic Guide to English Poetry. London: Longman. Locke, John. 1689 (1975). An Essay Concerning Human Understanding. London. New ed., Oxford: Clarendon Press. Lyons, John. 1963. Structural Semantics. Oxford: Blackwell. —. 1977. Semantics. Cambridge: Cambridge University Press. McGough, Roger. 2002. Everyday Eclipses. London: Viking (It. transl. by Franco Nasi, Eclissi quotidiane, S.Giorgio a Cremano: Medusa, 2003). Mureddu, Patrizia. 2013. “Tradurre Omero.” In Tradurre è un’intenzione, ed. Nicoletta Da Crema, 31-42. Milan: Marcos y Marcos. Palmer, Frank Robert. 1976. Semantics. Cambridge: Cambridge University Press. Porzig, Walter. 1934. “Wesenhafte Bedeutungsbeziehungen.” Beiträge zur deutschen Sprache und Literatur 58: 70-97.

Michele Prandi

103

Prandi, Michele. 1992. Grammaire philosophique des tropes. Paris: Les Editions de Minuit. —. 2004. The Building Blocks of Meaning. Amsterdam-Philadelphia: Benjamins. —. 2007. “Works preliminary to translation.” In Rassegna Italiana di Linguistica Applicata XXXIX, 1-2: Voices on Translation. Linguistic, Multimedia and Cognitive Perspectives, ed. Annalisa Baicchi, 33-59. —. 2012. “A Plea for Living Metaphors: Conflictual Metaphors and Metaphorical Swarms.” Metaphor and Symbol, 27 (2): 148-70. Prandi, Michele and Elisa Caligiana. 2007. “Métaphores dans le lexique: verbes appropriés et supports de noms de sentiments.” In Verbum XXIX, 1-2, Verbes et classes sémantiques, ed. Aude Grezka and Françoise Martin-Berthet, 127-42. Quintilianus. 1920. Institutio Oratoria. Engl. transl. by Harold Edgeworth Butler, Loeb Classical Library, Cambridge/Mass.: Harvard University Press, Richards, Ivor Armstrong. 1936. The Philosophy of Rhetoric. Oxford: Oxford University Press. Ricoeur, Paul. 1975 (1978). La métaphore vive. Paris: Editions du Seuil. Engl. Transl.: The Rule of Metaphor: Multi-Disciplinary Studies in the Creation of Meaning in Language, University of Toronto Press: Toronto. Saussure, Ferdinand de. 1916 (1974). Cours de linguistique générale. Paris: Payot. Critical edition by T. de Mauro, Paris: Payot, 1972. Engl. transl.: Course in General Linguistics. London: Fontana/Collins. Searle, John. 1979. “Metaphor.” In Expression and Meaning, ed. John Searle, 76-116. Cambridge: Cambridge University Press Semino, Elena and Gerard Steen. 2008. “Metaphor in Literature.” In Gibbs (ed., 2008): 232-46. Sperber, Dan and Deirdre Wilson. 1986a. Relevance. Communication and Cognition. Oxford: Blackwell. Sperber, Dan and Deirdre Wilson. 1986b. “Façons de parler.” Cahiers de linguistique française 7: 9-41. Sperber, Dan and Deirdre Wilson. 2008. “A deflationary account of metaphor.” In The Cambridge Handbook of Metaphor and Thought, ed. Raymond Gibbs, 84-105. Cambridge: Cambridge University Press. Spinolo, Nicoletta and John Christopher Garwood. 2010. “To kill or not to kill: Metaphors in simultaneous interpreting.” FORUM, International journal of interpretation and translation VIII (1): 181-211. Sweetser, Eve. 1990. From Etymology to Pragmatics. Cambridge: Cambridge University Press.

104

Translating Metaphors

Todorov, Tzvetan. 1970 (1979). “Synecdoques.” Communications 16. Repr. Semantique de la poésie, eds. Tzvetan Todorov et al., 7-26. Paris: Editions du Seuil. Trier, Jost. 1932 (1973). “Sprachliche Felder.” Zeitschrift für Deutsche Bildung 8. Repr. in Jost Trier, Aussätze und Vorträge zur Wortfeldtheorie. Herausgegeben von A. van der Lee & O. Reichmann, 93-109. The Hague-Paris: Mouton. Van den Broeck, Raymond. 1981. “The limits of translatability exemplified by metaphor translation.” Poetics Today 2 (4): 73-87. Vendler, Zeno. 1970. “Say what you think.” In Studies in Thought and Language, ed. Joseph Cowan, 79-97. Tucson: The University of Arizona Press. Vinay, Jean-Paul and Jean Darbelnet. 1958. Stylistque comparée du français et de l’anglais. Methode de traduction. Paris: Didier. Vinsauf, Geoffrey de. 1924. “Poetria Nova.” In Les arts poétiques du XIIe et du XIIIe siècles, ed. Edmond Faral. Paris: Champion. Weinrich, Harald. 1958. “Münze und Wort. Untersuchungen an einem Bildfeld.” In Romanica. Festschrift Rohlfs, eds. Heinrich Lausberg and Harald Weinrich, 508-521. Halle: Niemeyer. —. 1963. “Semantik der kühnen Metapher.” Deutsche Vierteljahrsschrift für Literaturwissenschaft und Geistesgeschichte 37: 325-44. —. 1964. “Typen der Gedächtnismetaphorik.” Archiv für Begriffsgeschichte IX: 23-6. —. 1967. “Semantik der Metapher.” Folia Linguistica 1: 3-17.

TRANSLATION STRATEGY AND THE CONSTRUCTED READER: ITALIAN TRANSLATIONS OF CONTEMPORARY IRISH POETRY DEBORA BIANCHERI NATIONAL UNIVERSITY OF IRELAND, GALWAY

This article intends to outline some pivotal traits of the strategy most commonly adopted in Italy for the translation of poetry. While, on the one hand, textual analysis of individual poems will reveal distinctive traits of each translation, on the other it will highlight translation norms that typically apply to this specific literary genre. It will be noted that the inherent “opaqueness” of poetry as a genre hampers the pursuit of a “transparent” result, which is conventionally the goal of many prose translations. Therefore, the most accepted policy amongst Italian publishers of poetic works is one articulating a manifest “mediation” of the source text. This means that translators often attempt to decode the unfamiliar in terms that are imagined to be familiar to the target readers. Such a process will be illustrated by assessing an Italian edition of poems by Belfast-born poet Derek Mahon published by Trauben in 2000. Appraisal of this particular collection will serve to highlight the fact that the target texts are treated as merely one component of the translation project, which is meant to achieve its significance through a comprehensive approach to the book and its different parts.

This article engages primarily with Italian translations from the work of Derek Mahon, born in Belfast in 1941 and generally acclaimed as one of the most innovative Irish poets of the twentieth century. His poetry skilfully combines classical references with Northern Irish themes and postmodern sensibility. While this analysis intends to explore the practice of translation in Italy in relation to contemporary Irish poetry, the results can acquire their full significance only when juxtaposed with the dominant translation norms in the domain of prose, an area that could be further

106

Translation Strategy and the Constructed Reader

investigated on its own.1 For the purposes of this study, however, it should suffice to mention that many Italian translations of contemporary Irish novelists strive to produce texts marked by a high level of conformity to the target language grammar, syntax and idioms, thus achieving what Lawrence Venuti calls a “transparent” translation (1995). Recently, Venuti addressed the dichotomy existing between poetry translation and fiction more directly in the following statement, which also implies the close interdependence of market and reception (2011: 217): The marginality is in fact the first reason to move poetry closer to the centre of translation studies. Poetry translation attracts a narrow audience and therefore occupies a tenuous position in the process of commodification that allows other literary genres, notably the novel, to become lucrative investments on the foreign rights market.

This consideration also introduces the importance of reception as a fundamental premise of the theoretical argument made here: the translation strategy is ultimately determined by different conceptions of the intended readership of specific literary genres. For this reason the conceptual notion of “constructed reader” will be used in order to refer to the publishers’ projections and translators’ expectations of the reader’s profile, with the goal of demonstrating how a priori ideas about readership may actively affect the translation practice. It will be shown that the “constructed reader” is a function embedded within the target texts, revealed by translation choices both at a micro-level and at the macrolevel of publication format. The following analysis of Irish poetry in Italian translation will explore the most accepted policy amongst Italian publishers, showing that it is one of articulating a manifest “mediation” of the source texts, a term that indicates an attempt to bridge the gap between the source language and target language cultures. Specifically, this strategy will be assessed in terms of its capacity to construct an alternative linguistic-cultural dimension where differences can be reconciled, thus initiating a critical assessment of poetry translation that resembles Homi Bhabha’s discussion of cultural encounters in terms of Third Space (2000). The act of translation will be examined as a challenge to retain the ambivalence of the poetic voice without completely alienating the response of the target readers, as Mahon’s Italian collection L’Ultimo Re del Fuoco, published in 2000 by Turin-based Trauben, will quite clearly exemplify. Before 1

For a more in-depth critical assessment of the translation of contemporary Irish fiction into Italian see Biancheri (2013)

Debora Biancheri

107

engaging with the actual textual analyses, however, which will serve to promote a deeper understanding of the mediation issues that may mark the source texts, it is necessary, briefly, to contextualise the translation of contemporary Irish poetry into Italian. Irish poetry from the second half of the twentieth century onwards is generally brought to the Italian publishing market in the form of collections translated by Italian scholars working either in Italy or Ireland. For both established and emerging scholars, translation seems to assume the value of an act of self-affirmation as an authoritative voice in their discipline of study. Far from being an invisible art—as Venuti (1995) famously argued—it is rather the focus of attention, as it is taken as evidence of the translator’s ability and competence. After all, Venuti himself recognised poetry translation to be the most fertile ground on which to challenge the invisibility of the translator (2011: 217): Released from the constraint to turn a profit, poetry translation is more likely to encourage experimental strategies that can reveal what is unique about translation as a linguistic and cultural practice. It is the uniqueness of poetry as a form of language use that occasions any such revelations.

This is precisely what makes poetry translation distinctive from the translation of fiction, where “explicitation” devices are widely employed to meet the reader’s “loaded expectations” (Hermans 1999), thereby improving connectivity and adjusting evaluative items and other linguistic aspects to culturally appropriate levels. Nonetheless, this does not mean that the translation of poetry is immune to those “deforming tendencies” that Antoine Berman outlined in relation to the translations of novels (1992). Poetry and novels share at least the same linguistic variety and creativity, and the former is subject to similar problems of equivalence in translation, if not more intricate ones. As Nabokov commented about poetry translation (from Venuti 2000: 77): [t]he problem […] is a choice between rhyme and reason: can a translation, while rendering with absolute fidelity the whole text, and nothing but the text, keep the form of the original, its rhythm and its rhyme? To the artist whom practice within the limits of one language, his own, has convinced that matter and manner are one, it comes as a shock to discover that a work of art can present itself to the would-be translator as split into form and content, and that the question of rendering one but not the other may arise at all.

By now, the tension between form and content has been widely acknowledged, and it is more of a conundrum than a “shock.” Italian

108

Translation Strategy and the Constructed Reader

publisher Trauben partly surmounts the problem by expanding Nabokov’s idea of translation as “the whole text, and nothing but the text.” The inclusion of critical tools in the paratext (Genette 1997: 1) is fairly common in edited collections of poetry translation and, as they play a crucial role in the transmission of the significance of the source text to the target reader, they should be considered an integral part of the translation strategy. Through mediation, for instance, “clarification” is a fairly common adjustment occurring in the target texts. Berman defines this as the attempt to make clear what is not meant to be clear in the source text. This can mean that the personal interpretation of the translator, however competent and informed, is presented as the “authentic” one, since a range of other possibilities, which a poem generally leaves open, are not mentioned in his/her critical intervention. Or, more simply, these possibilities are excluded by the linguistic preferences expressed in the target text. Related to the tendency towards “clarification” is a certain reliance on “expansion,” also mentioned by Berman. It consists in the addition of explanations which, when inserted in the body of a poem, can break the rhythm and effectively diminish the strength of the poetic voice. Generally, the forms of “explicitation” that “clarification” may entail are the recovery of ellipsis, the direct expression of implicit attitudes and the strengthening of cohesive or collocation networks. Mahon’s poem “The Dawn Chorus” and the corresponding Italian translation “Coro all’Alba” (90-1) provide a preliminary illustration of how the normalisation of language, in itself an almost automatic form of “clarification,” can interfere with the rhythmic patterns of the source text. At the same time it offers some insights into the difficulty of achieving real transparency, as a mere look at the title of the poem and its translation might reveal: while the English expression “dawn chorus” is quite readily associated with the singing of songbirds at the start of a new day, the Italian expression does not have such immediate associations. Perhaps coro mattutino would have led some readers to make the connection. However, without the use of the designated Italian noun cinguettio, birds remain a mere speculation in the Italian title: one that the choice of coro all’alba does not necessarily help. This underlying reference is further neglected in the last stanza quoted below, where the simple past of the verb “to fly” is substituted by the Italian evadere, literally “to evade,” which meaning “escape” does achieve the sense of the English text and yet fails to create the same figurative allusions.

Debora Biancheri (1) It is not sleep itself but dreams we miss, Say the psychologists, and the poets too. We yearn for that reality in this. […] But, wide awake, clear-eyed with cowardice, The flaming seraphin we find untrue. It is not sleep itself but dreams we miss. Listening heartbroken to the dawn chorus, Clutching the certainty that once we flew, We yearn for that reality in this.

109

È per i sogni non per il sonno in sé, il rimpianto, dicono gli psicologi; è proprio anche dei poeti dirlo. È quella realtà cui noi aneliamo in questa. […] Ma noi, ben svegli con occhi chiari di viltà, il serafino fiammeggiante non lo si dà per vero. È per i sogni, non per il sonno in sé, il rimpianto. Mentre ascoltiamo il coro dell’alba a cuore infranto, serriamo le certezze che era in noi una volta evadere: è quella realtà che noi aneliamo in questa.

However, Giuseppe Serpillo’s review—published in Il Tolomeo 2004-5— of this translation by Roberto Bertoni, focused more specifically on the perceived failure to reproduce the formal structure of the source text, by pointing out the “extremely long lines” (2004-5: 142) of the Italian translation. Although this consideration is certainly valid, there are several other factors to take into account when translating poetry. Indeed, on closer analysis, Bertoni’s choices appear to be far from arbitrary: his text displays an exact reproduction of the repetitions in exactly the same positions as they appear in the source text. Moreover, although the rhyme between “miss” and “this” is lost due to their Italian rendition as rimpianto and questo/a, their obsessive recurrence at the end of the verse is maintained by exerting some pressure on the Italian syntax. This may be seen as an attempt to recover some of the poetic quality of the source text, as with the solitary rhyme between rimpianto/infranto. The collection contains another poem displaying similar features, “Global Village”/“Villaggio globale,” (132-133) whose translation, also by Bertoni, is similarly decried by the review. A look at the beginning of the text provides further evidence of the friction between form and content inherent to poetry and the delicate balancing acts required by translation:

110

Translation Strategy and the Constructed Reader

(b) This morning, from beyond abandoned piers Where the great liners docked in former years, A foghorn echoes in deserted sheds Known to Hart Crane, and in our vigilant beds.

Stamani dal mare al di là dei moli abbandonati, Dove attraccavano in anni passati navi passeggeri imponenti, Penetrano nelle baracche abbandonate note ad Hart Crane Entrano nei nostri letti vigili gli echi di sirene di nebbia.

Evidently, the end rhymes are suppressed in the target text, in favour of a more literal, prosaic rendition. Nonetheless, it seems to me that the translator is making a conscious choice of “expanding” the Italian text in order to incorporate the meanings expressed more subtly in English. For instance, the translation incorporates the additional information dal mare, from the sea, which arguably serves to depict more vividly the harbour setting of the poem. By the same token, two verbs absent from the source text are inserted in the translation, penetrano and entrano, literally “to penetrate” and “to enter,” which give a more physical presence to the “echoing” of the foghorn, nominalised in Italian as echi di sirene di nebbia. Overall, the English text gains its poetic qualities and enthralling atmosphere from elusiveness and a certain conciseness of language, besides rhythm and rhyme. The Italian translator instead succeeds in creating a vivid evocation of the same setting by increased semantic specificity and a peculiar use of syntax, like the VS constructions of attraccavano navi passeggeri in the subordinate clause and, more significantly, “penetrano... entrano (nei nostri letti) gli echi di sirene di nebbia,” where the main subject is moved to the very end of a long passage. In other words, consistently with the goal of achieving a thorough comprehension of the source text, the translator prefers to ensure intelligibility of certain nuances of content, thus effecting what in Italian is often referred to as a translation brutta e fedele, ugly but faithful. Its “ugliness” represents the conscious sacrifice of an elegant reproduction, or indeed re-invention, of the formal poetic structure of the original poem. From this perspective it could be argued that the constructed reader of this particular collection, rather than a keen reader of Italian poetry—who would miss a more sophisticated use of the language—is either an Italian student of Irish literature or a student of Italian. This would, after all, be consistent with Bertoni’s profile: Senior Lecturer in the Department of Italian at Trinity College Dublin, and the publisher’s strong academic orientation: Trauben has active links with Turin University, where the wife of its director runs a Centre for Celtic Studies.

Debora Biancheri

111

A further example of how the translation gives priority to clarity of meaning is found in “Smoke”/“Fumo” (144-147): (c) Bone-idle I lie listening to the rain, not tragic now nor yet “to frenzy bold” – must I stand out in thunder yet again who have thrice come in from the cold? Sold on sobriety, I turn to the idea of nicotine, my opium, hashish, morphine and cocaine, […] Now closing time and the usual commotion, crowds and cars as if to a revolution; blue in the face behind my veils of smoke I try to recapture pool dreams or evoke aesthetic rapture, images of felicity, […]

Ozioso fino al midollo, steso in ascolto della pioggia non tragica adesso, né ancora a rovesci convulsi tra i tuoni devo resistere di nuovo io, venuto dal freddo tre volte? – Appresa la sobrietà rivolgo la mente all’idea della nicotina, mio oppio e hashish, cocaina e morfina. […] Ecco l’ora della chiusura, con la solita agitazione di folle e auto manco andassero a una rivoluzione; blu (di tristezza?) in viso dietro veli di fumo, cerco di ritrovare sogni di biliardo, di rievocare rapimenti estetici, immagini di felicità, […]

First of all the compound “bone-idle,” which in English can be considered a “dormant metaphor” (see Prandi, this vol.) is partly revived in its Italian translation: ozioso fino al midollo “idle till the [bone] marrow”, which by recovering the literality of the metaphor in a context where the overuse has not diminished its effect creates a less common, more striking image. Similarly, the ambiguous “blue in the face” is rendered with blu (di tristezza) in viso “blue (of sadness) in the face”. Once again the reader is brought to the source text, rather than the other way around. The colour “blue” in Italian is related to other emotions, such as fear, as in ho una fifa blu, hence the necessity to add di tristezza in order to spell out the implied meaning, while simultaneously keeping the impressionistic depiction of a face behind “veils of smoke” created by the source text. Serpillo’s review, however, does not rate as successful this attempt to keep some of the imagery of the source text while making the meaning more explicit to Italian readers who may be unaware of it: in his view, the parenthetical additions “lack a function.” More importantly, they also clearly impinge on the rhythm of the poems, as shown in “A Bangor Requiem” (148-151):

112

Translation Strategy and the Constructed Reader

(d) Shall we say the patience of an angel? No, not unless angels be thought anxious too and God knows you had reason to be; for yours was an anxious time of nylon and bakelite, market-driven hysteria on every radio, your frantic kitsch décor designed for you by thick industrials and twisted ministers (‘Nature’s a bad example to simple folk’); [...]

La chiameremo pazienza di un angelo? No, no a meno che anche gli angeli si possano pensare ansiosi e Dio sa se tu avevi ragione di esserlo; perché i tuoi erano tempi ansiosi di nylon e bachelite, di isteria provocata dai mercati in ogni radio, i tuoi frenetici interni kitsch progettati per te da ottusi industriali e perversi ministri (di Dio) (“la Natura e di cattivo esempio per i semplici”); [...]

Here “twisted ministers” becomes perversi ministri (di Dio), as the translator discounts the option of treating “ministers” as a false friend and employing the common Italian title sacerdoti, which would have tightened up the rhythm and created assonance with the preceding progettati. Yet his translation is acceptable, as sacerdoti can be properly referred to as ministri di un culto religioso, ministers of a religious cult. Although it is doubtless true that the poetic effects of the source text are repeatedly compromised by the Italian rendition, Serpillo’s attitude is illustrative of a widespread, yet far too categorical, theoretical belief that the implicit priority of every target text has to be an aesthetically pleasing outcome. However, a resolution to address specific constructed readers might dictate otherwise. According to this critical stance, it is possible to justify Bertoni’s strategy of privileging the content and providing target texts which are almost glosses—plus additional clarifications—in view of the fact that the Italian edition includes all the source texts as well. Considering that those texts are in English, it is legitimate to presume a level of knowledge on the readers’ part which should allow at least a reasonable appreciation of their sound patterns, if not metrics and rhyme schemes. The straightforward dismissal of this strategy as a “failed” translation suggests an unquestioned reliance on the preference for a creative appropriation of the source text, even if this would imply substantial re-working. Bertoni’s premise is different: he acknowledges the dependency of his own work on the aesthetics of the source text. He aims to produce a translation that is not an “appropriation” of, or substitute for, the source, as often happens when the translator’s role is taken up by another poet rather than a scholar. He does not use Mahon as the source of

Debora Biancheri

113

inspiration to create new poetry in Italian, but acts as a mediator between Mahon’s work and the Italian audience. Therefore, he does not try to achieve equally accomplished final products, but rather “uses” his translations in the best way possible to make accessible and highlight the significance of the source texts for the benefit of a readership eager for new knowledge. Since the translation becomes indissolubly linked to its source text, Bertoni’s method is the embodiment of the perfect pursuit of a symbiotic relationship between source and target culture as posited by the concept of Third Space. This concept, besides being an abstract goal in the theoretical exploration of translation as an ontological process, can therefore be relevant when assessing practical translation strategies.

1. Reaching for a Third Space: embracing differences between languages and cultures as the starting point of translation Overall, a criticism as harsh as that articulated in the review of L’Ultimo Re del Fuoco seems to miss the strength of the collection as a whole, which lies largely in the presence of the source texts and the support given by the critical apparatus, which also provides supplementary “expansion” and “explications” of linguistic aspects and cultural references in the collection. Indeed the presence of any paratext intended to supply relevant critical intervention is another crucial aspect of what constitutes the translation process in the context of this article. Introductions in particular provide the most indicative elements of translators’ efforts to mediate specific cultural items in the source language poems, which further consolidate the idea of constructed readers, outlined above. The volume L’Ultimo Re del Fuoco is equipped with two robust opening essays by the two translators, Roberto Bertoni and Giovanni Pillonca. These cover the themes that Mahon most frequently engaged with and focus on those aspects that are crucial to a cross-cultural understanding of his poetry, such as the relationship between the poet and his community (2000: 10). Bertoni’s introduction, rather than being an attempt to lay the foundations for fuller comprehension of the published poems, is conceived as a general assessment of Mahon’s work, in the sense that it transcends the scope of the poems contained in this specific Italian collection. For instance, the translator is keen to highlight the poet’s ironic attitude to the “new world order,” not only by engaging, from a rather broad perspective, with the themes of alienation, exploitation, technological fetishism, virtual reality and subsequent human detachment

114

Translation Strategy and the Constructed Reader

from authentic contact with nature and history, but also by relating it to other poems that are not present in the selection published by Trauben, which was, at the time of its release, the only collection of poetry by Mahon translated into Italian. For instance, he mentions The Hudson Letter, a collection which deals with American socio-economic policies and which Bertoni himself translated for another Italian publisher in 2003. Likewise he recalls “The Yellow Book,” a poem concerned with the “tourist age” and the “economic development” of Ireland, which gave him the opportunity to stress the nostalgia of the narration voice for the conditions preceding the Irish economic boom of the 1990s. The complete picture of Mahon’s achievement also includes a commentary on the more recent developments in his poetry, which was becoming like the world as Mahon describes it, “a forest of intertextuality” (27). Bertoni’s wide-ranging approach, besides giving a comprehensive account of the general concerns animating Mahon’s poetry, also gives him the opportunity to comment on the general features of Mahon’s style, characteristic of poems that are present in L’Ultimo Re del Fuoco as well. On the one hand, Mahon mocks postmodern pastiches of writing forms and modalities that lack content and purpose, on the other his poems manifest some features typically associated with post-modernism (26). More importantly, Bertoni’s insistence on Mahon’s polemical attitude towards the dehumanising aspects of postmodernity can be seen as the careful construction of a proper reception for “The Last of the Fire Kings,” the eponymous poem of the Italian collection. Here Mahon openly exposes his willingness to “break with tradition,”2 something that was perceived as “heretical” when the poem first appeared in Ireland. Most of the “committed” poetry of the late 1960s and early 1970s, especially when rooted in the North of Ireland, consciously focused on local rather than cosmopolitan aspects of Irish culture. Nowadays, however, the “pioneering” postmodern attitude which led to the poem’s initial rejection has been accepted as the foundation of a new critical condition increasingly concerned with the global, existential dimension of violence and alienation (Bertoni 2000: 25). The strict interconnectivity between local and universal, Irish and European dimensions, past and present, defines several of Mahon’s poems. This is exemplified by “Death and the Sun”/“La Morte e il Sole,” a poem which, through Camus’ interpretation of the myth of Sisyphus, allows a comparison to be developed with the unemployed of modern Belfast. By 2

“Last of the fire kings, I shall/ Break with tradition and/ Die by my own hand/ Rather than perpetuate/ The barbarous cycle” (50).

Debora Biancheri

115

stressing this and other European topoi which are part of the collective cultural repository of the Italian educated classes, the introduction effectively uses the poetry’s complexity as a way to make it more accessible to Italian readers. A comprehensive list of authors who have influenced Mahon, at times barely discernible when reading his poems, is also supplied: Ovid, Rimbaud, De Quincey, Seferis, Malcom Lowry, Ezra Pound, Albert Camus, Edvard Munch, Van Gogh and Paolo Uccello. The eclectic variety of these names recalls Latin and French cultures, as well as the Modernist and Impressionist movements: all traditions prominent amid the Western cultural paradigm to which Italy belongs. It therefore constitutes another attempt to frame the reader's interpretation by way of highlighting references in the poetry collection more attuned to his/her cultural profile. Acknowledging the abundance of information supplied in the introduction is crucial to assessing Bertoni’s strategy fairly. With this in mind, this final example may serve to reveal the logic behind the translations. From “Beauty and the Beast”/“La Bella e la Bestia” (142f.): (e) I sit here like an old child with a new toy Or a creature from outer space, Saturn perhaps, Admiring the ingenuity of the planet of the apes when (look!) the huge gorilla, the size of a fly [...] A black speck outlined against the morning sky Clutching Fay, said Noel Coward, ‘like a suppository’. […] ...The little bi-planes come gunning for him now And Kong, by Jove, knock one of them out of the sky With a hairy hand. They wear him out, of course, And he falls to extinction among the crowds below. […] Fay, born in Alberta, you were also in Dirigible And ‘existed more forcefully when faced with terror’

Siedo qui, come un ragazzo invecchiato, con un nuovo giocattolo, O come una creatura proveniente dallo spazio, che so? Da Saturno, Ammirato dall’ingegnosità del pianeta delle scimmie Quando eccolo! Il gorillone, ridotto alle dimensioni di una mosca [...] Piccolo punto nero delimitato sullo sfondo del cielo mattutino, Tiene tra le grinfie Fay, “come,” cito Noel Coward, “una supposta”. […] …Ora arrivano i piccoli biplani a bersagliarlo: Per Giove!, Kong ne spazza uno dal cielo, Affibbiandogli una zampata pelosa. Spezzano la sua resistenza, Certo, lui cade, si estingue in mezzo alla folla là sotto. [...] Fay, nata nell’Alberta, eri anche in Dirigibile Ed «esistevi con massima forza faccia a faccia con il terrore»

116

Translation Strategy and the Constructed Reader

Says Video Guide – like most of us, probably. Well, Kong and I dedicate this one to you, old girl, Wherever you are, pushing 90 and hanging in there, We want you to know we love you and root for you still.

Dice la «Video Guide» – come la maggior parte di noi, probabilmente –. Be’, vecchia mia, io e Kong ti dedichiamo questo pezzo, Ovunque tu sia; vicina come sei ai 90 anni, sempre arzilla, Vogliamo tu sappia che ti amiamo e teniamo ancora per te.

At first inspection, the colloquial terminology displayed in the translation may sound almost inappropriate in a poetic context. Yet, it ultimately achieves the effect of shocking the reader by virtue of its quirkiness: something that is fairly close to what the hectic language of Mahon’s postmodern poem achieves by very different means. The strange image of “the huge gorilla,” for instance, becomes il gorillone, a translation that, by employing the augmentative suffix –one, acquires an almost childish dimension. The expression could have been more literally and naturally rendered as l’enorme gorilla, yet Bertoni’s unusual choice is actually coherent with a poem that carries the title of a fairytale. Similarly, sempre arzilla, always sprightly, translates the less flamboyant expression “hanging in there”—referring to actress Vina Fay Wray in her nineties— which again brings an almost cartoonish touch to the poem. These notes are even more remarkable because of the contrast between formal and informal language. The Italian text swings from expressions like spazza, affibbiare and zampata, which are all extremely colloquial, to the epic-sounding sentence spezzano la sua resistenza, which literally means “they broke his resistance.” However, overall, the apparently bizarre quality of the target text is an attempt to create a sense of polylingualism which is a hallmark of many of Mahon’s poems, as the reader should know from the introduction. To judge it according to conventional aesthetic standards means failing to see that the translation is consciously carrying across traits of hybridity: something that is shocking before being beautiful ࡳ but, once accepted, it can be more enriching than translations indiscriminately aiming at stylistic elegance. Indeed, the Third Space can be reached only by breaking free from the boundaries of established canons of translation, so as to allow two different linguistic and cultural traditions to merge. In this sense, the appreciation of Mahon’s idiosyncratic features, typical of literary post-modernism, that is fostered by Bertoni’s introduction and translation, should trigger a more truthful and fruitful dialogue between his poetry and the Italian literary canon.

Debora Biancheri

117

2. Challenging the readers: poetry translation as the preservation of foreignness “released from the constraints of turning a profit” The attention to the socio-cultural context of the work and its author displayed in a series like “Poesia Irlandese” is in sharp contrast to normative behaviour relating to the translation of Irish contemporary fiction, where there is no perceived need for paratextual elements to complement the meaning of the translation proper. This means that meaningful elements belonging to the cultural area in which the source text is rooted are regularly left floating like inexplicable “black holes” in the familiar universe of a highly domesticated language. One brief example to illustrate the issue can be found in the regular use of acronyms referring to paramilitary groups in Northern Ireland—such as UDA or UVF (Ulster Defence Associations and Ulster Volunteer Force)— in contemporary novels by authors from the six counties, such as Eoin McNamee, Bernard McLaverty and Liam McLiam Wilson. The Italian translations, while leaving those items unaltered, frequently do not display any kind of glossary or notes to explain the references of those abbreviations, let alone their historical significance in the geographical and cultural context of the novels. In this sense, the Italian policy in relation to fiction apparently contravenes the rule of immediate “recognisability” of the target text and the promotion of “hegemony” which Venuti postulates as main features of the fluency typical of transparency. Rather, by leaving indicators of an alternative cultural specificity such as UVF and UDA untranslated, spaces of unintelligibility may be created. Effectively, elements of the source texts which do not have a direct equivalent in the cultural arena of the receiving culture may remain arguably “unheard” by the target readers if deprived of their context, making the transparency of the target texts ultimately “silent.” The “resistancy” hypothesised by Venuti’s preferred mode of translation, should instead, while re-configuring the message to be accepted within the target culture, foster an appreciation of its significance reaching back towards the context in which such a message was originally framed. In the specific examples provided, for instance, an understanding of the city of Belfast, or Northern Ireland more generally, as a geographical and political location, is essential for the appreciation of implicit meanings that contribute to the overall significance of the novels. In order to cultivate a “heterogeneous discourse” (Venuti 1998: 11) target texts should vary in accordance with the identity of the “narrator” being translated. Such an achievement is rendered impossible both by the

118

Translation Strategy and the Constructed Reader

straightforward domestication posited by “transparency” as such, and by the illusion of transparency pursued by many translations of contemporary fiction in Italy. Both effectively erase the “sense of foreignness” that in most cases has invited translation in the first place. It is important to remember, however, that such a translation strategy is rarely a conscious choice on the translators’ part, as it depends on external pressures that are mostly moulded by pre-determined ideas about the reader’s profile for different types of literary texts. With regard to contemporary fiction, the need to meet the alleged demands of readers that are likely to exist in large numbers is dictated by a publishing market which is firmly located within an entrepreneurial mindset. As its own survival is dependent on sales, the act of translation is heavily influenced by economic rather than artistic concerns. From this perspective it is relatively unimportant that the conservation of strangeness is only apparent because the lack of supplementary mediation often deprives the source texts of their significance. Such use of transparency is rather indicative of a conscious construction of the appeal of contemporary fiction mainly in terms of “diversion and entertainment.” In this light, the “elitist” production and limited circulation which characterises Trauben volumes of Irish poetry can be understood in terms of different priorities of the publishing house, as well as of an alternative delineation of the audience profile. If Bertoni’s translations analysed in this article were displayed in the context of a “pocket” poetry edition, with no introduction and no source texts, with the target texts left to stand and to “speak” for themselves, then they might indeed be weak by virtue of the lack of poetic appeal of the target texts alone. Yet, if restored to the auxiliary position that the translations occupy within the Trauben publication, the translator’s endeavour can be assessed not so much as a work intended to appeal to the reader’s aesthetic sensibilities, but rather as one intended to enrich the reader’s knowledge of the linguistic and cultural universe of the poetry’s origins. From this perspective, the introduction ultimately remains the core of the interpretative framework sustaining this volume of Mahon’s translations. Summing up, in this instance, poetry translation—albeit finding its expression by very different means—can be understood in terms compatible with Pierre Bourdieu’s definition of poetry itself, as the “disinterested activity par excellence,” surviving not on economic, but rather “charismatic legitimation” (1993: 51). This means that the domain of poetry, more so than fiction, can foster alternative ways of dealing with the “non-identitarian nature of things and ideas” (Fuchs 2009: 5) that is the premise of translation. The concept of hybridity posited by Third Space

Debora Biancheri

119

allows a discussion of mediation as a negotiation between continuity and difference which provides a useful theoretical underpinning to postulate a dimension that ultimately transcends both of the contact cultures. An earlier formulation of “fusion of horizons” of source and target contexts is to be found in Hans-Georg Gadamer, who theorised this Verschmelzung with regard to contemporary interpretations of historical documents (1989). The notion of “horizon” employed here derives from phenomenology: it stands for the larger context of meaning in which any particular meaningful presentation is situated. A text which has become alien because of temporal distance has to be recovered through an appraisal of the difference between past and present horizons. In other words, a dialogue with the past is possible only if both “the alien horizon of the text and the interpreter’s own horizon” (Machor and Goldstein 2001: XIII) are taken into account. For Gadamer, the gap between these two horizons provides an opportunity to discover the link between the contemporary and the earlier context, a sort of new dimension in which both could coexist. Susan Bassnett and André Lefevere also adopt a diachronic point of view when they stress the necessity for a translation to be regularly updated in order to meet new horizons of expectation (Bassnett and Lefevere 1995). From their perspective, the “fusion,” rather than embodying a form of interrelations, would reflect a representational stance. This seems to posit the target culture as the one univocally dictating the norms that will establish the final outcome of the translation process. In this sense, “acceptability” (Toury 1995) seems to be privileged, as it is almost taken for granted that meeting the standards of the receiving culture is what will determine the “appropriateness” of a translation. Although in theoretical terms this recalls quite closely the “reassertion of self-identity” of the receiving system, something that Zygmunt Bauman (1999) attributed to the process of “assimilation,” the interactional nature of the translation process should not be played down. The rearrangement of the source culture for the benefit of the target culture may determine a mutual opening up: on one side the source culture needs to adapt to the target culture in order to be accepted, on the other side the target culture has to find the decoding structures which will render the foreignness acceptable. As Bauman maintained “…both partners emerge changed from every successive attempt at translation […] No act of translation leaves either of the partners intact” (1999: XLVIII). The strategy of mediation, as outlined in relation to the translations of Mahon’s poetry, which is indeed representative of translation norms widely employed in relation to this literary genre in Italy, opens up a space where heterogeneous perspectives on translation can be protected from

120

Translation Strategy and the Constructed Reader

being assimilated by the domestic culture. It could be seen as a form of “minoritising” translation, as it favours particularity and respects differences by promoting the awareness of asymmetries in the transmission process—if not always in terms of power relations, as discussed by Venuti. The aim is still to resist the “assimilationist ethic by signalling the linguistic and cultural differences in the text—within the major language” (Venuti 1998: 12) or, in this case, simply the target language. In conclusion, this assessment of a poetry collection in translation should have served the purpose of demonstrating that the policy of mediation, through its acknowledgement that meanings are inevitably changed by the transfer process, promotes non-equivalence as the very essence of translation and accordingly integrates the source texts in ways which reveal their translational dimension. As a final note, however, it should be stressed that the practical utility of such effort of opening “the self towards the other, thus extending and developing target and source languages” (Fuchs 2009: 5) is somewhat undermined by its very nature. The reason a collection such as L’Ultimo Re del Fuoco can be praised in an academic context is the very reason that might alienate a wider audience and cause series like “Poesia irlandese” to be perceived by many as scholarly exercises for students or other scholars to read. Ultimately, those translations striving hardest to reach the interstitial hybridised dimension of the Third Space are those most likely to have poor potential for sale. With this in mind, this brief analysis of the Italian collection of Mahon’s work on the one hand offers insights into the stimulating opportunity to think of “Irish poetry in Italian translation” as a category animated by interaction between the Italian and the Irish literary canons which actually transcends the borders of both. On the other hand, it highlights the double-edged nature of mediation when translation is only one of the many factors constituting a market-driven publishing environment. The “ugliness” that is the cost of fidelity can hinder the visibility of source texts belonging to a literary genre still largely valued in terms of the aesthetic qualities of the language used.

Acknowledgments Grateful acknowledgement is made to Trauben Edizioni s.a.s. for the permission to reproduce, and to the Irish Research Council for funding my research activity. I also wish to mention the generous support of Dr. Louis de Paor, whose guidance was crucial for my engagement with Irish poetry.

Debora Biancheri

121

References Bassnett, Susan, and Andre Lefevere. 1995. Translation, History, and Culture. London: Cassell. Bauman, Zygmunt. 1999. Culture as Praxis. Theory, Culture & Society. New ed. London: SAGE. Bhabha, Homi K. 2000. “How Newness Enters the World: Postmodern Space, Postcolonial Times, and the Trials of Cultural Translation.” In Writing Black Britain, an Interdisciplinary Anthology, ed. James Procte, 300-6. Manchester: Manchester University Press. Berman, Antoine. 1992. The Experience of the Foreign: Culture and Translation in Romantic Germany. Albany: State University of New York Press. Bertoni, Roberto. 2000. “Figurazione del millennio: le raccolte degli anni novanta.” In L'ultimo Re del Fuoco, ed. Roberto Bertoni and Giovanni Pillonca, 21-8, Torino: Trauben. Biancheri, Debora. 2013. “Translating Irish Literature into Italian: The Challenges of Decoding the Unfamiliar.” In Authorial and Editorial Voices in Translation 2 - Editorial and Publishing Practices, ed. Hanne Jansen and Anna Wegener. Montréal: Éditions québécoises de l’oeuvre, collection Vita Traductiva. http://hdl.handle.net/10315/26584. Bourdieu, Pierre, and Randal Johnson. 1993. The Field of Cultural Production: Essays on Art and Literature. Cambridge: Policy Press. Fuchs, Martin. 2009. “Reaching out; or, Nobody exists in one context only: Society as translation.” Translation Studies 2 (1): 21-40. Gadamer, H. George. 1989 [1975]. Truth and Method. London: Sheed and Ward. Genette, Gérard. 1997 [1987]. Paratexts. Thresholds of interpretation. Cambridge: CUP. Hermans, Theo, ed. 1999. Translation in Systems. Descriptive and SystemOriented Approaches Explained. Michigan: St. Jerome. Machor, James L., and Philip Goldstein. 2001. Reception Study: from Literary Theory to Cultural Studies. New York: Routledge. Mahon, Derek. 2000. L’Ultimo Re del Fuoco. Trans. by Roberto Bertoni and Giovanni Pillonca. Torino: Trauben. Nabokov, Vladimir. 2000. “Problems of Translation: Onegin in English.” In Translation Studies Reader, ed. Lawrence Venuti, 71-83. London: Routledge. Serpillo, Giuseppe. 2004-5. “Review of ‘Poesia Irlandese’.” Il Tolomeo 8: 142.

122

Translation Strategy and the Constructed Reader

Toury, Gideon. 1995. “The Nature and Role of Norms in Translation.” In Descriptive Translation Studies and Beyond, ed. Gideon Toury, 53-69. Amsterdam: Benjamins. Venuti, Lawrence. 1995. The Translator's Invisibility: a History of Translation. London: Routledge. —. 1998. The Scandals of Translation: Towards an Ethics of Difference. London; New York: Routledge. —. 2000. The Translation Studies Reader. New York: Routledge. —. 2011. “Introduction.” Translation Studies 4 (2): 127-32.

MOOD AND MODAL VERBS IN THE ENGLISH, GERMAN, AND GREEK OFFICIAL VERSIONS OF THE CHARTER OF FUNDAMENTAL RIGHTS OF THE EUROPEAN UNION URSULA STEPHANY UNIVERSITY OF COLOGNE

In studies of translational science, it is usually admitted that different linguistic versions do not represent exactly the same text, neither in form nor content. Important insights into “the ways in which speakers of different languages depict the same events in words” (Slobin 1996: 76) may not only be reached by translational studies, but also by comparing different authentic linguistic versions of legal texts produced in multilingual contexts such as the European Union. While most linguistic studies of legal language have focused on nouns and noun phrases so far, modality is an especially rewarding field for the cross-linguistic study of legal texts (Burr and Gallas 2004; Lenz 2006; Panaretou 2009; Tacke 2011). In the present paper, deontically modal notions expressed by the categories of mood and modal verbs will be studied in three authentic versions of the Charter of Fundamental Rights of the EU. It will be shown that despite a similarly rich inventory of modal verbs in English and German the use of modal verbs and mood in legal texts is strikingly different and that mood is used in a more similar way in the German and Greek versions of the Charter than in its English and German versions.

1. Introduction In studies of translational science it is a commonplace to admit that “the version from a text A in a verbal language Alpha into a text B in a verbal language Beta” (Eco 2003: 1) does not result in exactly the same text, neither in content nor, of course, in form. Therefore, terms such as “transfer”, “equivalent text”, “dire quasi la stessa cosa [saying about the same thing]” abound in definitions of translation (see Nida 1964; Eco 2003; Ramat 2007). As rightly remarked by Ramat (2007), it would

124

Official Versions of the Charter of Fundamental Rights of the EU

however be difficult to accept “that some concepts and ideas could be formulated and expressed only in language A but not in language B.” Slobin (1996: 76) explains the fact that “translations of the same text cannot help but add or remove nuances in accord with the characteristics of the given language” by what he calls “thinking for speaking”, namely “a special form of thought” for expressing “experience in linguistic terms”. One way of investigating the particular ways of thinking for speaking “is to compare the ways in which speakers of different languages depict the same events in words” (Slobin 1996: 76). This approach is not only particular to translational science, but important insights may also be reached by studying the authentic linguistic versions of legal texts produced in the multilingual context of the European Union in which the methods of translation and co-editing play an essential part. According to Ramat (2007), an important goal of translation from a pragmatic point of view is “isofunctionality”. A translation is functionally equivalent to the original text if it obtains “the same pragmatic effect”. Pragmatic isofunctionality is fundamental in the production of authentic legal texts (see below) in the multilingual European Union with its different judicial systems.

2. Multilingual Legal Texts in the European Union The European Union presently consists of 28 member States and comprises 24 official languages. Its linguistic and cultural diversity is explicitly acknowledged in art. 22 of the Charter of Fundamental Rights of the European Union, which reads: “The Union shall respect cultural, religious and linguistic diversity.” Official multilingualism was already established in art. 314 of the European Community in Rome in 1957 (Burr and Gallas 2004: 241) and continues to be considered as a constituting characteristic of European identity, which is taken into account in the legal texts of the European Community (Burr 2006: 187; see also Burr and Gallas 2004: 195). With 24 official languages having been recognized since July 1st 2013, multilingualism of the European Union has in the meantime expanded exponentially. A further important feature of the European Union besides official multilingualism is its diversity of judicial systems. While Visconti (2013) considers these as constituting “major obstacles to European legal integration”, Burr (2009: 758) is more optimistic and delineates the gradual development of a common legal system in the European Union in which “new concepts particular to European Law are being developed and linguistically expressed.” According to Burr and Gallas (2004: 240), this

Ursula Stephany

125

autonomous European legal language may differ significantly from the diverse national legal languages present in the European Union so that given legal concepts used in common European Law and in diverse national kinds of law do not necessarily bear the same meaning (Burr 2009: 758). The multilingual law of the European Union may even lead to the abolishment of terminological distinctions in certain languages because the underlying facts have changed. Thus, in the Lisbon Treaty, the legal distinctions between German Entscheidung ‘decision’ and Beschluss ‘resolution’ and Dutch beschikking and besluit have been disposed of in favor of German Beschluss and Dutch besluit (Burr 2009: 753). It is a special characteristic of European Law (Art. 314 EU; see also Burr 2006) that legal texts are produced in all official languages of the European Union. Each of these texts constitutes an official or authentic linguistic version and is legally binding to the same degree as the other authentic linguistic versions of a given text.1 This means that all authentic linguistic versions of legal texts of the European Union are legally equal. Burr (2006: 196-198, 2009: 750, 757) stresses that, in such a situation, European Comparative Legal Linguistics may contribute to optimizing both the production of legal texts in legislation and their interpretation in administration of justice by discovering apparent convergence and divergence of the different linguistic versions. Burr (2006: 187) not only considers the Charter of Fundamental Rights of the European Union having been edited in 24 authentic linguistic versions in the meantime as a genuine “European text” constituting a milestone “on the Union’s way to a community of fundamental rights”, but also as an important pillar for the further development of European Legal Linguistics. The production of the numerous authentic linguistic versions of legal texts of the European Union represents a major challenge for lawyers, translators and linguists. As pointed out by Burr (2006), legal texts of the European Union owe much to the production of such texts in officially multilingual countries such as Switzerland or Canada. In such countries, a principle of “integrated multilingualism” is applied. This is characterized by the development of a common legal language “in multilingual shape”. Such a legal language is developed in the process of co-editing legal texts 1

Linguistic versions of legal texts which are legally equivalent are qualified as “being legally authentic” (Burr 2013: 1468). “The change of status from translations to authentic language versions is explicitly laid down in Art. 28.2” of the Court of Auditors’ Rules of Procedure (Burr 2013: 1479). Legislative texts in the 23 EU languages are of “equal authenticity and authoriticity” which is based on “the parity of the legal cultures and the official languages of the Contracting States” (Burr 2013: 1487-1488).

126

Official Versions of the Charter of Fundamental Rights of the EU

in the different languages of the country in question. In co-editing, multilingual legal texts do not start from a source language and move on to a certain target language in a linear way. Rather, integrated multilingualism is a dynamic principle in which multilingual legal texts form “a multiple textual space” constituting the legal plausibility space (Burr 2006). As stated by Burr and Gallas (2004: 199) the method of convention or co-editing becomes impracticable once more than three languages are involved. In contrast to the number of four official languages in multilingual countries such as Switzerland or the European Economic Community EEC of over 50 years ago, the European Union presently has to face the problem of handling 24 national languages, all of which enjoy the status of official languages of the European Community (see Burr 2009: 758). Although Burr and Gallas (2004: 199) consider the method of convention or co-editing to be “certainly the more efficient and also more elegant one because it takes the génie de la langue, the peculiar characteristics of each language into consideration”, the method of translation plays an important part in the production of legal texts in the European Union besides co-editing (Burr and Gallas 2004: 198-199). In mere translation, the source text is considered as the authentic text, while in co-editing the processes of translation help to produce and structure the different authentic texts by a simultaneous and common endeavor (Burr and Gallas 2004: 199, 201). In practical law-making within the European Union, the legal text is negotiated by the European Council, the European Parliament, and the European Commission in one of the most common official languages, usually English or French and, in later discussions, this text is again and again referred to in spite of the fact that it has already been translated into the different official languages of the European Union so that everybody participating in the negotiations may intervene in his or her own native language (Burr and Gallas 2004: 199). According to Burr and Gallas (2004: 227-228) the constitution of a common legal text by transferring it into different linguistic versions while following the principles of linguistic adequacy, usefulness, and comprehensibility, may even offer the chance of improving the conclusiveness of the original text. Possible errors or ambiguities of the original text may thus be eliminated (Burr and Gallas 2004: 230).2 In view of the large number of typologically and genetically diverse official languages represented in the European Union, Burr and Gallas (2004: 241) suggest that legal texts should be produced in genetically diverse groups of languages rather than in genetically related 2 It may even at times be preferable to change the original text rather than its translations (Burr and Gallas 2004: 230, 240).

Ursula Stephany

127

ones. In this way, the syntactic or discourse-specific dominance of single languages may be avoided. At the time when the text of the Charter of Fundamental Rights of the EU was produced by the “method of convention”, this was a novelty in the production of a legal text in the Community (Burr 2006: 189; see also Burr and Gallas 2004: 196-197). While in the beginning phases of committee work translation of the contributions of different members speaking different native languages was important, in the final deliberations, authentication of the different linguistic versions of the Charter was achieved by a cooperation of specialists in law and languages (Burr 2006: 189).

3. Legal language Legal language may be called a sociolect, register, or professional jargon (Lenz 2006: 31, Moser and Panaretou 2011: 33). Legal texts thus belong to a specific genre serving the communicative purposes of the discourse community of lawyers engaged in legislation and its application. The core of legal language is the language used in laws (Lenz 2006: 31). The genre of legal language is characterized by a number of special linguistic features partly depending on a specific language. In all languages, it possesses a special terminology. Some typical features of German legal language mentioned by Lenz (2006: 33) are its nominal style rich in functional verb constructions such as ein Urteil fällen ‘to pronounce judgment’ rather than urteilen ‘to judge’ or Anwendung finden (lit. ‘to find application’) rather than angewandt werden ‘to be applied’ as well as the presence of complex nominal compounds such as Grunderwerbssteuerbefreiung ‘tax exemptions on the purchase of land’ or Bundesausbildungsförderungsgesetz ‘federal law for the promotion of training and education’. In the syntactic domain, German legal language is characterized by an impersonal style leading to the frequent use of passive constructions in which desired or undesired actions including their goals are more prominent than personal agents (Lenz 2006: 34). There are many similarities between German and Modern Greek legal language. These may in part be explained historically since Modern Greek Law mainly goes back to the German lawyer Georg Ludwig von Maurer, one of the four regents of King Otto I of Greece from 1832 to 1834. Besides a special terminology and the more frequent use of nouns as compared to the standard language, Modern Greek legal language is

128

Official Versions of the Charter of Fundamental Rights of the EU

characterized by use of the passive voice3 and learned grammatical forms and structures (Panaretou 2009: 76-77). Further characteristics of Greek legal language are its long and complex sentences and nominal rather than pronominal anaphor (Moser and Panaretou 2011: 32; see also Panaretou 2009). When Moser and Panaretou (2011) compared directives used by Greek mothers setting rules for their children and legal rules expressed by laws, they found important differences in the use of the categories of tense, aspect, and mood. These could reach such a degree that certain constructions which are acceptable in legal language are judged to be unacceptable in standard Greek. Accordingly, the authors’ hypothesis is that “what makes deviant uses acceptable in law texts is the specific genre, as conventionalized by the members of the relevant discourse community within its socio-cultural and institutional context” (Moser and Panaretou 2011: 12). Since most linguistic studies of legal languages have focused on nouns and noun phrases so far (Visconti 2013), studies of other domains of legal language are most desirable. Thus, Visconti (2013) is concerned with connectives such as notwithstanding or subject to, which play an important part in the “uniform interpretation and implementation of EU legislation”. She points out that connectives are inconsistently used in the English, Italian, French, and German authentic linguistic versions of different legal texts and that in many cases debated by the European Court of Justice the judgment hinges on connectives and may cause legal dispute.4 As shown by Panaretou (2009) and Moser and Panaretou (2011) for Modern Greek, besides nominal terminology and connectives, the grammatical categories of tense, aspect, and especially deontic modality are of utmost importance in interpreting legal texts. The importance of deontic modality in legal language has also been emphasized by Burr and Gallas (2004: 235-236) for several European legal languages and for German legal language by Lenz (2006) and Tacke (2011). 3

Panaretou (2009: 99-100) points out an interesting difference between the use of the active and passive voice in Greek legal texts. While the passive voice is typically found in formal laws issued by the legislative organs of the State (e.g. ĮʌĮȖȠȡİȪİIJĮȚ /apaȖorévete/ ‘it is forbidden’, İʌȚIJȡȑʌİIJĮȚ /epitrépete/ ‘it is allowed’), the active voice is chosen for informal decisions and suggestions (e.g. ĮʌȠijĮıȓȗȠȣȝİ /apofasízume/ ‘we decide’, ıȣıIJȒȞȠȣȝİ /sistínume/ ‘we recommend’). The author explains this by the fact that in the case of formal laws the obvious power of the State makes it unnecessary to mention him as an agent, whereas informal decisions and suggestions must be attributed to an agent. 4 This is even true of the connectives and, or, and but, which play an especially important role in legal language (Visconti 2013; see also Gibbons 2003: 59-60).

Ursula Stephany

129

The present paper is concerned with comparing the expression of deontic modality in the English, German, and Greek authentic linguistic versions of the Charter of Fundamental Rights of the European Union. The results are preliminary.

4. Modality in legal language and in the English, German, and Greek versions of the Charter of Fundamental Rights of the European Union 4.1 Modality, mood, and tense in legal language5 According to Sayatz (1996, quoted by Burr and Gallas 2004: 235), legal texts “lay down standards for action or conduct concerning potential future actions in a certain domain.” Legal language is thus inherently future-oriented, a feature also characteristic of deontic modality (see Moser and Panaretou 2011: 22-23). Legal texts are not concerned with the subjective qualification of the status of events by the speaker “with respect to their validity, truth, or factuality” (Stephany 1986: 375), but with the necessity and possibility of facts and events, and therefore with doing rather than being. In legal texts, it is not the legislator’s degree of certainty about circumstances which is at stake but on the contrary the “it is-so” character of the law (Lyons 1977: 750; see also Panaretou 2009: 86). This also explains why legal rules are expressed in the indicative present (see below). Epistemic possibility plays a role in legal language only when the conditions for applying legal rules are laid down, since such rules have to take all possible and even rare or exceptional circumstances into consideration (e.g. an accident with dangerous materials may be an air crash) (Panaretou 2009: 87; Moser and Panaretou 2011: 22). Interestingly, there is not a single example of epistemically used may, kann/können ‘can, may’ or ȝʌȠȡİȓ /borí/ ‘may’ to be found in the text of the Charter of Fundamental Rights of the EU. It is deontic rather than epistemic modality which prevails in legal language (Panaretou 2009: 92; Moser and Panaretou 2011: 22).

5

In languages with a grammaticized category of verbal aspect such as Greek it would be interesting to study the role of aspect in legal language in addition to mood and tense and also compare perfective and imperfective verb forms to their expressions in other languages with a grammatical category of aspect (e.g. Polish) and to languages which do not possess the grammatical category of verbal aspect (e.g. German) or in which it is grammaticized in a different way (e.g. English).

130

Official Versions of the Charter of Fundamental Rights of the EU

Wüest (1993, quoted by Burr and Gallas 2004: 235) found that historically the diversity of deontic expressions in normative legal texts was much larger than it is nowadays. There are indeed only two major linguistic categories occurring in the three authentic linguistic versions of the Charter of Fundamental Rights of the European Union, namely modal verb constructions and the present indicative. A striking difference between Greek legal and standard language is that the very grammatical category exclusively admitting a deontic interpretation, the imperative, is not used in legal language. Moser and Panaretou (2011: 23; see also Panaretou 2009: 88) attribute this to the impersonal and general character of legal rules expressed by the third person of the verb, which precludes the imperative, since it is limited to directives in the second person. According to Moser and Panaretou (2011: 23), this “reflects the lack of interaction between the addresser and the addressee, i.e. the state and the citizens.” In Greek legal language, mediopassive present indicative forms of full verbs such as ĮʌĮȖȠȡİȪİIJĮȚ /apaȖorévete/ ‘it is forbidden’ or ȣʌȠȤȡİȠȪIJĮȚ /ipoxreúte/ ‘it is obligatory/necessary’ as well as the modal verbs ʌȡȑʌİȚ /prépi/ ‘it must’ and ȝʌȠȡİȓ /borí/ ‘it can, may’ are used instead of the imperative (Panaretou 2009: 88-89). A further difference between legal and standard Greek concerning mood is the following: While the subjunctive, which occurs in all three persons, is very frequently used for expressing polite directives in standard Greek, it does not occur in Greek legal language in this function (Moser and Panaretou 2011: 23-24).6 The reason is to be found in the general validity of legal rules which do not leave the option to the addressee, i.e. the citizen, not to commit to them. The most frequently used tense in legal texts is the present (indicative), to which we will turn in the next section. There are a few tokens of the future tense to be found in the English and Greek versions of the Charter, but not a single one in the German version.7 German expressions corresponding to the Greek or English future are either nominal expressions or present tense verb forms (examples 1).8

6

On conditional sentences in Greek legal language see Panaretou (2009: 97-99). On the opposition of the indicative and the subjunctive for the distinction of factuality and non-factuality in Greek legal language see Moser and Panaretou (2011: 20). 7 Wilder (2012a: 89) recommends to limit use of the English will future to future events in contracts and to avoid its use for representing duties in such texts. 8 For the future tense in Greek legal language see Panaretou (2009: 96-97) and Moser and Panaretou (2011).

Ursula Stephany (1)

131

Concluding statement of the Charter of Fundamental Rights of the European Union (a) The above text adapts the wording of the Charter proclaimed on 7 December 2000, and will replace it as from the date of entry into force of the Treaty of Lisbon. (b) Der vorstehende Wortlaut übernimmt mit Anpassungen die am 7. Dezember 2000 proklamierte Charta und ersetzt sie ab dem Zeitpunkt des Inkrafttretens des Vertrags von Lissabon. (c) ȉȠ ĮȞȦIJȑȡȦ țİȓȝİȞȠ ĮȞĮʌĮȡȐȖİȚ, ʌȡȠıĮȡȝȩȗȠȞIJȐȢ IJȠȞ, IJȠȞ ȋȐȡIJȘ țĮIJȐ IJȘ įȚĮțȒȡȣȟȒ IJȠȣ ıIJȚȢ 7 ǻİțİȝȕȡȓȠȣ 2000 țĮȚ șĮ IJȠȞ ĮȞIJȚțĮIJĮıIJȒıİȚ Įʌȩ IJȘȞ ȘȝȑȡĮ ȑȞĮȡȟȘȢ ȚıȤȪȠȢ IJȘȢ ȈȣȞșȒțȘȢ IJȘȢ ȁȚııĮȕȫȞĮȢ.

Law is future-oriented since it refers to circumstances which may occur from the moment it comes into force until it will be annulled (Panaretou 2009: 96; Moser and Panaretou 2011: 26). It may therefore seem surprising that the future tense only plays a minor role in Greek legal texts (Panaretou 2009: 96) or may even be completely absent from certain linguistic versions of the Charter. As far as the future orientation of legal rules is concerned, it must not be forgotten that deontically used modal verbs, which frequently occur in legal texts and also in the Charter, are intrinsically future-oriented. In English, the future tense is furthermore periphrastically expressed with the help of the modal auxiliaries will or shall and is thus strongly modal. However, as is commonly asserted, “in legal language shall does not indicate futurity, but is instead employed to express a command or obligation, and can thus be paraphrased with must” (Tiersma 1999). As will be shown below, modal constructions with shall amount to more than half of all finite verbal expressions in the English version of the Charter. In spite of the fact that also the German werden future has been taken to carry a strong modal character (Vater 1975), it is not used in the German version of the Charter. The Modern Greek future has a strong modal component, too, not only for historical reasons.9 In Modern Greek standard usage, it is frequently employed in an imperative function “by virtue of its factuality, which lends the command a greater degree of obligatoriness, presenting as it does the action as a future fact” (Moser and Panaretou 2011: 23-24). In Greek legal texts, the perfective future has a strong deontically modal character (e.g. ıİ ʌİȡȓʌIJȦıȘ İȞIJȠʌȚıȝȠȪ ĮȡȤĮȚȠIJȒIJȦȞ șĮ įȚĮțȠʌȠȪȞ ȠȚ İȡȖĮıȓİȢ /se períptosi edopismú arxeotíton șa įiakopún i erȖasíes/ ‘in the case of localization of

9

The Modern Greek future particle șa originates from an Ancient Greek verb for ‘will’.

132

Official Versions of the Charter of Fundamental Rights of the EU

archeological sites construction works shall be interrupted’10) (Panaretou 2009: 96).11

4.2 Modality in the English, German, and Greek versions of the Charter of Fundamental Rights of the European Union In the three linguistic versions of the Charter studied in the present paper, three different modal verbs are used in the English version, five modals plus the modal expression ist/sind zu ‘must’ in the German version, and two modal verbs in the Greek version (Table 1). In contrast to English and German, both of which possess rich inventories of modal verbs, Modern Greek has only two modal verbs marking the poles of necessity (ʌȡȑʌİȚ /prépi/ ‘must’) and possibility (ȝʌȠȡȫ /boró/ ‘may, can’)12 of the modal continuum.13 The modal verb ʌȡȑʌİȚ /prépi/ ‘must:3S’ is a defective verb limited to the third person singular.14 Table 1. Modal verbs/constructions occurring in the English, German, and Greek versions of the Charter of Fundamental Rights of the European Union Modal verbs/construc-tions

English shall 45 must

may

12

10

German soll/sollen 3SG/3PL muss/müssen 3SG/3PL ist/sind zu 3SG/3PL kann/können 3SG/3PL darf/dürfen 3SG/3PL

Total (tokens) 10

67

Greek 1 8

ʌȡȑʌİȚ /prépi/ 3SG

16

ȝʌȠȡİȓ/ȝʌȠȡȠȪȞ /borí//borún/ 3SG/3PL

14

8 8

14 39

30

At least in contemporary contract drafting this should read “is to be interrupted” (Keith F. Wilder, p.c.). 11 This modal meaning is limited to the perfective future (Panaretou 2009: 96-97). Thus, șĮ įȚĮțȩʌIJȠȞIJĮȚ ȠȚ İȡȖĮıȓİȢ /șa įiakóptode i erȖasíes/ ‘the construction works will (always) be interrupted’ is a non-modal statement about repeated future actions. 12 Modern Greek does not possess an infinitive so that the citation form of verbs is the 1S present form. 13 In a legal sense, ‘may’ expresses “privilege” (Keith F. Wilder, p.c.). 14 For a detailed analysis of Greek modality see Iakovou (1999).

Ursula Stephany

133

While use of modal verbs expressing deontic necessity predominates in the English and Greek versions of the Charter, in the German version, modal verbs expressing deontic possibility slightly prevail (Table 2). In legal texts, deontic necessity expresses duties or obligations and deontic possibility rights or “privileges”.15 In English legal language, shall and the stronger modal verb must are used for duties and to be entitled to and may for rights and privileges, respectively (Wilder 2012b: 89-90). The modal verb can does not occur in the English version of the Charter and is generally avoided in English legal texts for creating privileges (Wilder 2012b: 90). The main reason for the large difference between the frequency of modal verbs expressing deontic necessity in English on the one hand and Greek and German on the other, seems to lie in the use of shall, one of the fundamental characteristics of more traditional English legal language. According to Tiersma (1999) legal documents have at least one performative, express or implied, that indicates the overall force of the document: enacting (a statute), promising (a contract), and so forth. That which is being enacted or promised is indicated by shall.

He adds that the real reason for the pervasiveness of shall in legal language is that “it unambiguously indicates that something is intended to be legally binding” […], that it “is an enforceable legal obligation, not just an informal rule or unenforceable agreement.” Use of the modal verb must instead of shall “adds emphasis to the duty in question” (Wilder 2012a: 90). This explains why it is much less used in the English version of the Charter than shall (see Table 1). Table 2. Deontic necessity vs. possibility in the English, German, and Greek versions of the Charter of Fundamental Rights of the European Union Deontic modality necessity

shall/must

possibility

may

n (tokens) 15

English

German 85 % 15 % 67

muss/ist zu/soll kann/darf

Greek 44 % 56 % 39

ʌȡȑʌİȚ /prépi/ ȝʌȠȡİȓ /borí/

53 % 47 % 30

In contrast to rights, privileges are rights “without a duty corollary” giving “its holder a choice or permission to act” (Wilder 2012b: 90).

134

Official Versions of the Charter of Fundamental Rights of the EU

The prevalence of the use of the modal verb shall in the English version of the Charter results in another striking difference between the English version on the one hand and the German and Greek versions on the other. This consists in the role of the present indicative of full verbs as opposed to constructions containing a modal auxiliary (see Figure 1). The present indicative occurs in 61% of verbal expressions not containing a modal auxiliary in the Greek version and in 53% of such expressions in the German version, while in the English version it only amounts to 16%. Fig. 1. Expressions of deontic modality in the English, German, and Greek versions of the Charter of Fundamental Rights of the European Union (n = 85 Engl., 86 Germ., 85 Greek tokens)16

90% 80% 70% 60% 50%

Pres. ind.

40%

Modal verbs Other

30% 20% 10% 0% English

German

Greek

The main function of the present indicative of full verbs in standard language is to express non-modalized statements of fact. In standard German, the present tense may also be used in directives leaving no alternative to the addressee (example 2).

16

Besides the present indicative of full verbs and modal verb constructions other categories occurring in the Charter are the future tense and the conditional mood.

Ursula Stephany (2)

German

135

Du hörst jetzt sofort auf zu schreien. ‘You will stop shouting at once.’

In legal texts of a number of languages the deontically neutral present indicative is used atemporally and expresses the undisputed general validity and binding nature of laws (Panaretou 2009: 96; Moser and Panaretou 2011: 26; Tacke 2011; see also Wilder 2012a: 89). According to Lenz (2006: 73), the present indicative may function as a “semantic imperative” in such texts. Tacke (2011) speaks of the “imperative” or “legal present tense” which is adequately characterized by von Alemann (2006: 103; quoted by Tacke 2011) as follows: zusammen mit der rechtlichen Geltung wird aus einem einfachen Aussagesatz in einem performativen Akt eine Rechtsnorm” [along with its legal force, a simple statement is turned into a legal norm by a performative act].

Or in Cornu’s terms (Cornu 2000: 272, quoted by Burr and Gallas 2004: 235), “l’indicatif vaut l’impératif. C’est une particularité de l’énoncé législatif.” Cornu (2000: 272) provides an interesting explanation of this special feature of legal language: Psychologiquement, l’indicatif présent offre d’ailleurs des avantages. Il occulte celui qui donne l’ordre et ne brandit pas le pouvoir d’ordonner. C’est une façon plus discrète, plus douce et plus diplomatique de commander. La référence à ce qui est pourrait même faire imaginer que la règle énoncée n’est pas arbitrairement imposée, mais naturellement fondée, que le droit est proche de la nature des choses.

According to Burr and Gallas (2004: 236), a preference for using the present indicative in legal texts has recently been developing crosslinguistically. This is especially true of German, but also of French17 and Italian and even English,18 whereas in Swedish use of the modal auxiliary skall prevails. In examples (3), the present indicative found in article 9 of the German and Greek versions of the Charter typically corresponds to a modal construction with shall in the English version, obviously rendered in a more traditional legal style. It is this type of correspondence which is most frequently found in the Charter. 17

For French see also Bast (2006). In recent developments of legal English, use of the present tense has become more habitual. It is maintained that under the common law “the contract is always speaking” (Wilder 2012a: 89). 18

136

Official Versions of the Charter of Fundamental Rights of the EU (3) The Charter of Fundamental Rights of the European Union, Art. 9 (a) The right to marry and the right to found a family shall be guaranteed in accordance with the national laws governing the exercise of these rights. (b) Das Recht, eine Ehe einzugehen, und das Recht, eine Familie zu gründen, werden nach den einzelstaatlichen Gesetzen gewährleistet, welche die Ausübung dieser Rechte regeln. (c) ȉȠ įȚțĮȓȦȝĮ ȖȐȝȠȣ țĮȚ IJȠ įȚțĮȓȦȝĮ įȘȝȚȠȣȡȖȓĮȢ ȠȚțȠȖȑȞİȚĮȢ įȚĮıijĮȜȓȗȠȞIJĮȚ ıȪȝijȦȞĮ ȝİ IJȚȢ İșȞȚțȑȢ ȞȠȝȠșİıȓİȢ ʌȠȣ įȚȑʌȠȣȞ IJȘȞ ȐıțȘıȒ IJȠȣȢ.

A detailed analysis of the Charter of Fundamental Rights shows that there is no one-to-one correspondence between different modal expressions or the use of the present tense of full verbs in the three linguistic versions. Rather, the present indicative as well as each modal verb occurring in the three languages corresponds to a number of different expressions in each of the other ones. Thus, English shall is not only rendered by the German and Greek present indicative of full verbs, but also by German können ‘can’, dürfen ‘may’, and ist zu ‘must’ and by Greek ʌȡȑʌİȚ /prépi/ ‘it must’ as well as ȝʌȠȡİȓ /borí/ ‘it can, may’. In spite of the fact that the present indicative of full verbs is very frequently found in both the German and Greek versions of the Charter and even more frequently in Greek than in German (see Figure 1), its use in the Greek version is parallel to the German version in only 80% of tokens while the remaining 20% of Greek present tense tokens are rendered by four different modal verbs or modal constructions in German. The following examples point out such parallels and differences in the three authentic versions of the Charter of Fundamental Rights of the European Union. In examples (4), parallel constructions are found in the three languages: English must corresponds to German muss and to Greek ʌȡȑʌİȚ /prépi/.19 (4) The Charter of Fundamental Rights of the European Union, Art. 24 paragraph 2 (a) In all actions relating to children, whether taken by public authorities or private institutions, the child's best interests must be a primary consideration. (b) Bei allen Kinder betreffenden Maßnahmen öffentlicher Stellen oder privater Einrichtungen muss das Wohl des Kindes eine vorrangige Erwägung sein. 19

While English must und German müssen both express deontic necessity and thus strong obligation in this context, the overall functions of these genetically related verbs in the English and German systems of modal verbs differ.

Ursula Stephany

137

(c) Ȉİ ȩȜİȢ IJȚȢ ʌȡȐȟİȚȢ ʌȠȣ ĮijȠȡȠȪȞ IJĮ ʌĮȚįȚȐ, İȓIJİ İʌȚȤİȚȡȠȪȞIJĮȚ Įʌȩ įȘȝȩıȚİȢ ĮȡȤȑȢ İȓIJİ Įʌȩ ȚįȚȦIJȚțȠȪȢ ȠȡȖĮȞȚıȝȠȪȢ, ʌȡȦIJĮȡȤȚțȒ ıȘȝĮıȓĮ ʌȡȑʌİȚ ȞĮ įȓȞİIJĮȚ ıIJȠ ȣʌȑȡIJĮIJȠ ıȣȝijȑȡȠȞ IJȠȣ ʌĮȚįȚȠȪ.

In the German version of Art. 1, the modally stronger expression ist zu has been preferred to the modal verb müssen ‘must’ (examples 5). This expression makes it absolutely clear that there is no alternative to respecting and protecting human dignity. The same can be said of the English modal verb must as compared to the deontically weaker shall (Wilder 2012a: 90). (5) The Charter of Fundamental Rights of the European Union, Art. 1 (a) Human dignity is inviolable. It must be respected and protected. (b) Die Würde des Menschen ist unantastbar. Sie ist zu achten und zu schützen. (c) Ǿ ĮȞșȡȫʌȚȞȘ ĮȟȚȠʌȡȑʌİȚĮ İȓȞĮȚ ĮʌĮȡĮȕȓĮıIJȘ. ȆȡȑʌİȚ ȞĮ İȓȞĮȚ ıİȕĮıIJȒ țĮȚ ȞĮ ʌȡȠıIJĮIJİȪİIJĮȚ.

In spite of the fact that in most instances both German and Greek use the present indicative of full verbs in a parallel way, in Article 52 paragraph 4, the modal construction with ʌȡȑʌİȚ /prépi/ explicitly expressing a legal duty corresponds to the imperatively used legal present tense of a full verb in German. The English modal construction with shall also renders the obligation in an explicit way (examples 6).20 (6) The Charter of Fundamental Rights of the European Union, Art. 52 paragraph 4 (a) In so far as this Charter recognises fundamental rights as they result from the constitutional traditions common to the Member States, those rights shall be interpreted in harmony with those traditions. (b) Soweit in dieser Charta Grundrechte anerkannt werden, wie sie sich aus den gemeinsamen Verfassungsüberlieferungen der Mitgliedstaaten ergeben, werden sie im Einklang mit diesen Überlieferungen ausgelegt. (c) ȈIJȠ ȕĮșȝȩ ʌȠȣ Ƞ ʌĮȡȫȞ ȋȐȡIJȘȢ ĮȞĮȖȞȦȡȓȗİȚ șİȝİȜȚȫįȘ įȚțĮȚȫȝĮIJĮ ȩʌȦȢ ĮʌȠȡȡȑȠȣȞ Įʌȩ IJȚȢ țȠȚȞȑȢ ıȣȞIJĮȖȝĮIJȚțȑȢ ʌĮȡĮįȩıİȚȢ IJȦȞ țȡĮIJȫȞ ȝİȜȫȞ, IJĮ İȞ ȜȩȖȦ įȚțĮȚȫȝĮIJĮ ʌȡȑʌİȚ ȞĮ İȡȝȘȞİȪȠȞIJĮȚ ıȪȝijȦȞĮ ȝİ IJȚȢ ʌĮȡĮįȩıİȚȢ ĮȣIJȑȢ.

20

As far as English language contracts are concerned, Wilder (2011) suggests to reserve the use of shall for expressing the parties’ duties and to otherwise replace it by other constructions (e.g. is to).

138

Official Versions of the Charter of Fundamental Rights of the EU

In Article 52 paragraph 6 the German ist zu construction corresponds more closely semantically to Greek ʌȡȑʌİȚ /prépi/ ‘it must’ than either of them do to English shall (examples 7). Both the German and Greek texts add emphasis to the duty of taking account of national laws and practices, something which English shall does not (see Wilder 2012a: 90). (7) The Charter of Fundamental Rights of the European Union, Art. 52 paragraph 6 (a) Full account shall be taken of national laws and practices as specified in this Charter.21 (b) Den einzelstaatlichen Rechtsvorschriften und Gepflogenheiten ist, wie es in dieser Charta bestimmt ist, in vollem Umfang Rechnung zu tragen. (c) ȅȚ İșȞȚțȑȢ ȞȠȝȠșİıȓİȢ țĮȚ ʌȡĮțIJȚțȑȢ ʌȡȑʌİȚ ȞĮ ȜĮȝȕȐȞȠȞIJĮȚ ʌȜȒȡȦȢ ȣʌȩȥȘ ȩʌȦȢ țĮșȠȡȓȗİIJĮȚ ıIJȠȞ ʌĮȡȩȞIJĮ ȋȐȡIJȘ.

In Article 41 paragraph 4, the expression of deontic possibility by English may corresponds to German kann (examples 8). While in Standard English, may incorporates the meanings of both possibility and permission, in legal English, may represents “privilege” and can “possibility”.22 It is interesting to note that kann rather than darf has been chosen in the German version. German darf would only have rendered permission (“privilege”) while kann expresses both possibility and permission. English must, German muss and Greek ʌȡȑʌİȚ /prépi/ all express a legal duty. (8) The Charter of Fundamental Rights of the European Union, Art. 41 paragraph 4 (a) Every person may write to the institutions of the Union in one of the languages of the Treaties and must have an answer in the same language. (b) Jede Person kann sich in einer der Sprachen der Verträge an die Organe der Union wenden und muss eine Antwort in derselben Sprache erhalten. (c) ȀȐșİ ʌȡȩıȦʌȠ ȝʌȠȡİȓ ȞĮ ĮʌİȣșȪȞİIJĮȚ ıIJĮ șİıȝȚțȐ ȩȡȖĮȞĮ IJȘȢ DzȞȦıȘȢ ıİ ȝȓĮ Įʌȩ IJȚȢ ȖȜȫııİȢ IJȦȞ ȈȣȞșȘțȫȞ țĮȚ ʌȡȑʌİȚ ȞĮ ȜĮȝȕȐȞİȚ ĮʌȐȞIJȘıȘ ıIJȘȞ ȓįȚĮ ȖȜȫııĮ.

It should be mentioned that legal rights are explicitly stated in certain articles of the Charter by English to be entitled to, German Anspruch 21

In modern legal language, “is to be taken” would be more correct (Keith F. Wilder, p.c.; see also fn. 10). 22 Keith F. Wilder, p.c. See also fn. 13.

Ursula Stephany

139

haben auf ‘to be entitled to’, ein Recht darauf haben ‘to have a right to’ and Greek įȚțĮȚȠȪȞIJĮȚ /įikeúde/ ‘they are entitled to’, ȑȤİȚ įȚțĮȓȦȝĮ /éçi įikéoma/, ‘he/she has a right to’ (examples 9). (9) The Charter of Fundamental Rights of the European Union, Art. 47 (a) Everyone is entitled to a fair and public hearing within a reasonable time by an independent and impartial tribunal previously established by law. Everyone shall have the possibility of being advised, defended and represented. (b) Jede Person hat ein Recht darauf, dass ihre Sache von einem unabhängigen, unparteiischen und zuvor durch Gesetz errichteten Gericht in einem fairen Verfahren, öffentlich und innerhalb angemessener Frist verhandelt wird. Jede Person kann sich beraten, verteidigen und vertreten lassen. (c) ȀȐșİ ʌȡȩıȦʌȠ ȑȤİȚ įȚțĮȓȦȝĮ ȞĮ įȚțĮıșİȓ Ș ȣʌȩșİıȒ IJȠȣ įȓțĮȚĮ, įȘȝȩıȚĮ țĮȚ İȞIJȩȢ İȪȜȠȖȘȢ ʌȡȠșİıȝȓĮȢ, Įʌȩ ĮȞİȟȐȡIJȘIJȠ țĮȚ ĮȝİȡȩȜȘʌIJȠ įȚțĮıIJȒȡȚȠ, ʌȠȣ ȑȤİȚ ʌȡȠȘȖȠȣȝȑȞȦȢ ıȣıIJĮșİȓ ȞȠȝȓȝȦȢ. ȀȐșİ ʌȡȩıȦʌȠ ȑȤİȚ IJȘ įȣȞĮIJȩIJȘIJĮ ȞĮ ıȣȝȕȠȣȜİȪİIJĮȚ įȚțȘȖȩȡȠ țĮȚ ȞĮ IJȠȣ ĮȞĮșȑIJİȚ IJȘȞ ȣʌİȡȐıʌȚıȘ țĮȚ İțʌȡȠıȫʌȘıȒ IJȠȣ.

5. Conclusion A major result of the study of mood and modal verbs in the English, German, and Greek official and authentic versions of The Charter of Fundamental Rights of the European Union seems to be that pragmatic equivalence or isofunctionality does not mean that these texts are linguistically identical. Definitions of translations such as the one implicit in the title of Eco’s book “Dire quasi la stessa cosa” [Saying nearly the same thing] (Eco 2003) or of its German translation “Quasi dasselbe mit anderen Worten” [virtually the same in other words] therefore also apply to legal texts produced by co-editing or translation. Since the ways in which different linguistic communities and cultures express experience in linguistic terms differ (Slobin 1996), exactly identical meanings of different linguistic versions of a given text are excluded even for legal texts, which strive for maximum clearness and unambiguity. In consequence, “the legal policy that all language versions of EU law must be treated as equally authentic forces the interaction between the different language versions, as these are assumed to have a level of parity with each other” (Burr 2013: 1506).

140

Official Versions of the Charter of Fundamental Rights of the EU

Acknowledgements I would like to thank Eleni Panaretou for drawing my attention to the interesting field of modality in legal language, Isolde Burr for suggesting The Charter of Fundamental Rights of the European Union as a text more easily accessible to non-specialists in law, Keith E. Wilder for pointing out recent developments in legal English and the intricacies of the functions of modal verbs in this genre, two anonymous reviewers for further suggestions, and Sarah Agnes Marshall for correcting language inadvertences. All remaining inadequacies are my own responsibility.

References Bast, Jürgen. 2006. Grundbegriffe der Handlungsformen der EU entwickelt am Beschluss als praxisgenerierter Handlungsform des Unions- und Gemeinschaftsrechts. Berlin-Heidelberg-New York: Springer. Burr, Isolde. 2006. “Die Grundrechte-Charta: Ein europäischer Text.” In Gemeinschaftskommentar zur Europäischen Grundrechte-Charta, ed. Peter J. Tettinger and Klaus Stern, 187-198. München: C. H. Beck. —. 2009. “Linguistische Aspekte zu authentischen mehrsprachigen Rechtstexten.” Aktuelle Juristische Praxis (AJP) 6: 750-60. —. 2013. “Article 55 [Languages and Deposit of the Treaty].” In The Treaty on European Union (TEU), A Commentary, ed. Hermann-Josef Blanke and Stelio Mangiameli, 1461-525. Berlin-Heidelberg-New York-Dordrecht-London: Springer. Burr, Isolde and Tito Gallas. 2004. “Zur Textproduktion im Gemeinschaftsrecht.” In Rechtssprache Europas: Reflexion der Praxis von Sprache und Mehrsprachigkeit im supranationalen Recht, ed. Friedrich Müller and Isolde Burr, 195-242. Berlin: Duncker & Humblot. Charter of Fundamental Rights of the European Union/Charta der Grundrechte der Europäischen Union/ȋȐȡIJȘȢ IJȦȞ ĬİȝİȜȚȫįȦȞ ǻȚțĮȚȦȝȐIJȦȞ IJȘȢ ǼȣȡȦʌĮʀțȘȢ DzȞȦıȘȢ. Official Journal of the European Union, vol. 53, 30 March 2010; 2010/C 83/02, Amtsblatt der Europäischen Union, 53. Jahrgang, 30. März 2010; 2010/C 83/02, ǼʌȓıȘȝȘ ǼijȘȝİȡȓįĮ IJȘȢ ǼȣȡȦʌĮʀțȘȢ DzȞȦıȘȢ, 53Ƞ ȑIJȠȢ, 30 ȂĮȡIJȓȠȣ 2010; 2010/C 83/02. Cornu, Gérard. 2000. Linguistique juridique. 2nd ed. Paris. Eco, Umberto. 2003. Dire quasi la stessa cosa: Esperienze di traduzione. Milano: Bompiani.

Ursula Stephany

141

Gibbons, John. 2003. Forensic Linguistics: An Introduction to Language in the Judicial System. Oxford: Blackwell. Iakovou, Ȃaria. 1999. Tropikes katigories sto rimatiko sistima tis Neas Ellinikis [Modal Categories in the Modern Greek Verbal System]. Doct. Diss. University of Athens. Lenz, Magdalena. 2006. Grammatik und Stil: Das Passiv als stilistisches Mittel im Vergleich zu konkurrierenden grammatischen Konstruktionen. Doct. Diss. Technical University Berlin. Lyons, John. 1977. Semantics. 2 vols. Cambridge: Cambridge University Press. Moser, Amalia and Eleni Panaretou. 2011. “Why a Mother’s Rule is Not a Law: The Role of Context in the Interpretation of Greek Laws.” In Context and Contexts, ed. Anita Fetzer and Etsuko Oishi, 11-40. Amsterdam: Benjamins. Nida, Eugene A. 1964. Towards a Science of Translation: With Special Reference to Principles and Procedures Involved in Bible Translating. Leiden: Brill. Panaretou, Eleni. 2009. ȃȠȝȚțȩȢ ȜȩȖȠȢ: īȜȫııĮ țĮȚ įȠȝȒ IJȦȞ ȞȩȝȦȞ [Legal Language: Language and Structure of Laws]. Athens: Ekdoseis Papazisi. Ramat, Paolo. 2007. “The Problem of (Un)Translatability.” Rivista Italiana di Linguistica Applicata 39 (1-2): 21-32. Sayatz, Ulrike. 1996. “Modale Referenz in Gesetzen und Gesetzeskommentierungen: Ein textvergleichender Ansatz.” In Ebenen der Textstruktur: Sprachliche und kommunikative Prinzipien, ed. Walter Motsch, 275-81. Tübingen: Niemeyer. Slobin, Dan I. 1996. “From Thought and Language to Thinking for Speaking.” In Rethinking Linguistic Relativity, ed. John J. Gumperz and Stephen C. Levinson, 70-96. Cambridge: Cambridge University Press. Stephany, Ursula. 1986. “Modality.” In Language Acquisition: Studies in First Language Development, ed. Paul Fletcher and Michael Garman, 375-400. Cambridge: Cambridge University Press. Tacke, Konstantin. 2011. “Das imperative Präsens in der deutschen Rechtssprache.” Paper presented at the Linguistic Workshop Linguistik im Schloss V, Zeit, Schloss Wartin, June, 17-19, 2011. Tiersma, Peter M. 1999. Legal Language. Chicago: Chicago University Press. Visconti, Jacqueline. 2013. “European integration: Connectives in EU legislation.” International Journal of Applied Linguistics 23 (1): 44-59.

142

Official Versions of the Charter of Fundamental Rights of the EU

Vater, Heinz. 1975. “Werden als Modalverb.” In Aspekte der Modalität, ed. Joseph P. Calbert and Heinz Vater, 71-148. Tübingen: Narr. Von Alemann, Florian. 2006. Die Handlungsformen der interinstitutionellen Vereinbarung: Eine Untersuchung des Interorganverhältnisses der europäischen Verfassung. Berlin-Heidelberg-New York: Springer. Wilder, Keith F. 2011. “‘Bauhaus Drafting’: Tips and Tricks for Modern English Language Contracts.” Bonner Rechtsjournal (BRJ) (2): 79-80. —. 2012a. “‘Bauhaus Drafting’: Tips and Tricks for Modern English Language Contracts.” Bonner Rechtsjournal (BRJ) (1): 88-90. —. 2012b. “‘Bauhaus Drafting’: Tips and Tricks for Modern English Language Contracts.” Bonner Rechtsjournal (BRJ) (2): 89-90. Wüest, Jakob. 1993. “Die Sprache der Gesetze: Ein Beitrag zu einer vergleichenden Pragmatik.” In Studien zum romanisch-deutschen Sprachvergleich, ed. Giovanni Rovere and Gerd Wotjak, 103-20. Tübingen: Niemeyer.

THE TRANSLATION OF CONVERSATION AND FILM DUBBING AS A DISCOVERY PROCEDURE: EVIDENCE FROM DEMONSTRATIVES MARIA PAVESI UNIVERSITY OF PAVIA

It is notoriously difficult to obtain information about the translation of conversation, humans’ main form of verbal communication. This chapter argues that, among the various genres and registers that may be taken to approximate conversation, audiovisual dialogue provides the closest match. English film language is sociolinguistically rich, considerably overlaps with face-to-face conversation and presents selected features that are entrusted with the representation of orality. Despite the many constraints and limitations, the dubbing of audiovisual products thus offers a privileged access to the transfer of spontaneous spoken language. To exemplify how dubbing translation can be explored to search for translation patterns of spoken language, the second part of the chapter focuses on the translation of demonstrative pronouns in a parallel corpus of British and American films and their Italian translations. The analysis of the translation strategies will show that demonstratives are systematically reduced when dialogues are transferred from the source to the target language. The considerable number of omissions and substitutions observed in the corpus points to the relevant role of cross-linguistic contrast, since such translation shifts mainly take place within areas in which English and Italian differ syntactically and pragmatically. This investigation may open up novel avenues of research by showing the advantages of resorting to audiovisual translation as a means to shed light on transferability across spoken languages.

1. Introduction Conversation or, more generally, spontaneous spoken language is the basic type of human verbal communication, children’s first approach to their mother tongue as well as speakers’ most extensive experience of

144 Translation of Conversation and Film Dubbing as a Discovery Procedure

verbal language throughout their lives.1 Nonetheless, in its purely interactional and non-transactional realizations, conversation typically remains untranslated. This gap arguably results from the little compatibility between spontaneous conversation and translation. Conversation is in fact a transitory language expression, embedded in a shared spatio-temporal context, where the co-participants jointly contribute to topic development relying on the support of non-verbal codes. Most importantly, conversation is characterized by spontaneity, free topic development and free turn taking. Translation, on the contrary, is goal-oriented, normally a single person’s non-dialogic accomplishment; it is asynchronous and notably a pre-defined and other-initiated activity. Any act of translation will thus introduce variables that markedly differentiate the translation of conversation from conversation itself. The need is there, however, to gain information about the transferability of this key language variety. Admittedly, there are approximations to the translation of conversation. In cross-cultural encounters with immigrants and refugee seekers, for instance, portions of face-to-face conversation are translated, although the general aim of the interpretation process is not primarily that of reproducing the oral make-up of the original message. Among the many features that drastically distinguish these interactions from conversation, we find their highlighted transactional, institutional and humanitarian goals, together with the ingrained linguistic asymmetry between interactants. Other forms of translation come closer to the translation of spoken language and could be subjected to linguistic scrutiny. In television interviews with foreign guests, for example, a seemingly spontaneous verbal exchange is staged for the benefit of the non-participating audience. Several factors, however, limit the comparability of this genre with reference to translation. Among them, we can mention the lack of synchrony in the interpreter’s intervention, which is detached from the original language production. In addition, the interpreter is likely to deliver his/her translation while concentrating on the referential rather than the phatic and conative aspects of communication. Finally yet importantly, only one of the two interactants is usually translated in these crosslinguistic encounters. The unstructured, two-way interaction found in natural conversation is clearly not represented in these interactions on screen. More recently, supposedly live TV programmes (e.g. dog-training programmes, cooking programmes) are being dubbed. They offer some degree of overlap with conversation and could be used for the study of its 1

I would like to thank the editors of the volume for their valuable comments and suggestions, which were helpful in improving the chapter.

Maria Pavesi

145

translation. Yet, they mostly represent instances of language-in-action, where speech is contingent on what people are doing, on the setting and the objects present in the situational context. For this reason, such programmes represent just one subgenre of spontaneous spoken language. These last instances of media communication bring us close to film translation, by which I will here refer to dubbing. This translation modality entails the substitution of the original soundtrack with an “equivalent” target language soundtrack in films, TV series, soap operas, etc. , that is all those products that are structured around various forms of fictive dialogue (monologues, duologues, polylogues), embedded in a multimodal context and shown on screen. Cary (1960) has defined dubbing as “traduction totale”, with this label underlying the fact that everything in a film needs to be translated, i.e. intonation, voice quality, pragmatics, in keeping with the rigid and unchangeable frame provided by the visual images. Following Richet (2000: 92), it is my contention that cinema offers a privileged means for the study of features of spoken language in translation and contrastively. It must be pointed out, moreover, that films represent a natural and matchless source of parallel texts, parallel in the sense that the target language dialogue is supposed to reproduce the content as well as the modality of delivery of the source dialogue (Tomaszkiewicz 2001, 2009). In film translation, we are given the possibility of examining, turn by turn or utterance by utterance, what can be considered functionally equivalent expressions in the source and target texts. At the same time, texts are anchored to fixed scenes. These represent locations, situations, movements, gestures, gazes, dresses etc., a whole set of semiotic signs that do not change with the language change (Chaume 2004, 2012). Films hence offer unique opportunities for the synchronous reproduction of the matching of verbal and nonverbal languages. They provide invariant,2 often realistic situations, in contrast with the variable and abstract circumstances in which cross-cultural pragmatic use is often investigated (Coulmas 1979: 245). In the rest of the chapter, I will at first briefly present audiovisual dialogue and suggest why it can be useful in the investigation of spoken language. Dubbing translation as a discovery procedure will be discussed next, bringing attention to the features that can be profitably investigated through dubbing. In the second part of the chapter, I will focus on the 2

Situational identity across the language divide, however, is to be ruled out. Different cultural interpretations of the same visual input will apply, as given, among other factors, by cross-cultural variation in codified social events and by dissimilar collective values associated with the same images and representations.

146 Translation of Conversation and Film Dubbing as a Discovery Procedure

translation of demonstrative pronouns in a corpus of film dialogue as an exemplification of how audiovisual translation may shed light on issues of transferability across spoken languages.

2. Naturalness of audiovisual language Linguistic verisimilitude or naturalness are central to film dialogue as they are required for viewers to become immersed in the fictional world represented on screen (e.g. Pavesi 2008; Baños 2013). The naturalness of the original dialogue is also the premise for the study of the translation of spoken language through dubbing. It would be naïve, however, to start from the assumption that audiovisual language fully reproduces or reflects spontaneous spoken language and can be straightforwardly relied on to investigate the features of conversation in both source and target language. Audiovisual language is a register of its own, a hybrid variety in its collocation in the space defined by the spoken and written modes. Audiovisual language is constrained by the teleological structure of film (Remael 2003, 2008) and its communicative, narrative, generic and aesthetic functions (Kozloff 2000). For these reasons, it displays specific syntactic, lexical, discoursal, and pragmatic features (Taylor 1999; Rossi 1999, 2011), and is characterized by very few disfluency phenomena, increased discourse immediacy and reduced vagueness (Quaglio 2009; Forchini 2012; Pavesi 2012), as well as increased emotionality and conflictuality (Bednareck 2010, 2012; Freddi 2011).

2.1. Representation of orality Although film dialogue is fictive and exhibits register-specific features, several authors have issued a call to use the language produced by filmmakers and heard on TV as a source of reliable data for linguistic analysis. Murphy back in 1978 highlighted the sociolinguistic richness of audiovisual language, whereas Lakoff Tolmach and Tannen (1984) shifted the attention to the artistic realism of film dialogue. While introducing an analysis of conversational strategies in Ingmar Bergman’s film Scenes from a marriage (1973: 323), they argued that: [A]rtificial dialog may represent an internalized model or schema for the production of conversation – a competence model that speakers have access to. If, then, we are interested in discovering the ideal model of conversational strategy, there is much to be gained by looking at artificial conversation first, to see what these general, unconsciously-adhered-to

Maria Pavesi

147

assumptions are; and later returning to natural conversation to see how they may actually be exemplified in literal use.

In a similar vein, after carrying out an investigation of the TV series Star Trek, Rey (2001) proposed that audiovisual dialogue represents a valuable indicator of how conversation is perceived. Following AlvarezPereyre (2011: 48), film dialogue can be examined both as a language artefact, deriving from an artistic and social endeavour, as well as a specimen, that is, language data that can contribute to our language knowledge.

2.2. Approximation to spontaneous spoken language and metonymic orality Quantitative corpus investigations have recently confirmed some degree of approximation of audiovisual dialogue to spontaneous conversation. Quaglio (2009) carried out a multi-dimensional analysis of a large set of language features found in the sitcom Friends; Forchini (2012) adopted a similar approach in her study of a corpus of 11 American movie dialogue transcriptions, whereas Rodríguez Martín (2010) investigated the written-spoken divide in a corpus of 10 American films. These three major investigations share the large number of features considered and the systematic comparison with reference corpora of spoken English. Overall, they have found audiovisual conversation to be as involved and interactive, as affective and informal as face-to-face conversation. A degree of alignment between contemporary audiovisual dialogue and spontaneous spoken language can be taken as warranting scholars to investigate the translation of spoken language through dubbing. At the same time, an additional perspective is offered by a crucial characteristic of audiovisual dialogue: its selectedness. Audiovisual talk, in fact, has been posited to stage privileged carriers of orality (Pavesi 2005; 2008; cf. also Heyd 2010; Rossi 1999), well-recognizable markers capable of evoking spokenness in viewers’ minds (Fowler 2000). Thanks to a metonymic process, a few features conceived and perceived as typically oral are taken to stand for the whole in filmic discourse (cf. Bollettieri Bosinelli et al. 2005). In other words, selected cues or triggers activate the impression of spontaneous spoken language conveyed in film. They are individual linguistic features of texts—words, expressions, syntactic or morphological phenomenathrough which viewers cognitively reconstruct the missing parts of spoken communication and mentally experience orality. “Staged orality”, as it has also been called (Heyd 2010; Bednareck

148 Translation of Conversation and Film Dubbing as a Discovery Procedure

2010), allows researchers to identify well-represented features of spoken language and submit them to linguistic examination.

2.3. Dubbing as a discovery procedure Research has brought to the fore the sociolinguistic relevance of the data coming from the screen together with a certain degree of overlap between audiovisual language and spoken language. If a group of features can be identified that flag orality in audiovisual language, dubbing can then be capitalized on as a source of information on the translation of conversational language. This will apply in spite of the limitations of fictive orality to fully reproduce spokenness. In what follows, I will focus on English—the language of most audiovisual products produced and distributed around the world—and Italian, a dubbing language boasting a tradition of professional excellence as well as a vast array of scholarly and professional publications. Given the specificity of dubbing, it is not possible, however, to study the translation of spontaneous spoken language tout court. As mentioned above, various phenomena such as eufluency disalign original and dubbed film dialogue from spontaneous conversation. They should not be considered in the investigation of the translation of spoken language or should be handled with great care. Moreover, we cannot straightforwardly generalize the patterns emerged in dubbing to spontaneously occurring interactions, in that audiovisual translation is constrained and normed by specific factors that do not apply to real life situations. This in particular concerns lip synchronization and isochrony, i.e the same duration of the utterances in the original and the dubbed texts (Pavesi 2005; Chaume 2012). If, however, due attention is paid to the choice and the analysis of individual features or clusters of related features performing a key role in speech, the study of dubbing can unveil relevant and systematic translational patterns. Such study may help us define the degree of translatability of given phenomena, highlight problems of translation of spoken language and identify the consequences of cross-linguistic contrasts. Firstly, spoken language in translation can be studied focussing on the source language and investigating how typical spoken features are transferred to the target language. Examples from English include that deletion before complement clauses, contractions and situational ellipsis, all phenomena that have to do with the situational parameters of online production and sharedness of context (e.g. Quaglio and Biber 2006). Secondly, the emphasis may be placed on the target language and we may

Maria Pavesi

149

reasonably ask if typical features of Italian speech, such as multifunctional che ‘that’ or the negative particle mica ‘at all’, ‘a bit’,3 are introduced in Italian translations to render English spoken features. Thirdly, some phenomena can be recognized as shared between source language and target language as linguistic correlates of the situational parameters of face-to-face communication across languages (e.g. Miller and Weinert 1998; Biber et al. 1999; Rühlemann 2007). Among them, we find greater freedom in word orders, high frequency of deictic expressions, and the construction of discourse through small blocks of meaning. To show how the study of dubbing can contribute to the investigation of the translation of spoken language, I will focus on demonstrative pronouns, a group of features that characterize both spoken English and spoken Italian.

3. Demonstrative pronouns Demonstrative pronouns strongly bind the dialogue to the linguistic and extra-linguistic situation shared by the interlocutors, thus creating context-embeddedness and discourse cohesion (e.g. Halliday and Hasan 1976; Levinson 1983; Diessel 1999, 2006). More specifically, Diessel (2006, 2014) identifies demonstratives as the means that prototypically and universally index speakers’ “joint focus of attention”. According to the author, demonstratives fulfil two, interrelated functions: the first one is to locate the referent relatively to the deictic centre, the second “to coordinate the interlocutors’ joint attentional focus” as essential to the “meeting of minds” that occurs in human communication (p. 468-469). In addition, demonstratives contribute to the co-construction of dialogue and convey emotionality. For this reason, they are key interactional features in conversation (Carter and McCarthy 2006). Being universal and very frequent features of spoken language (Dixon 2003; Diessel 2006, 2014), demonstrative deictics fulfil the requirement of the third category outlined above, which comprises conversational phenomena shared across languages. In addition, demonstratives qualify as an area to be profitably investigated in terms of the translation of spokenness, not only for the crucial functions they perform in speech (cf. Pavesi 2013), but also because of their considerable variation across

3 But also ‘(not) much’, ‘surely’. These translations do not cover all the uses of mica. In interrogatives, mica might also be translated by ‘perhaps’ or paraphrased with ‘do/did you happen to...?’ and ‘by any chance’; e.g. Mica ce l’hai una biglietto dell’autobus? ‘You don’t have a bus ticket by any chance?’. I thank Paolo Ramat for drawing my attention to this point.

150 Translation of Conversation and Film Dubbing as a Discovery Procedure

languages both as systems and in their uses (e.g. Diessel 1999; Dixon 2003; Da Milano 2005).

3.1. Demonstrative pronouns in English and Italian English and Italian share a binary system of proximal and distal demonstrative pronouns: this-these and that-those in English, the two series quest- and quell- in Italian.4 In both languages, their use depends among other factors on the proximity or distance of the reference object in relation to the speaker or the speaker and hearer (e.g. Da Milano 2005; Diessel 1999, 2006). The choice of demonstrative, however, is known to be also influenced by contextual and affective factors (e.g. Lakoff 1974; Carter and McCarthy 2006). English and Italian demonstratives can be distinguished into two main classes depending on their pragmatic uses (Halliday and Hasan 1976; Diessel 1999, 2006).5 Exophoric demonstratives are prototypical deictics focussing interlocutors’ attention onto items of the situational context and can include concrete objects, as well as abstract situations, conditions and states of mind. Endophoric demonstratives, on the other hand, refer to textual elements and can be anaphoric or discourse deictics. In both spoken English and Italian, they most often perform a discourse deictic function, i.e. they shift hearers’ focus of attention onto textual portions whose delimitations surpass individual elements (cf. Botley and McEnery 2001; Gaudino-Fallegger 1992). In terms of language contrast, English and Italian demonstrative pronouns diverge in several respects. English requires an overt subject, which in spoken language is often a demonstrative. By contrast, as Italian is a null-subject language, demonstrative pronouns carry a special pragmatic weight when they are used as subjects. Object pronouns as well exhibit a different syntactic behaviour in the two languages: in Italian they are most often placed preverbally if they are clitics, whereas they are placed postverbally in English (cf. Non ci credo vs. I don’t believe it). Clitics are weaker forms of reference, available when no emphasis is 4

The reinforcement particles qui/qua/lì/là ‘here’/’there’ are not taken into account in the present discussion due to the very limited occurrences in the database used for analysis. 5 It must be underlined that the behaviour of deictic systems relies on different parameters and varies greatly across languages. In Turkish for instance “a threeway demonstrative system (bu, úu, o) obligatorily encodes both distance contrasts (i.e. proximal and distal) and absence or presence of the addressee’s visual attention on the referent” (Küntay and Özyürek 2006: 303; see also Ramat 2014).

Maria Pavesi

151

intended. In general, Italian demonstrative pronouns are strong forms of reference, occurring when entities are emphasized or difficult to access, as when topics are introduced for the first time in discourse or two referents are in competition for anaphora (Berretta 1985; Serianni 1991; Cardinaletti 2009: 44).

3.2. Methodology and translation strategies The empirical basis for corpus analysis is provided by the first version of the Pavia Corpus of Film Dialogue (PCFD), a bilingual unidirectional parallel corpus comprising the transcriptions of 12 original film dialogues and their dubbing translations, for a total of about 230,000 words (Freddi and Pavesi 2009). Table 1 reports the most relevant information about the films included in the corpus. All English demonstrative pronouns in the corpus were manually annotated for syntactic role  subject, direct object, object of preposition  and pragmatic function  exophoric, endophoric  in the source texts, along with translation strategy from English into Italian. Exophoric demonstratives have been taken to include concrete entities as well as more abstract phenomena such as the situations portrayed on screen. Endophoric demonstratives in the corpus are almost exclusively represented by discourse deictics (Pavesi 2013). The degree of translatability of demonstrative pronouns in dubbing from English into Italian has been evaluated by comparing the source and the target texts. The emerging patterns point to translation regularities in spontaneous spoken language, which could be further investigated in other contexts of use. At the same time, dubbing translations are likely to bring to light equivalent resources available in the target language and may reveal cross-linguistic contrasts. The major strategies translators implement when rendering demonstrative pronouns into Italian can be grouped into full translation, omission, explicitation, compensation and substitution, as exemplified in table 2.

152 Translation of Conversation and Film Dubbing as a Discovery Procedure

Table 1. The Pavia Corpus of Film Dialogue Film title and year of release

Film director

Year No. of No. of Translatorof running running dialogue release words words writer in (Eng) (It) Italy Ae Fond Kiss Ken Loach 2005 8,624 8,828 Federica (2004) Depaolis Bend it like Gurinder 2002 9,582 9,729 Elettra Beckham (2002) Chadha Caporello Billy Elliot (2000) Stephen 2001 5,507 5,354 Carlo Casolo Daldry Crash (2004) Paul Haggis 2005 8,709 9,639 Filippo Ottoni Dead Man Tim 1996 12,447 11,382 Loretta Bertini Walking (1996) Robbins Erin Brockovich Steven 2000 13,714 12,842 Marco Mete (2000) Soderberg Finding Forrester Gus Van 2001 10,379 10,591 Elettra (2000) Sant Caporello Notting Hill Roger 1999 10,438 9,661 Francesco (1999) Mitchell Vairano Ocean’s Eleven Steven 2001 9,784 9,273 Marco Mete (2000) Soderberg One Hour Photo Mark 2002 5,731 5,231 Carlo Valli (2002) Romanek Secrets and Lies Mike Leigh 1996 13,229 11,926 Elisabetta (1996) Bucciarelli Sliding Doors Peter Howitt 1998 8,882 8,339 Francesco (1997) Vairano Total 117,956 111,865

Maria Pavesi

153

Table 2. Translation strategies for demonstrative pronouns in the Pavia Corpus of Film Dialogue Full translation Girl That’s the best bit!

Omission Ed This is the only thing you got?

Explicitation Erin I’m.. I’m telling you the, the truth, and I, I will get to the bottom of all of this.

Quella è la parte migliore! that is the part best [That is the best part!] (Bend it like Beckham) È l’ unica cosa che hai? is the only thing that havePRS.IND.2SG [Is it the only thing you have?] (Erin Brockovich) Ti dico, ti dico you say-PRS.IND.1SG you sayPRS.IND.1SG la verità e andrò in the truth and go-FUT.1SG in

Jess

[…] They think I’ve got a job at HMV.

fondo a questa storia. bottom to/at this story [I’m telling you the truth and I will go to the bottom of this matter.] (Erin Brockovich) […] Credono che abbia believe-PRS.IND.3PL that haveAUX.PRS.SBJV.1SG trovato un lavoro per l’ estate found a job for the summer [They think I have a summer job.]

Mel

Blimey, that’s not on.

Le bugie non si dicono the lies not PRON.REFL.3PL sayPRS.IND.3PL [You shouldn’t tell lies.] (Bend it like Beckham)

154 Translation of Conversation and Film Dubbing as a Discovery Procedure

Compensation sono lavori dove Rosalind Well, there may be jobs Forse ci where you can disappear Perhaps there are jobs where for days at a time, but puoi scomparire per this isn’t one of them. can-PRS. IND.2SG disappear for giorni e giorni ma qui days and days but here non puoi far=lo. not can-PRS.IND.2SG do=it [Perhaps there are jobs where you can disappear for days, but here you can’t.] (Erin Brockovich) Substitution Roisin I don’t know, but who knows that?

Non lo so! Chi not it know-PRS.IND.1SG who può saper-lo! can know=it [I don’t know! Who can know it?] (Ae Fond Kiss)

Erin No, you told me I’d be set. No, mi ha detto no me have-AUX.PRS.IND.3SG sayPST.PTCP

Ed

Never said that.

che ero a cavallo. that be-PST.IPFV.1SG at horse [No, you told me I was set.] Non ho mai detto not have-AUX.PRS.1SG never sayPTCP.PST questo this [I have never said that.] (Erin Brockovich)

Maria Pavesi

155

English demonstrative pronouns may be rendered with corresponding ones in Italian, i.e. proximals with proximals and distals with distals. In most cases, however, translation involves omissions and translation shifts. When omission occurs, the demonstrative disappears from the text and is not compensated for, nor replaced by, other deictic or referential devices in the immediate context. Demonstrative pronouns may also be translated via explicitation, realised partially as the replacement of the demonstrative pronoun with a demonstrative determiner within a more explicit noun phrase, or fully through the substitution of the demonstrative with nondeictic, full lexical items. In addition, the loss of a demonstrative pronoun can be compensated for within the same utterance by means of other deictic or focussing structures. Moving now to a less radical restructuring of the utterance where a demonstrative was originally embedded, two further strategies appear to be available to translators. Firstly, when in object position, demonstrative pronouns can be replaced by a weak, clitic pronoun performing a lower focussing function, similar to English it. Secondly, a demonstrative pronoun in the original text can be replaced with a pronoun of the other series: i.e. a distal demonstrative is translated with a proximal one, or vice versa. In the present context, a mainly qualitative and cross-linguistic approach will be pursued by examining the translational behaviours through the investigation of the most common and significant translation strategies, whereas detailed quantitative analysis of the results can be found in Pavesi (2013). The following discussion of translation strategies will focus at first on the strategies that reveal the low degree of translatability of demonstrative pronouns in spoken language: omission, explicitation and compensation. Subsequently, I will pay attention to exchanges of pronouns, as indexing interesting cases of cross-linguistic contrast. 3.2.1. Formal equivalence For the most part, English demonstrative pronouns are not translated with Italian demonstrative pronouns. This is the most striking finding of the analysis. Altogether, no more than a fifth of the 1678 demonstrative pronouns in the source texts are transferred to the target texts. This in itself is a very noticeable result that suggests a low degree of translatability of the feature from spoken English into spoken Italian. Cross-linguistic contrast is further shown by the small percentage of demonstrative pronouns (13%) that is translated with formally equivalent pronouns in Italian. The mismatch between the two languages more specifically

156 Translation of Conversation and Film Dubbing as a Discovery Procedure

pertains to distal pronouns. Only 6% distal that and those pronouns in the original films are rendered with the corresponding distal pronouns in Italian, whereas proximal this and these pronouns are translated with the corresponding proximal demonstratives relatively more frequently, i.e. 28% of the times. 3.2.2. Omission The very high proportion of omissions or deletions provides an unequivocal indication that demonstrative pronouns are an area of low degree of translatability between the two spoken languages. Quite remarkably, omission is indeed the most common translation strategy in the corpus, accounting for 42% of all translations and affecting both proximal and distal forms. The trend applies to distal pronouns to a greater extent than to proximal pronouns, with respectively 46.2% and 31.5% of all translations being omissions. For sure, the constraints of lip synchronization and isochrony favour the omission of the pronouns.6 However, the considerable frequency of this strategy and the different figures for the deletion of that and this cannot be fully accounted by such constraints. Hence, systematic omission also brings to light a lack of syntactic and pragmatic equivalence between these deictics in the two languages. Due to the diffusion of omissions in the corpus, a closer look was given to the syntactic roles of the demonstrative pronouns that are not transferred to the target texts. The aim of the analysis was to verify whether the syntactic-pragmatic mismatch between the two languages could partially account for the observed translational behaviour (Pavesi 2013). The distribution of the translation strategies according to the syntactic role of the demonstrative pronoun in the source text does substantiate the prediction that the structural and pragmatic differences between the two languages have an impact on translation outcomes. First, omission is by far the most frequent treatment of demonstrative subject pronouns: 39.4% this and 58.7% that subject pronouns disappear in the Italian translations. Second, demonstratives are deleted most frequently when they are in subject rather than direct or prepositional object position.

6 First, to a dental voiced fricative in English (th-) corresponds a voiceless labiovelar in Italian (qu-); second, the one-syllable this and that correspond to the two-syllable quest- and quell-, with obvious differences in the articulation of these items in the two languages. The constraints of lip synchronization essentially apply only in close-ups and full close-ups.

Maria Pavesi (1)

TESS:

(2)

TOPHER:

157

I don’t have a cell phone. This isn’t mine. Io non ho un cellulare. Non è mio. [I don’t have a cell phone. It’s not mine] (Ocean’s Eleven) Oh hey Russ aah we got another player, if that cool with you. Oh ehi! Rusty, abbiamo un altro giocatore, se per te va bene. [Oh hey! Rusty, we have another player, if for you it’s fine] (Ocean’s Eleven)

This is hardly surprising, since Italian is a null subject language and many of the English demonstratives introduce conversational routines that are translated idiomatically in Italian by dropping the subject demonstrative (cf. Ex. (2)). What is worth pointing out is the general impact on translation. Since subjects are by and large the most frequent syntactic role in the Anglophone dialogues, their deletions contribute to explaining why omissions are so widespread in the whole corpus. The strategy, however, applies to all syntactic positions, although to a lesser extent than in subject position: (3)

ANNA:

(4)

RUSTY:

You don’t mean that. Non mi dirai sul serio? [You are not being earnest?] (Sliding Doors) Slightly more complicated than that. Appena appena un po’ più complesso. [Just a bit more complex.] (Ocean’s Eleven)

3.2.3. Explicitation Explicitation is another strategy by which demonstrative deixis is reduced when translating the English dialogues into Italian. 12% of all demonstratives in the source language are rendered through explicitations. The incidence of the strategy provides additional evidence of the low degree of translatability of these deictic devices from English into Italian. Explicitation mostly affects direct and prepositional objects rather than subjects. In particular, it is prepositional objects that are made more explicit when shifting from English into Italian. This was to be expected since objects are syntactically compulsory in Italian when required by the argument structure of the verb and, unlike subjects, cannot be left unexpressed. For reasons that deserve further investigation, demonstrative pronouns in these positions often appear to be felt as unnatural and are

158 Translation of Conversation and Film Dubbing as a Discovery Procedure

substituted with lexically more explicit forms of reference. In many of these cases, the original demonstrative picks out a given situation or refers to filmic narration. The Italian translation hence makes the diegetic dimension of the fictive dialogue more explicit by replacing the demonstrative pronoun with a demonstrative determiner followed by a descriptive lexical label. In (5), questa vicenda ‘this affair’ more clearly identifies convict Matthew Poncelet’s vicissitudes and in (6), l’accaduto ‘what happened’ exophorically refers to the events shown on screen (somebody has just tried to steal a book in William’s shop). (5)

HELEN’S MOM:

That has nothing to do with that, I’m simply curious. Helen, what has drawn you to this? Questo non è importante, ma toglimi una curiosità che cosa è che ti attrae tanto in questa vicenda? [This is not important, but out of curiosity, what is it that attracts you so much in this affair?] (Dead Man Walking)

(6)

WILLIAM:

I’m sorry about that. Mi dispiace per l’accaduto.7 [I’m sorry for what happened.] (Notting Hill)

Explicitation also applies to fixed phrases in the source texts. In questo modo ‘in this way’ is a case in point, used repeatedly to render the vague expressions like this / like that, very frequent in conversational English. (7)

CASIM:

You’re not hanging about here dressed like that in front of me. Just leave! Non avrai mica intenzione di rimanere qui vestita in questo modo? Fila a casa. [You don’t intend to remain here dressed in this way, do you? Get straight home.] (Ae Fond Kiss)

3.2.4. Compensation That demonstrative pronouns do not translate easily from English into Italian is also shown by the many instances of compensation. With this translation strategy, the demonstrative pronoun is replaced with another focussing device that realises a “joint focus of attention” (Diessel 2006). In this way, deixis is preserved but expressed differently. That is, the Italian translation holds the audience’s attention verbally but resorts to other linguistic means to highlight entities in discourse. Through compensation, 7

A translation like Mi dispiace per questo / di ciò would indeed sound very odd.

Maria Pavesi

159

discourse correspondences beyond formal equivalences across languages come to the fore, especially when more than one translator operates the same choice. For example, demonstrative adverbs such as lì ‘there’, qui ‘here’, and così ‘so’ / ‘like this’ / ‘like that’ often translate demonstrative pronouns in English. (8)

WILLIAM:

This is the place. Qui sei al sicuro. [Here you are safe] (Notting Hill)

(9)

CHRISTINE:

What? I got scared, Cam! It’s not like I... I haven’t been pulled over before. You know? But not like that. Che c’è? Ho preso paura, Cam! Certo, è vero... già altre volte sono stata fermata, lo sai. Però non così. [What is it? I got scared, Cam! Sure, it’s true… I was stopped other times in the past, you know. But not like this.] (Crash)

(10)

JESS:

You shouldn’t say that about your dad. Non devi parlare così di lui.8 [You mustn’t speak like this about him.] Bend it like Beckham)

Various focussing devices can also be used instead of demonstratives, such as the imperative verbs guarda ‘look’ or senti ‘listen’ used as discourse markers. They represent interesting solutions to the low degree of transferability of demonstrative pronouns in this representation of orality. (11)

8

RUKHSANA:

Oy! I heard that he’s over six feet tall, he’s a fair, has a fair complexion and he’s got a full head of hair. Ehi! Ho sentito dire che è alto più di un metro e ottanta, che ha un bel portamento, molto elegante, e che ha tantissimi capelli.

The more literal and acceptable (although more formal) translation would have been Non dovresti dire questo/ciò di tuo padre. Yet, the translator in this case too opts for a different solution that gets rid of the pronoun. Morphologically, the difference between say (trans.) and parlare (intrans.) obliges Italian to use an ADV instead of an OBJ like that.

160 Translation of Conversation and Film Dubbing as a Discovery Procedure

TAHARA:

RUKHSANA:

[Hey! I’ve heard that he is more than six feet tall, that he has a good posture, very elegant, and he’s got a full head of hair.] Ewww! Uuhh! [Ewww!] What d’you mean? That sounds nice! Perché fai cosi? Guarda che è carino! [Why do you behave like this? Look how nice he is!] (Ae Fond Kiss)

Another interesting case is provided by the presentative adverb ecco, roughly ‘here’/‘there’ (Serianni 1991: 509-511). Ecco is frequent and natural in spoken Italian (cf. Rossi 2011: 33-34) where it may perform the functions of pointing to a new referent or drawing attention to something previously said. The adverb can be followed by a clitic pronoun, like in eccola ‘here/there she/it is’ or eccoci ‘here we are’: (12)

DONNA:

Because I said so, that’s why! Now get out! Out! Now! That’it! Okay. Okay. Okay girl. Perché lo dico io, ecco perché! Uscite e basta! Avanti! Su! [Because I say it, that’s why! Just get out! Come on! Hurry up!] (Erin Brockovich)

(13)

WILLIAM:

Er, eighteen yards. That’s my house there, with the blue front door. Oh, diciassette metri. Eccola casa mia, quella con il portone blu. [Oh, seventeen meters. There is my house, the one with the blue front door.] (Notting Hill)

(14)

SCOTT:

Oh, well, uh.. yeah, yeah, th-thanks. Eh. Well, this is it. Oh? Sì.. grazie. Eccoci qua. [Oh? Yes .. thanks. Here we are.] (Erin Brockovich)

Although compensation accounts for only 10.3% of all translations, it is a meaningful strategy as it indexes what appears to be textual or discourse synonymy across spoken languages. In other words, compensation discloses how different phenomena may perform similar highlighting and pointing functions in conversation in different languages.

Maria Pavesi

161

3.2.5. Substitution Another strategy that involves the loss of demonstratives is the replacement of demonstratives with clitics. Translation by clitics applies to 12.0% of all original demonstrative pronouns, and, as expected, is concentrated on direct objects for which it provides a natural solution. Distal pronouns again are involved in more shifts than proximal pronouns and more than 40% of all instances of that in direct object position become clitics in the target language. Due to their frequency, clitics can hence be considered equivalent resources available in spoken Italian both when reference is made to an object present in the perceptible context or to a segment of discourse. Translation by clitics reflects the role played by these weak pronouns in Italian, where they may perform the same endophoric and exophoric functions as demonstrative pronouns, although with a lower deictic force (cf. Berretta 1985; Gaudino-Fallegger 1992). Notice that in (15) and (16) the pre-verbal clitic lo ‘it’ translates the demonstrative that of the original dialogue. In the Italian translations, clitic pronouns also appear within left or right dislocations, like in (17) where the clitic lo refers to the same object picked out by the proximal demonstrative questo.9 (15)

GEORGE:

Hey, baby. Let me tell you something. Let me get that out of your hair. Ehi, tesoro, devo dirti una cosa. Aspetta che te lo tolgo dai capelli. [Hey, Darling, I must tell you something. Wait, I’ll take it out of your hair] (Erin Brockovich)

(16)

NINA NORKIN:

Because, Will’s been trying to get me to go digital and…

MR PARRISH:

9

Will vuole convincermi a prenderne una digitale ma io… [Will wants to convince me to get a digital one but I …] Oh, don’t do that, I’d be out of a job! Oh, you have one shot left. Oh, non lo faccia. Perderei il lavoro! Oh, è rimasta una posa. [Oh, don’t do it. I would lose my job. Oh, there is one shot left.] (One Hour Photo)

The right-dislocated demonstrative provides a very natural translation solution in Italian, whereas the simple demonstrative (i.e. Ti ricordi questo) would sound sociolinguistically marked.

162 Translation of Conversation and Film Dubbing as a Discovery Procedure (17)

CYNTHIA:

I thought I might move in here. Front view. See the world go by. Remember this? Pensavo di spostarmi in questa camera con la vista sull’esterno, vedi la gente che passa. Te lo ricordi questo? [I thought of moving into this room with a view on the outside, you can see people go by. Do you remember this?] (Secrets and Lies)

The replacements within the class of demonstratives bring to the fore other interesting cross-linguistic differences in the pragmatic use of these pronouns. If English proximal pronouns are always translated by Italian pronouns of the same series, English distal pronouns are more frequently translated with demonstratives of the other series. That is, 65 of all that and those become quell- in dubbing, while more than the double  i.e. 150  have been shifted to Italian quest-. In addition, a closer look at the individual occurrences shows that the great majority of shifts from distal that to proximal quest- involve an endophoric reference,10 almost exclusively cases of discourse deixis. Conversely, only a minority of endophoric that are translated with endophoric quell- in Italian. In English that is employed to pick up some other speaker’s discourse, thus serving an interactional purpose and tightening conversational turns together. By contrast, the same function of dialogic endophora is mostly performed by questo in the Italian translations, in line with what has been suggested for non-translated spoken Italian (Gaudino-Fallegger 1992). (18)

BELLA:

HONEY:

Yeah. But on the other hand, her best friend is Anna Scott. Però, d’altro canto, la sua migliore amica è Anna Scott. [But, on the other hand, her best friend is Anna Scott.] Oh well, that’s true, I can’t deny it. I mean, she needs me. What can I say? Oh, questo è vero, non posso negarlo, ha bisogno di me, che posso dire? [Oh, this is true, I cannot deny it, she needs me, what can I say?] (Notting Hill)

The textual function of that is so evident that the pronoun is also used within turns, to refer back to something just uttered by the same speaker. In this case too, the Italian translations contain a proximal form.

10

113 out of 148 shifts.

Maria Pavesi (19)

WILLIAM:

163

I suppose that in the, er, dream... dream scenario, I just, er, change my personality because you can do that in dreams and, um, walk over and, er, kiss the girl. But, er... Immagino che nel… sogno… di mia sceneggiatura io, ecco… cambio personalità, perché… nei sogni questo lo puoi fare. Mi avvicino e… bacio la ragazza, ma… [I imagine that in my …dream scenario, I, just ... change personality, because you can do this in dreams. I go closer and kiss the girl.] (Notting Hill)

3.2.5.1. Contrasts in exophoric deixis The analysis of exophoric that revealed additional cross-linguistic contrasts between English and Italian. The multimodality of films is very useful here since it presents the wider context in which interaction takes place and, more specifically, where demonstratives are embedded. Characters’ movements, postures, gazes and pointing gestures are accessible to viewers and translators alike. The same is true of the size, shapes, physical locations, etc. of the reference entities. At the same time, the interactional dimension  which has emerged as crucial for the choice of the demonstrative  is represented on screen for the translator to evaluate and keep into account. Starting with straightforward equivalences between the two languages, 31 exophoric that pronouns are translated with equivalent distal pronouns in Italian. In the following extract, for example, reference is made to a person who was present in the scene but then moved away. Both English and Italian express the spatial as well as temporal distance from the speakers with a distal pronoun. (20)

JESS:

Was that the club chairman? Quello era il presidente? [Was that the president?] (Bend it like Beckham)

More generally, distal pronouns are used in both languages when speaker and hearer are oriented towards an entity falling outside the trajectory linking their two positions. In the following extract from the corpus, the distal demonstrative in the two languages is employed to bring interlocutors’ attention to converge on an entity removed from the dyad’s sphere of influence. (21)

RYAN:

But there’s nobody else here yet and that’s gasoline there. We need to get you outta here right away. Okay? Ma non c’è nessun altro qui adesso. E quella lì è benzina. Perciò io devo tirarla fuori subito. Okay?

164 Translation of Conversation and Film Dubbing as a Discovery Procedure [But there is nobody else here now. And that there is petrol. For this reason I must get you out right away. Okay?] (Crash)

Although to a more limited extent, however, exophoric that pronouns too are shifted in the translations from English into Italian. In the PCFD, 35 that pronouns are rendered by a proximal questo o questa. The instances in which translations introduce a replacement of demonstrative suggest a systematic divergence in the two conversational languages. Typically, a situation is portrayed in which both speaker and hearer are close one to the other and are both near the reference object. In the birthday party scene in Secrets and Lies (Mike Leigh, 1996), Roxane, who has just arrived, hands Maurice, her uncle, a bottle of champagne, while uttering “I got you that”. The distal demonstrative in English brings attention onto the receiver, whose point of view is taken. In Italian, by contrast, spatial proximity and speaker’s orientation appear to prevail, with the proximal demonstrative being the only possible choice. Notice that Roxane is actually still holding the bottle while she utters Ti ho portato questo ‘I brought you this’. In another exchange from One Hour Photo (Mark Romanek, 2002), the different coding in the two languages is quite striking. Sy is sitting in a restaurant inside the mall where he works. He is looking at some photos while a waitress he apparently knows well approaches him. She takes the pictures in her hands and when pointing at one of them, she remarks “That’s a good shot!”, which in Italian becomes bella questa! ‘This is a nice one!’ (22)

WAITRESS, pointing specific picture:

to

a

That’s a good shot! Bella questa! [This is a nice one!] (One Hour Photo)

Immediately afterwards, when the waitress asks him if the people in the photos are his relations, Sy, almost touching one of the pictures and orienting his gaze on the photo, replies “Yes. That’s, my little nephew, Jake”. In Italian it becomes Sì! Questo è il mio nipotino Jake, ‘Yes. This is my little nephew, Jake’. (23)

WAITRESS:

These your, uh, relations? Sono tuoi parenti? [Are they your relatives?]

Maria Pavesi MR PARRISH:

165

Yes. That’s, my little nephew, Jake Sì! Questo è il mio nipotino Jake. [Yes! This is my little nephew Jake.] (One Hour Photo)

The two scenes show the dyad of interlocutors focussing on relevant objects situated within a sphere comprising both speaker and hearer: a present and some photos. The interactional function of the demonstrative is clear as the pronoun indexes a verbal demonstration by the speaker for the benefit of the hearer. The interactional function of the demonstrative used in close face-toface exchanges also applies to the following extract, where, however, the demonstrative is set in a context of challenge. The scene is taken from Notting Hill (Roger Mitchell, 1999). William is trying to get to the press conference where Ann is being interviewed. To that aim, he claims he is a member of the press and shows a Blockbuster card to the hotel receptionist. The reception’s gaze is clearly focussed on the card while he coldly and disapprovingly utters “that’s a Blockbuster Video Membership card, sir”. In the Italian dialogue, the English distal pronoun that is translated by the proximal questo: Questo è l’abbonamento a Blockbuster Video, signore ‘This is the Blockbuster Video Membership card, sir’. (24)

MALE RECEPTIONIST:

WILLIAM:

MALE RECEPTIONIST:

Are you an accredited member of the press? Lei è un membro accreditato della stampa? [Are you an accredited member of the press?] Yeah. There you go. ((showing a card)) Sì. Ecco qua. ((showing a card)) [Yes. Here you are.] Yes, that’s a Blockbuster Video Membership card, sir. Questo è l’abbonamento a Blockbuster Video, signore. [This is the Blockbuster Video Membership card, sir.] (Notting Hill)

The data presented here suggest that the mismatch between English that and Italian quello / quella affects the translation of both endophoric and exophoric reference. In addition, the translation patterns show a specificity of English: i.e. that appears to have moved away from the strictly spatial meaning of ‘far away from the speaker / close to hearer or far from both’. Its conversational use, as reflected in the film dialogues, shows the unmarked status of the distal pronoun in both its textual and spatial meanings. That has emerged as the unmarked demonstrative of

166 Translation of Conversation and Film Dubbing as a Discovery Procedure

face-to-face conversation, which may neutralize distinctions in terms of proximity to either one of the conversation partners or both. It is also clear that the pronoun is often charged with an interactional and interpersonal component, whereby when the addressee’s point of view is taken, the distal demonstrative is employed, although physical proximity in itself could justify the choice of the proximal pronoun. Space is thus construed as shared by both interactants. The use of that, however, may apply to both instances of converging or diverging emotional stances. In Italian, on the other hand, the same interactional functions appear to favour the demonstrative prototypically encoding physical proximity to the speaker, and questo prevails.11

4. Conclusions It is notoriously difficult to obtain information about the translation of conversation. This chapter has argued that, among the various genres and registers that may be taken to approximate conversation, audiovisual dialogue provides the closest match, thus allowing the researcher to gain a unique insight into the translation of spontaneous spoken language. English film language is sociolinguistically rich and overlaps considerably with face-to-face conversation along parameters of interactiveness and informality. As a result, it may be taken as an internalized schema offering a useful representation of the production of conversational language. More specifically, some selected features can be profitably investigated since they are entrusted with the function of expressing and evoking orality in film. Admittedly, owing to the specific functions performed by audiovisual dialogue, but also due to various production constraints, there are a number of shortcomings in relying on dubbing to study the cross-linguistic rendering of spoken language. Provided caution is paid when interpreting the findings, audiovisual translation can however offer a privileged access to the translation of spontaneous spoken language, access that is not easy to obtain otherwise. To exemplify how dubbing can be explored to search for translation patterns in orality, the second part of the paper has focused on the translation of demonstrative pronouns in a parallel corpus of British and American films and their Italian translations. The analysis of the 11

In spontaneous spoken Italian, the demonstrative could be accompanied by a locative qui or qua ‘here’ > questo qui, questo qua or even the reduced ‘sto qui/qua.

Maria Pavesi

167

translation strategies has shown that demonstratives are systematically reduced when the dialogue is transferred from the source to the target language. In particular, the considerable number of omissions and substitutions observed in the PCFD points to the relevant role of crosslinguistic contrast since such translation shifts mainly take place within areas where English and Italian differ both syntactically, in the expression of the subject and the position of the object, and pragmatically, with both endophora and exophora. This investigation may open up novel avenues of research as it has shown some advantages of resorting to audiovisual translation to uncover translation trends together with translation problems as applied to spoken language. These trends and cross-linguistic contrasts could be further explored in spontaneous spoken Italian. Are there patterns and regularities in Italian conversations that have escaped researchers’ attention and that may emerge more clearly when the translation of fictive orality is closely examined? The divergence between the two languages in terms of exophoric deixis is a case in point, worth investigating more in depth in both spoken English and spoken Italian, as well as in different translation productions.

Acknowledgement The research for this article was carried out within the international project English and Italian Audiovisual Language: Translation and Language Learning funded by the Fondazione Alma Mater Ticinensis, University of Pavia for the years 2010-2012.

References Alvarez-Pereyre, Michael. 2011. “Using Film as Linguistic Specimen.” In Telecinematic Discourse: Approaches to the Language of Films and Television Series, ed. Roberta Piazza, Monika Bednarek and Fabio Rossi, 47-67. Amsterdam: Benjamins. Baños, Rocío. 2013. “‘That’s so Cool’: Investigating the Translation of Adverbial Intensifiers in English-Spanish Dubbing Through a Parallel Corpus of Sitcoms.” Perspectives. Studies in Translatology 21 (4): 526-42. Bednarek, Monika. 2010. The Language of Fictional Television. Drama and Identity. London: Continuum.

168 Translation of Conversation and Film Dubbing as a Discovery Procedure

—. 2012. “‘Get us the Hell out of Here’: Key Words and Trigrams in Fictional Television Series.” International Journal of Corpus Linguistics 17 (1): 35-63. Berretta, Monica. 1985. “I pronomi clitici nell’italiano parlato.” In Gesprochenes Italienisch in Geschichte und Gegenwart, ed. Günter Holtus and Edgar Radke, 185-224. Tübingen: Narr. Biber, Douglas, Johansson Stig, Geoffrey Leech, Susan Conrad and Edward Finegan. 1999. Longman Grammar of Spoken and Written English. London: Longman. Bollettieri Bosinelli, Rosa Maria, Elena di Giovanni and Ira Torresi. 2005. “Visual and Verbal Aspects of Otherness: From Disney to Coppola.” In Identity, Community, Discourse. English in Intercultural Settings, ed. Giuseppina Cortese and Anna Duszak, 405-27. Bern: Peter Lang. Botley, Simon and Tony McEnery. 2001. “Demonstratives in English. A Corpus-based Study.” Journal of English Linguistics 29 (1): 7-33. Cardinaletti, Anna. 2009. “Si impersonale e dimostrativi: due casi di influenza dei dialetti sull’italiano?” In Italiano, italiani regionali e dialetti, ed. Anna Cardinaletti and Nicola Munaro, 29-54. Milano: Franco Angeli. Carter, Ronald and Michael McCarthy. 2006. Cambridge Grammar of English. A Comprehensive Guide. Spoken and Written English. Grammar and Usage. Cambridge: Cambridge University Press. Cary, Edmond. 1960. “La traduction total.” Babel 6: 110-5. Chaume, Frederic. 2004. Cine y traducción. Madrid: Cátedra. —. 2012. Audiovisual Translation: Dubbing. Manchester: St. Jerome. Coulmas, Florian. 1979. “On the Sociolinguistic Relevance of Routine Formulae.” Journal of Pragmatics 3 (4): 239-66. Da Milano, Federica. 2005. La deissi spaziale nelle lingue d’Europa. Milano: Franco Angeli. Diessel, Holger. 1999. Demonstratives. Forms, Function, and Grammaticalization. Amsterdam: Benjamins. —. 2006. “Demonstratives, Joint Attention, and the Emergence of Grammar.” Cognitive Linguistics 17 (4): 463-89. —. 2014. “Demonstratives, Frames of Reference, and Semantic Universals of Space.” Language and Linguistics Compass 8 (3): 116-32. Dixon, Robert M.W. 2003. “Demonstratives: A Cross-linguistic typology.” Studies in Language 27 (1): 61-122. Forchini, Pierfranca. 2012. Movie Language Revisited. Evidence from Multi-Dimensional Analysis and Corpora. Bern: Lang. Fowler, Roger. 2000. “Orality and the Theory of Mode in Advertisements.” In Changing Landscapes in Language and Language

Maria Pavesi

169

Pedagogy, ed. Marie-Noëlle Guillot and Marie-Madeleine Kenning, 26-39. London: AFLS/CiLT. Freddi, Maria. 2011. “A Phraseological Approach to Film Dialogue: Film Stylistics Revisited.” Yearbook of Phraseology 2: 137-63. Freddi, Maria and Maria Pavesi. 2009. “The Pavia Corpus of Film Dialogue: Methodology and Research Rationale.” In Analysing Audiovisual Dialogue. Linguistic and Translation Insights, eds. Maria Freddi and Maria Pavesi, 95-100. Bologna: CLUEB. Gaudino-Fallegger, Livia. 1992. I dimostrativi nell’italiano parlato. Wilhelmsfeld: Egert Verlag. Halliday, Michael A.K. and Raqaiya Hasan. 1976. Cohesion in English. London: Longman. Heyd, Theresa. 2010. “How You Guys Doin’? Staged Orality and Emerging Plural Address in the Television Series Friends.” American Speech 85 (1): 33-66. Kozloff, Sarah. 2000. Overhearing Film Dialogue. Berkeley: University of California Press. Küntay, Aylın and Asli Özyürek. 2006. “Learning to use demonstratives in conversation: what do language specific strategies in Turkish reveal?” Journal of Child Language 33: 303-20. Lakoff, Robin. 1974. Remarks on This and That.” In CLS. Papers from the Tenth Regional Meeting of the Chicago Linguistics Society, eds. Michael W. La Galy, Robert A. Fox and Antony Bruck, 345-56. Chicago: Chicago Linguistic Society. Lakoff Tolmach, Robin and Deborah Tannen. 1984. “Conversational Strategy and Metastrategy in a Pragmatic Theory: The example of Scenes from a Marriage.” Semiotica 49 (3/4): 323-46. Levinson, Stephen C. 1983. Pragmatics. Cambridge: Cambridge University Press. Miller, Jim, and Regina Weinert. 1998. Spontaneous Spoken Language. Oxford: Clarendon. Murphy, Paul R. 1978. “Sociolinguistics in the Movies: A Call for Research.” Anthropological Linguistics 20 (5): 226-33. Pavesi, Maria. 2005. La traduzione filmica. Aspetti del parlato doppiato dall’inglese all’italiano. Roma: Carocci. —. 2008. “Spoken Language in Film Dubbing: Target Language Norms, Interference and Translational Routines.” In Between Text and Image. Updating Research in Screen Translation, ed. Delia Chiaro, Christine Heiss and Chiara Bucaria, 79-99. Amsterdam: Benjamins. —. 2012. “The Potentials of Audiovisual Dialogue for Second Language Acquisition.” In Translation, Technology and Autonomy in Language

170 Translation of Conversation and Film Dubbing as a Discovery Procedure

Teaching and Learning, ed. Pilar Alderete-Díez, Laura Incalcaterra McLoughlin, Labhaoise Ní Dhonnchadha and Dorothy Ní Uigín, 15574. Oxford: Lang. —. 2013. “This and That in the Language of Film Dubbing: a Corpusbased Analysis.” Meta 58 (1): 107-37. Quaglio, Paulo. 2009. Television Dialogue: The Sitcom Friends vs. Natural Conversation. Amsterdam: Benjamins. Quaglio, Paulo and Douglas Biber. 2006. “The Grammar of Conversation.” In The Handbook of English Linguistics, ed. Bas Aarts and April McMahon, 692-723. Oxford: Blackwell. Ramat, Paolo. 2014. “Transcategorization in deictic strategies.” Paper presented at the 8th International Conference on Construction Grammar (ICCG-8). University of Osnabrück, September, 3-6, 2014. Remael, Aline. 2003. “Mainstream Narrative Film Dialogue and Subtitling. A Case Study of Mike Leigh’s Secrets & Lies.” The Translator 9 (2): 225-47. Remael, Aline. 2008. “Screenwriting, Scripted and Unscripted Language: What do Subtitlers Need to Know?” In The Didactics of Audiovisual Translation, ed. Jorge Díaz Cintas, 57- 67. Amsterdam: Benjamins. Rey, Jennifer M. 2001. “Changing Gender Roles in Popular Culture: Dialogue in Star Trek Episodes from 1966 to 1993.” In Variation in English: Multi-Dimensional Studies, ed. Susan Conrad and Douglas Biber, 138-155. London: Longman. Richet, Bertrand. 2001. “Quelques données et réflexions sur la traduction des interjections.” Oralité et traduction, ed. Michel Ballard, 79-128 . Arras: Presses Université. Rossi, Fabio. 1999. Le parole dello schermo. Analisi linguistica del parlato di sei silm dal 1948 al 1957. Roma: Bulzoni. —. 2011. “Discourse Analysis of Film Dialogue: Italian Comedy between Linguistic Realism and Pragmatic non-Realism.” In Telecinematic Discourse: Approaches to the Language of Films and Television Series, eds. Roberta Piazza, Monika Bednarek and Fabio Rossi, 21-46. Amsterdam: Benjamins. Rühlemann, Christopher. 2007. Conversation in Context: A Corpus-driven Approach. London: Continuum. Serianni, Luca 1991. Grammatica italiana. Italiano comune e lingua letteraria. Torino: UTET. Taylor, Christopher. 1999. “‘Look Who’s Talking’. An Analysis of Film dialogue as a Variety of Spoken Discourse.” In Massed Medias. Linguistic Tools for Interpreting Media Discourse, Linda Lombardo, Louann Haarman, John Morley and Christopher Taylor, 247-78.

Maria Pavesi

171

Milano: Led. Tomaszkiewicz, Teresa. 2001. “La structure des dialogues filmiques: conséquences pour le sous-titrage.” In Oralité et traduction, ed. Michel Ballard, 381-399. Arras: Artois Presses Université. —. 2009. “Linguistic and Semiotic Approaches to Audiovisual Translation.” In Analysing Audiovisual Dialogue. Linguistic and Translation Insights, ed. Maria Freddi and Maria Pavesi, 19-29. Bologna: CLUEB.

INTERPRETARI, TRANSFERRE, TRA(NS)DUCERE, (CON)VERTERE, IMITARI: DIFFERENT VIEWPOINTS ON TRANSLATING PAOLO RAMAT IUSS PAVIA AND UNIVERSITY OF PAVIA

The first part of the article deals with some general problems of the translating operation. Key points of this part are: translation as a universal possibility of language(s); language as a ‘problem solving system’ and the ‘tertium comparationis’ (noeme); different expressions for ‘translation / to translate’ and different metaphors from Cicero to Manzoni; possible semantic map(s). The second part sums up the reflections done in the first part and tackles the following topics: contrastive grammar(s) and comparison between ‘same’ texts; the pragmatic difference between denotation (‘Bezeichnung’) and connotation (‘Bedeutung’); contrast between ‘semantic translation’ (“brutta fedele”) and ‘communicative translation’ (“bella infedele”).

1. Introduction Let us start this conclusive paper with some general reflections on translation and the operation of translating. As a matter of fact, we see that all kinds of texts are always translated in all languages. The most evident example—but certainly not the only one—is the Bible. We may have divergent opinions about the ‘quality’ of the translation but the denotative contents of the text have been (and are) transmitted from language to language. I will come back to this very basic point, but it is undeniable that there will always be a common content basis which among other things enables the translation from language A to language B. As Lazard (2012: 7) puts it: “L’intuition suggère en effet que toutes les langues [...] se ressemblent dans une certaine mesure: ce n’est pas par hasard que des langues exotiques ont été jadis décrites dans les termes de la grammaire latine” (cp. also Ramat 2014: 856). In fact, it is well-known that missionaries had much trouble in describing indigenous

174

Different Viewpoints on Translating

languages according the schemes of the Latin grammar; nevertheless they succeeded in writing useful and often very valuable descriptions. On the contrary, it is impossible to ‘translate’ Beethoven’s ‘pastoral symphony’ or the Michelangiolo’s ‘Pietà’. Music and sculpture make use of different semiotic codes, different communicative forms, different ‘languages’ (whereby the term ‘language’ is metaphorically used). As for Sign Languages, which have to be considered languages under all points of view, see below fn. 5. The fact that referential contents may pass from language to language is due to the fact that all languages are basically homologous semiotic systems which function in the same way—this means that they share the same properties in order to solve the same problem, namely to express the contents they are faced with: languages are ‘problems solving systems’. They make use of semantic-functional constructions expressing for instance ‘possession’, ‘location’, ‘apprehension’, ‘participation’, etc. (see Hansjakob Seiler’s ‘dimensions’: Seiler 1983, 1989, 1995, etc.).

2. Translation and typology We know that languages may use very different strategies for expressing the same function -though not in an unlimited number. But the identity of the function (say, the notion/‘dimension’ of ‘possession’) guarantees the translatability from language to language. This is the crosslinguistic 'tertium comparationis', i.e. what Klaus Heger (1990-91) called 'the noeme', the mental content, defined by Bernard Pottier as “figure de sens élémentaire, abstraite et universelle”(Pottier 2000: 14). I repeat: this is why it is basically possible to translate from language to language. Under this point of view the title of our workshop Translation: Language across Languages echoes the famous sentence of the ‘doctor mirabilis’ Roger Bacon (ca. 1214-1294): “gramatica una et eadem est secundum substantiam in omnibus linguis, licet accidentaliter varietur” (my emphasis; “according its substantial character, the ‘gramatica’ is one and the same for all languages, though it may vary by accident”): under the multifaceted variety of the human languages there exists the language as property of the human species. Note, anyway, that by the term ‘gramatica’ Bacon, following the Medieval tradition, meant actually ‘logica’—i.e. the ‘substance’, what matters—, applied to and implemented by the language. In conclusion isofunctionality (in the sense that all languages have to solve the problem of expressing in linguistic terms the ideas elaborated in the mind) and ‘tertium comparationis’ are the prerequisites for translating although they do not represent the only preliminary condition for a “good”

Paolo Ramat

175

translation. On the discussion of “good” or “bad translation” see further below. Note also that when speaking of translation we usually refer to texts, not to isolated items to be found in dictionaries: Benvenuto Terracini wrote in his fundamental essay on translation (1951, repr. 1957: 67) “Nel mettere a fronte non parole isolate, ma frammenti di discorso vivo molte discordanze che paiono insuperabili si risolvono” (“If we confront not isolated words but fragments of real discourse, many apparently unavoidable discordances will disappear”). In more recent times and in strictly linguistic terms that do not explicitly deal with the problem of translation this reminds us of the perspective adopted by George Lakoff, Charles Fillmore, Paul Kay and lastly by William Croft in the so-called 'Construction Grammar' which argues that the meaning of the whole is not the simple composition of the meaning of the parts (see Croft 2001).1 Constructions themselves have meanings in a ‘syntax-lexicon continuum’. This entails that the exact meaning of a construct can be defined only when looking at the whole of the construction. In the same vein it has been maintained by many linguists as well as critics of literature that a good translator must consider the whole source text before starting to translate. He has to understand all the denotations and connotations of the original (see below). Accordingly, Construction Grammar represents a step forward when compared to contrastive linguistics that was often limited to comparison of lexical structures and to Contrastive Grammar, inasmuch it entails, more or less implicitly, cross-textual comparisons.2 The necessity of considering entire texts is particularly evident in many situations where clarity is utmost necessary. The famous UNO resolution 242 speaks in the English version of the withdrawal of Israel armed forces from territories occupied in the recent conflict (i.e. the 1967 war of Israel against Egypt, Siria, and Jordany) whilst the French text is obliged to use an article before territoires and to make a choice between des territoires (plural) or de territoires (singular). The first—and officially adopted— option entails that Israel must withdraw from all the occupied territories, the second one means that Israel has to withdraw from some of the occupied territories (see Hagège 1987: 31f.). When speaking of translation we have in the present context to remember what Jakobson wrote in dealing with the ‘Übersetzungsfrage’: 1

According to Clausner and Croft 1997 the concept of ‘construct(ion) is allembracing in language. As Fedriani (2014: 7) writes: “[T]he notion of construction can be applied to virtually every unit of language, from simple words to more complex and abstract structures, like verbal argument structures, sentence patterns and even metaphors”. On metaphors see the discussion below. 2 See in this volume Stephany’s chapter, which goes much in this sense.

176

Different Viewpoints on Translating

languages differ not on what they can say but on what they must say (Jakobson 1959): faced with the French form je chante, (Ital. canto) a Spaniard has to decide whether to translate as estoy cantando (progressive form, I'm singing) or canto (simple present, I sing). On the other hand, if a French speaker really wants to underline the progressive aspect of the verbal action, she or he has at her/his disposal a periphrastic construct and can (but (s)he’s not obliged to) say je suis en train de chanter. The answer to the question ‘What are you doing ?’ can be Je chante (Ital. Canto) as well as Je suis en train de chanter (Ital. Sto cantando): see Malmberg 1973: 8f. The typological relevance of such a wider approach is evident. William Croft has posted (22.07.2013) an advertising for Multidimensional Scaling Programs (MDS at http://www.unm.edu/-wcroft/MDS.html) used in typology for cross-linguistic comparison of language-specific category data. This is not the place to discuss the (non-)universality of linguistic categories, which has been debated in many publications (see Ramat 2009), but I think that MDS can be applied not only to the analysis of language-specific categories, but also to more general and traditional categories, largely diffused cross-linguistically. As typologists know very well, the cross-linguistic differences may be extremely large: the examples quoted in this volume by Muro from an Upper North Wakashan language of the British Columbia and two other ‘exotic’ languages show how many different values such as Focalizer, Nominalizer, Inclusiveness etc., which do not have an English correspondent morphological expression, can or must be morphologically or syntactically expressed. Obviously, this makes the translation much more difficult than a translation between languages that belong to the same linguistic type, such as Hebrew and Arabic, or a well-established ‘Sprachbund’ like the so-called Standard Average European. As Sapir (1921) wrote in the typological chapter 6 of his book Language, “when we pass from Latin to Russian, we feel that it is approximately the same horizon that bounds our view, even though the near, familiar landmarks have changed.” Anisomorphism, says Muro, represents one of the main difficulties which translators must face. It is this anisomorphism (both lexical and morphosyntactic) that led Sapir and Whorf to advance the famous hypothesis that speakers of different languages might think in different ways.3 Though evidently relevant for the translation discussion, I will not deal here in depth with the ‘Sapir-Whorf hypothesis’ which is far 3 On anisomorphism applied to the distribution of metaphorical meanings, see also Prandi, this volume.

Paolo Ramat

177

from being settled and would require a special ‘ad hoc’ chapter. As Cardona (1976: 107) noted, the discussion has suffered from the fact that the hypothesis and its discussion preceded any linguistically, nonphilosophically carried out research upon language universals and deep semantic-functional structures (for further discussion see Ramat 2007: 25).

3. Translation and meaning Let’s now consider another relevant point tied to translation. We have to distinguish between denotation and connotation, in Coseriu’s (1978: 185) terms between ‘Bezeichnung’, i.e. the factual meaning of a linguistic expression, and its ‘Bedeutung’, i.e. its interpretation, its reference beyond the speech act (‘die einzelsprachliche Referenz’). To understand the ‘Bedeutung’ is of course the most difficult task, since, as we have already said, the lexical and morphosyntactic constraints of the target language play a very important role in shaping the ‘Bezeichnung’. We said above that all languages are able to express all possible denotations (‘Bezeichnungen’). But each language has its own rules, its own constraints—just as Jakobson showed. The distinction between factual meaning (denotation, ‘Bezeichnung’) and the connotations (‘Bedeutungen’) has led Peter Newmark (1984: 39f.) to make a distinction between ‘semantic translation’ and ‘communicative translation’. The former strives to reproduce “the exact contextual meaning of the original” (as if we would translate chien méchant by cane cattivo instead of attenti al cane). Anticipating a discussion which would have animated the 20th century, the Humanist Juan Luís Vives (1492-1540) had proposed three kinds of translation: (a) translations where only the ‘sensus’, what is said, is relevant (‘solus spectatur sensus’); (b) translations where only the way of the expression matters—i.e. how the meaning is linguistically realized (sola [spectatur] phrasis et dictio); and finally translations where (a) and (b) are united (ubi et res et verba ponderantur ‘where the things and the words are equally considered’ (De ratione dicendi, 3rd vol. chapt. 12: 233; quoted from Coseriu 1971a: 246; 1971b: 573). Vives clearly says that the three translation kinds apply to different text genres. Thus we come back to what I said before: there are texts where only the ‘sensus’ matters: ‘No smoking’, ‘Do not lean out’, ‘The sum of the angles of a triangle amounts to 180 degrees’. Other texts care more for the form than the substance—or, better, the form is here the substance, as in the case of Verlaine’s example below, and finally there are texts where res et verba ponderantur.4 4

Compare in this vol. the idea of a ‘third space’ proposed by Debora Biancheri.

178

Different Viewpoints on Translating

Moreover, it has to be noted that the same conceptual contents may be expressed even in languages which are typologically and culturally very similar by completely different linguistic structures: “... los mismos contenidos conceptuales pueden ser expresados por lenguas tipológica y culturalmente muy próximas mediante estructuras completamente divergentes”, Moreno 1995: 146 (my italics, P.R.; “the same conceptual contents may be expressed in languages that are very close from the typological and cultural point of view via completely diverging structures”): caue canem, beware of the dog, chien méchant, attenti al cane, ¡cuidado con el perro! are all warning signs which refer to the same state of affairs and do not show particular connotations. But they make use of very different linguistic means: imperative in Latin and English, adjective in French, substantive in Spanish and non-verbal sentence in Italian. Different means are used in Sign Languages which can be defined at a metalinguistic level as an “intersemiotic translation”.5 Note that Sign Languages usually represent the transposition of oral languages—although in principle they are an autonomous system of communication. But, though correctly considered an ‘intersemiotic translation’, they can’t ‘translate’ Beethoven’s symphonies nor Michelangelo’s statues—as well as oral languages cannot. Finally, we have to tackle the thorny problem of translating metaphors discussed by Michele Prandi in this volume: he correctly states that the common denominator of the metaphor is the transfer of a concept from a conceptual area to another one: when Latin writers used the terms transferre or tra(ns)ducere they were using consolidated metaphors, whose original meaning 'to bring over' was probably no longer present to their linguistic conscience. These 'dead metaphors' are called catachréseis: they refer to a metaphorical or metonymic meaning extension that produced a new sense of an old word (see Fontanier 1968: 213). When we use in Italian tramonto (< trans monte(m) 'beyond the mountain') for sunset and coucher du soleil we refer to a landscape where mountains are almost the norm, so to say they represent the unmarked situation, as its English or French equivalents which are also catachréseis whose original metaphoric sense went lost. The same holds for the famous verse of Sappho (frg. 94) dédyke mèn a salánna [...]. Égǀ dè móna kateúdǀ “the moonset has arrived […] and I'm sleeping alone'. The Italian translation by the Nobel Laureate Salvatore Quasimodo has ‘tramontata è la luna...’, but 5

See Giulia Petitta’s paper, which continues the important research tradition of the Rome CNR ‘Institute of Cognitive Sciences and Technologies’.

Paolo Ramat

179

actually dédyke means 'it dove/has dived' (scil. into the sea); this is quite understandable if we refer to an island like Lesbos where the sun disappears into the water, but it would sound rather odd to an Italian ear: ?si è tuffata la luna... The frozen metaphor of dédyke, if literally translated by 'si è tuffata', would become what Prandi (2010) calls a 'conflictual metaphor', a new poetic creation containing a vivid conceptual conflict since only leaving beings can dive, as well only living beings can coucher, ‘go to bed’. Nevertheless the common denominator -i.e. the 'tertium comparationis' which makes the translation possible- is that the moon has disappeared, has finished its ‘parcours’ in the sky. Now, frozen metaphors that belong to the common language and are no longer perceived as metaphors are very often the result of individual inventions: there is a continuous feeding process from the latter to the former and the two domains are not separate domains. Other examples of dead metaphors are offered by idioms and proverbs. We cannot translate They are as like as two peas by *Si assomigliano come due piselli, or *Ils se ressemblent comme deux petits pois since the same concept is expressed via a different comparison: Si assomigliano come due gocce d’acqua; Ils se ressemblent comme deux gouttes d’eau (see Montella 1989; Ramat 2007) Indeed, the aim of the translation is to reformulate the meaning of an expression into a different language. Accordingly, one does not translate the metaphor, but its semantic support, as illustrated in the case of sunset. It seems to me that such expressions (‘expressions figées’ as well as ‘dead metaphors’) can be easier translated than a complex text as a poem or a novel where the ‘Bedeutungen’ (connotations) play a major role. No matter whether the prohibition of smoking is expressed by an infinitive as in Vietato fumare and Rauchen verboten, or by a substantive as in No smoking. Greek has apagoreúetai to kapnízein, lit. 'is forbidden the to smoke', where the infinitive is substantivized by the article to and becomes the grammatical subject of a passive sentence. In these sentences, as in the previous example of beware of the dog, there are no connotations at all and the message has just a denotative value which, in spite of the different constructions, can be easily transmitted from language to language. On the contrary, translations of texts endowed with rich connotations may be difficult, even at the level of phonetics: Claude Hagège (1985: 48), was right in observing that the almost onomatopoeic effect of the nasal vowels in the five verses by Paul Verlaine les sanglots longs/ des violons/ de l'automne / blessent mon coeur/ d'une langueur/ monotone (Chanson d’automne) cannot be realized in Japanese since this language doesn't have nasal vowels. Even the English translation is capable to repeat just

180

Different Viewpoints on Translating

partially the ‘effet de sens’ of the original: 'When a sighing begins / In the violins / Of the autumn-song,/ My heart is drowned/ In the slow sound/ Languorous and long’ (transl. by the poet Arthur Symons (1865-1945)).

4. ‘The men who turn the wor(l)ds’: different viewponts on translating From the historical point of view, as far as we can go back with our documents, we find that the notion and the practice of translating are well documented—which is an unavoidable situation if one thinks of the relations and clashes of different cultures. Already at the end of the 4th millennium we find in Sumerian texts the word eme.bal ‘language (eme) (that) turns (bal)’, and inim.bal applies to the translator (‘[the man who] turns the words’, “che è un bell’antefatto cognitivo di lat. vertere e versio” (“a nice cognitive antecedent of Lat. vertere”), as Silvestri (2000: 84 [repr. 2011: 177]; see also 2009: 148 [repr. 2011: 217]) comments (see below also It. voltare). Silvestri quotes also It. dragomanno or turcimanno—an ‘interpreter’ who was in charge at the European courts for the Near Eastern embassies. Dragomanno and turcimanno go back via Aramaic tϷrgϷmƗnƗ ‘translator’ (cp. post-Biblical Hebrew tirgem‘ ‘to translate’, denominative verb, cp. also Arab. tar÷umƗn) and two paraetymologies (‘Volksetymologien’: drago and turco, respectively) to the Akkadian ta/rgumƗnu attested in the second half of the third millennium. (Cp. also Ge‘ez (Classical Ethiopic): targwama ‘to interpret, translate, expond, comment on’). From the Osman Turkish form tilmaç of the same word derive Rom. tălmaci, Hung. tolmács and Germ. Dolmetscher.6 Tullio De Mauro (1994: 86) wrote as follows: “[N]on è sbagliato affermare che l’attività traduttiva è coestensiva al linguaggio: antica e vasta, nella storia della specie e delle culture umane, quanto antico e vasto è il dominio del linguaggio umano” (“[I]t is not wrong to say that translating is inborn in the very nature of human language: it is as ancient and wide, within the history of the human species and cultures, as human language itself”). In fact, it would be very difficult at the theoretical, metalinguistic level to maintain the impossibility of linguistic translation for each and every language. What should be the prerequisites specific of language A or B that would guarantee their ability to translate, whereas language C would not possess these prerequisites?

6 See Leslau (1987: 579-580). Thanks to Prof. Martin R. Zammit, Dept. of Oriental Languages, University of Malta, for this bibliographical reference.

Paolo Ramat

181

At this point the relevant question is, however: how is the operation of translating conceived of? That translating is an activity present everywhere does not imply that it is seen everywhere with same eyes. I have quoted in my paper several synonyms referring to this operation and many contributions to the present volume underline that there are not only different strategies for translating but also different translation goals. Emanuele Banfi has hinted at the Chinese way of referring to the translation : fƗn yì as ‘to add wings to what is stranger’, a way of nobilitating what the ȕȐȡȕĮȡȠȚ say. Banfi adds also that Japanese has borrowed the Chinese expression: a good example of the Chinese cultural influence on the Raising Sun Empire. I have already mentioned the Sumerian inim-bal “word-turn(s)” for the translator’s activity. In other linguistic traditions we find more or less the same image of the operation of translating: Turkish uses the verb çevir-mek from *çev- ‘to turn, rotate” and the substantive çev-ir-me “rotation, change” and also “translation” (thanks to Emanuele Banfi for this information). As in the case of inim-bal, a semantic comparison with Lat. (con-)uertere comes immediately to the mind. Latin interpres –etis is an ਦȡȝȘȞİȪȢ, the person who explains (no clear etymology for this word, which has been kept in Modern Greek too, cp. the verb (įȚ)İȡȝȘȞİȪȦ).7 Originally the Latin word meant ‘inter-mediary, middleman’ but the etymology of -pres is uncertain (if connected with pretium ‘price’ the interpres would be ‘(the person who) establish(es) the medium price, the price in between’). Only later the substantive and the derived verb were metaphorically applied to the linguistic dimension. Russian uses perevodit’ ‘translate’ and perevodþik ‘translator’. The adjective perevodnòƱ means ‘closely followed’ and perevodnaja literatura is the translated literature. Here the idea is similar to what we find in Lat. imitari as a calque operation. Note that the very notion of ‘translation’ is linguistically expressed via metaphors in all the languages I know of. This is quite understandable if one thinks that ‘translation’ is a very abstract notion and abstract notions are usually drawn via metaphors from concrete terms and notions. Some of the Latin verbs I have used in the title refer to the spatial dimension, i.e. to something that can be sensorially perceived: tra(ns)ducere, (con)uertere, may be also inter-pretari, and trans-ferre (actually from its more regularized form trans-latare) with its ‘nomen actionis’ trans-latio. The same holds for über-setzen. Gianfranco Folena in a seminal article on ‘volgarizzare’ and 'tradurre' (Folena 1973) did an accurate analysis of the 7

See Banfi (this vol.) on the use of ਦȡȝȘȞİȪȢ in Herodotus’ narration.

182

Different Viewpoints on Translating

many terms used in Latin, in the Middle Age, and in the Renaissance. Along with the already mentioned interpretari and transferre the last one being a calque of the generic Greek verb meta-phérein, he mentions (con)uerto and trans-uerto used for poetic and literary translation. On the other hand, imitari 'imitate' does not belong to the spatial dimension. However, it is not the usual verb for ‘translate’ and it is used mainly for the literary translation; in fact, it correlates with imago ‘image, representation, comparison’ and evidently follows another semantic path. As noted by Folena (1973: 62), Latin had many different terms for the concept, since it was confronted with the problem of translatare into the poor native language (the famous ‘patrii sermonis egestas’) an impressive literary tradition that until later times did not care very much for the problem of translation and consequently had a rather poor and limited terminology (meta-phérein ‘transfer, transport’, meta-bibázein ‘take away, change’, meta-phrázein ‘paraphrase, explain’ and also meta-gráphein ‘transcribe’ and hence ‘translate’). Plautus says in the prologue of his Trinummus (v. 19) Philemo scripsit, Plautus uortit barbare. Thus, we find in Seneca not only conuerto and imitari but also mutare ‘mutate, change’ (Dante still uses tra(n)smutare) and explicare ‘explain’. These verbs allude in fact to an already sophisticated metalinguistic reflection on the translating operation which will find its highest points in Cicero, Quintilianus and Hieronymus—and, if we look outside the Latin and Late Latin tradition, in Martin Luther’s famous Sendbrief vom Dolmetschen (1515). Translating is no longer the simple (though by no means easy!) operation of the Dolmetscher, the interpreter. It entails different layers according to different goals and different audiences. Much in the spirit of Cicero and Quintilianus, Hieronymus (who died in Rome, 406), wrote that it is necessary “non verbum e verbo sed sensum exprimere de sensu” (“not to translate word for word but express the meaning from the meaning”, Epistula lvii ad Pammachium). St. Bonifacius (ca 680-754) said to a young boy who later became the bishop of Utrecht “Bene legis, fili, si intelligis quae legis” (“You read well, my son, if you understand what you read”) and he added: “Non ita, fili, quaero, uti mihi dicas modo lectionem tuam, sed secundum proprietatem linguae tuae et naturalem parentum tuorum locutionem edissere lectionem tuam” (“I don't want, my son, that you tell me your ‘lectio’—way of reading—, but that you explain your ‘lectio’ according the property of your own language and the way your parents naturally speak”, MGH, XV, Pars I; see Delbono 1967: 32f.; my emphasis). In the Middle Age the translation is often seen as an Umarbeitung rather than a Übersetzung . Suffice to think of the German poems Heliand and the Evangelienbuch of Otfrid von Weißenburg. The

Paolo Ramat

183

exegetic aspect of the translation is reflected in the imperative edissere and in the Italian (e)sporre ‘illustrate, explain’, frequently used in Italian 'translations/adaptations' of the 13th and 14th centuries (see Folena 1973: 78f.). The humanist Leonardo Bruni who translated many Greek texts into ‘vulgar’ (i.e. Italian) used the verb tradurre (from Lat. trans-ducere) that became the standard word for all the Romance languages: traduire, traducir, traduzir, traduir, a traduce (see Folena 1973: 102). Originally, the term wasn't limited to 'translate' and could mean what its etymology says, i.e. ‘transport (objects)’ -just as über-setzen which makes use of the same metaphor. In the Renaissance time it became the technical word for this operation, whereby in the Romance area trans-latare and trans-ferre became more and more obsolete. Recently, Giuseppe Polimeni has drawn attention to another metaphor for ‘translating’ (Polimeni 2012): still in the 19th century Alessandro Manzoni in The Bethroted (I promessi sposi) uses the expression voltare in volgare. Renzo says to Don Abbondio (Chap. 38) ‘me lo volti un po’ in volgare, ora (scil. your latinorum)’ “Now please turn it into ‘vulgar’”, where the metaphor of ‘voltare’, ‘to turn’, is semantically strictly bound to that of uertere we have already seen and volgare means ‘(understandable) Italian’—according to a tradition which goes back to the Middle Age (Folena 1973: 79).

5. Some words of conclusion It would be interesting to make a world-wide comparison of the linguistic expressions used in very different civilizations. Along with the world-wide Atlas of Linguistic Structures (WALS), one could dream of something like íon a world-wide scaleí the glorious Dictionary of Selected Synonyms in the Principal Indo-European Languages published 1949 by Carl Darling Buck with the telling subtitle A Contribution to the History of Ideas. From the ethnolinguistic viewpoint the cross-linguistic differences would elucidate how the translating operation -and finally the linguistic activity itself- may be considered in the cognitive strategies of the human mind (see Cardona 1976: 65ff.). We could arrive at a very comprehensive semantic map or a few semantic maps where the notion of translating is connected to some basic knots. The semantic networks of the notion will probably show that the maps it belongs to are not exactly isomorphic, though related to a not unlimited number of metaphors. But I don’t have the means for such a huge enterprise. It is to hope that the modern storage systems, the huge data banks and the typological researches on a world-wide scale will make the enterprise possible.

184

Different Viewpoints on Translating

In conclusion, I hope to have shown how multifaceted is the problem of translating or—better—how many problems are involved at different levels in the translating operations. Operations that do not concern the bare linguistic aspect but extend to pragmatics, sociology, sociology of science, politics etc. In short we can conclude, coming back to De Mauro’s quote: translating from language to language is inborn in the very nature of humans and is an essential component of their cognitive activity.

References Cardona, Giorgio Raimondo. 1976. Introduzione all'etnolinguistica. Bologna: Il Mulino. Clausner, Timothy C. and William A. Croft. 1997. “Productivity and schematicity in metaphors.” Cognitive Science 21: 247-82. Coseriu, Eugenio. 1971a. “Zur Sprachtheorie von Juan Luis Vives.” In Festschrift zum 65. Geburtstag W. Mönch, ed. Werner Dierlamm and Wolfgang Drost, 234-55. Heidelberg: Kerle Verlag. —. 1971b. “Das Problem des Übersetzens bei Juan Luis Vives.” In Interlinguistica. Sprachvergleich und Übersetzung. Festschr. Zum 60. Geburtstag von Mario Wandruszka, ed. Karl-Richard Bausch and Hans Martin Gauger, 571-82. Tübingen: Niemeyer. —. 1978. “Falsche und richtige Fragestellungen in der Übersetzungstheorie.” In Theory and Practice of Translation, ed. Lillebill Grähs, Gustav Korlén and Bertil Malmberg, 17-32. Bern: Peter Lang. —. 1981. “Kontrastive Linguistik und Übersetzung: ihr Verhältnis zueinander.” In Kontrastive Linguistik und Übersetzungswissenschaft. Akten des Intern. Kolloquiums Trier / Saarbrücken 1978, ed. Wolfgang Kühlwein, Gisela Thome, Wolfram Wilss, 183-99. München: Fink. Croft, William A. 2001. Radical Construction Grammar: Syntactic Theory in Typological Perspective. Oxford: Oxford University Press. Delbono, Francesco. 1967. “La letteratura catechetica di lingua tedesca.” Settimane di studio del Centro Italiano di studi sull'alto medioevo, 14: 3-47. De Mauro, Tullio. 1994. “Sette forme di adeguatezza della traduzione.” In Tullio De Mauro, Capire le parole, 81-95. Roma-Bari: Laterza. Folena, Gianfranco. 1973. “‘Volgarizzare’ e ‘tradurre’: idea e terminologia della traduzione dal Medioevo italiano e romanzo all’Umanesimo europeo.” In Gianfranco Folena, La traduzione. Saggi e studi, 59-120. Trieste: Università Degli Studi di Trieste, LINT. Fedriani, Chiara. 2014. Experiential Constructions in Latin. Leiden: Brill.

Paolo Ramat

185

Fontanier, Pierre. 1968. Les figures du discours. Paris : Flammarion. Hagège, Claude. 1985. L'homme de paroles. Contribution linguistique aux sciences humaines. Paris: Fayard. —. 1987. “La traduction, le linguiste et la rencontre des cultures.” Diogène 137: 24-34. Heger, Klaus. 1990-91. “Noeme as tertia comparationis in language comparison.” ALFA, Halifax (CND), 3-4: 37-61. Jakobson, Roman. 1959. “On linguistic aspects of translation.” In On translation, ed. Reuben A. Brower, 232-9. Harvard: Harvard University Press. Lazard, Gilbert. 2012. Études de linguistique générale, II: La linguistique pure. Leuven-Paris: Peeters. Leslau, Wolf. 1987. Comparative Dictionary of Ge‘ez (Classical Ethiopic) Ge‘ez-English / English- Ge‘ez, with an index of the Semitic roots, Wiesbaden: Otto Harrassowitz. Malmberg, Bertil. 1974. “Observations théoriques sur la traduction.” In Bertil Malmbeg, La traduzione. Saggi e studi, 3-21. Trieste: Università degli Studi di Trieste-LINT. Montella, Clara. 1989. “Traduttore traditore. Alcuni aspetti teorici della traduzione del proverbio.” In La pratica e la grammatica. Viaggio nella linguistica del proverbio, ed. Cristina Vallini, 117-37. Napoli: Istituto Universitario Orientale. Moreno Cabrera, Juan Carlos. 1955. La lingüística teórico-tipológica. Madrid: Gredos. Newmark, Peter. 1984. Approaches to Translation. Oxford: Pergamon. Polimeni, Giuseppe. 2012. “‘Me lo volti un po’ in volgare ora’: insegnare e tradurre le rivoluzioni scientifiche.” In Fuori l’italiano dall’università? Inglese, internazionalizzazione, politica linguistica, ed. Nicoletta Maraschio e Domenico De Martino, 250-3. Roma-Bari: Accademia della Crusca/Laterza. Prandi, Michele. 2010. “Typology of Metaphors: Implications for Translation.” Mutatis mutandis 3: 304-32. Ramat, Paolo. 2007. “The problem of (un)translatability.” Rivista Italiana di Linguistica Applicata 39: 21-32. Ramat, Paolo. 2009. “How universal are linguistic categories?” In Universals of Language Today, ed. Sergio Scalise, Elisabetta Magni and Antonietta Bisetto, 1-11. Heidelberg: Springer, Heidelberg. —. 2014. “Review of Gilbert Lazard, Études de linguistique générale, II: La linguistique pure (2012).” Linguistics 52: 852-63 Sapir, Edward. 1921. Language. An Introduction to the study of speech. New York: Harcourt, Brace.

186

Different Viewpoints on Translating

Seiler, Hansjakob. 1983. Possession as an operational dimension of language. Tübingen: Narr. —. 1989. “The dimension of participation.” Funciòn 7. Universidad de Guadalajara, México. —. 1995. “Cognitive-conceptual structure and linguistic encoding: Language Universals and Typology in the UNITYP framework.” In Approaches to language typology, ed. Masayoshi Shibatani and Theodora Bynon, 273-325. Oxford: Clarendon. Silvestri, Domenico. 2000. “Identità, varietà e alterità linguistiche nel mondo antico.” In Linguistica storica e sociolinguistica, ed. Palmira Cipriano et al., 79-111. Roma: Il Calamo. —. 2009. “Nomi di popoli e nomi di lingue nel mondo antico. Convergenze, divergenze e illusioni onomastiche.” In Lingue, ethnos e popolazioni: evidenze linguistiche, biologiche e culturali. Atti del XXXII Convegno SIG, ed. Paola Cotticelli Kurras e Giorgio Graffi, 139-56. Roma: Il Calamo. Terracini, Benvenuto. 1957. “Il problema della traduzione.” In Benvenuto Terracini, Conflitti di lingua e di cultura, 49-121. Venezia: Nistri Lischi. Visconti, Jacqueline. 2013. “European integration: connectives in EU legislation.”. International Journal of Applied Linguistics 23:44-59.

CONTRIBUTORS

Emanuele Banfi Retired University Professor University of Milano-Bicocca [email protected]

Giulia Petitta Post-Doc Fellow Istituto di Scienze e Tecnologie della Cognizione -CNR [email protected]

Debora Biancheri Irish National Council Post-Doc Researcher National University of Ireland, Galway [email protected]

Michele Prandi Director of the Department of Language and Modern Culture University of Genoa [email protected]

Emanuele Miola Post-Doc Fellow University of Milano-Bicocca [email protected]

Paolo Ramat Retired University Professor University of Pavia [email protected]

Alessio Muro Research Fellow University of Padua [email protected]

Ursula Stephany Retired University Professor University of Cologne [email protected]

Maria Pavesi Full Professor University of Pavia [email protected]