2009 - Audiovisual Translation: Language Transfer on Screen

Audiovisual Translation Language Transfer on Screen Edited by Jorge Díaz Cintas and Gunilla Anderman Audiovisual Translation This page intentionally left blank Audiovisual Translation Language Transfer on Screen Edited by Jorge Díaz Cintas Imperial College London and Gunilla Anderman © Jorge Díaz Cintas and Gunilla Anderman 2009 Chapters © their individual authors 2009 All rights reserved. No reproduction, copy or transmission of this publication may be made without written permission. No portion of this publication may be reproduced, copied or transmitted save with written permission or in accordance with the provisions of the Copyright, Designs and Patents Act 1988, or under the terms of any licence permitting limited copying issued by the Copyright Licensing Agency, Saffron House, 6-10 Kirby Street, London EC1N 8TS. Any person who does any unauthorized act in relation to this publication may be liable to criminal prosecution and civil claims for damages. The authors have asserted their rights to be identified as the authors of this work in accordance with the Copyright, Designs and Patents Act 1988. First published 2009 by PALGRAVE MACMILLAN Palgrave Macmillan in the UK is an imprint of Macmillan Publishers Limited, registered in England, company number 785998, of Houndmills, Basingstoke, Hampshire RG21 6XS. Palgrave Macmillan in the US is a division of St Martin's Press LLC, 175 Fifth Avenue, New York, NY 10010. Palgrave Macmillan is the global academic imprint of the above companies and has companies and representatives throughout the world. Palgrave® and Macmillan® are registered trademarks in the United States, the United Kingdom, Europe and other countries. ISBN-13: 978–0–230–01996–6 hardback ISBN-10: 0–230–01996–X hardback This book is printed on paper suitable for recycling and made from fully managed and sustained forest sources. Logging, pulping and manufacturing processes are expected to conform to the environmental regulations of the country of origin. A catalogue record for this book is available from the British Library. Library of Congress Cataloging-in-Publication Data Audiovisual translation : language transfer on screen / edited by Jorge Díaz Cintas and Gunilla Anderman. p. cm. Includes bibliographical references and index. ISBN 978–0–230–01996–6 (alk. paper) 1. Dubbing of motion pictures. 2. Motion pictures – Titling. I. Díaz Cintas, Jorge II. Anderman, Gunilla M. TR886.7.A9285 2009 778.535—dc22 10 9 8 7 6 5 4 3 2 1 18 17 16 15 14 13 12 11 10 09 Printed and bound in Great Britain by CPI Antony Rowe, Chippenham and Eastbourne 2008029937 Gunilla Anderman In Memoriam It was with great personal sadness that we learnt of the unexpected death of Professor Gunilla Anderman, on 21 April 2007. Gunilla had been ill for some time but chose to keep her illness very private and continued doing what she loved; writing about and teaching Translation Studies. Gunilla was co-founder of the University of Surrey’s Centre for Translation Studies (CTS) and remained its Director for over 20 years. She nurtured the CTS from very small beginnings to create the internationally respected Centre that we know today. As one-time Chair of the ITI Education & Training Committee, she was also very keen on actively fostering links between the profession and academia. Gunilla herself was a distinguished translator of drama between Swedish and English, as well as an inspirational teacher and scholar of international repute. We had been working together on this volume for some time and know how pleased Gunilla was to see the manuscript submitted to the publishers. This volume is dedicated to her memory. Gunilla was a gifted communicator, full of natural charisma, with a wonderful warmth of character and generosity of spirit. She will be remembered with affection and respect by all who knew her. We all feel privileged to have worked with Gunilla and will miss her dearly. Jorge Díaz Cintas Gillian James Margaret Rogers This page intentionally left blank Contents ix Acknowledgements x Notes on Contributors 1 Introduction Jorge Díaz Cintas and Gunilla Anderman 1 Part I Subtitling and Surtitling 2 Subtitling for the DVD Industry Panayota Georgakopoulou 21 3 Subtitling Norms in Greece and Spain Stavroula Sokoli 36 4 Amateur Subtitling on the Internet Łukasz Bogucki 49 5 The Art and Craft of Opera Surtitling Jonathan Burton 58 6 71 Challenges and Rewards of Libretto Adaptation Lucile Desblache Part II Revoicing 7 Dubbing versus Subtitling: Old Battleground Revisited Jan-Emil Tveit 8 The Perception of Dubbing by Italian Audiences Rachele Antonini and Delia Chiaro 9 Transfer Norms for Film Adaptations in the Spanish–German Context Susana Cañuelo Sarrión 10 Voice-over in Audiovisual Translation Pilar Orero vii 85 97 115 130 viii Contents 11 Broadcasting Interpreting: A Comparison between Japan and the UK Tomoyuki Shibahara 140 Part III Accessibility to the Media 12 Interlingual Subtitling for the Deaf and Hard-of-Hearing Josélia Neves 151 13 Audio Description in the Theatre and the Visual Arts: Images into Words Andrew Holland 170 14 186 Usability and Website Localisation Mario De Bortoli and Jesús Maroto Ortiz-Sotomayor Part IV Education and Training 15 Teaching Screen Translation: The Role of Pragmatics in Subtitling Erik Skuggevik 197 16 Pedagogical Tools for the Training of Subtitlers Christopher Taylor 214 17 229 Teaching Subtitling in a Virtual Environment Francesca Bartrina 18 Subtitling: Language Learners’ Needs vs. Audiovisual Market Needs Annamaria Caimi 240 Index 252 Acknowledgements The appearance of this volume owes much to the hard work and patience of many people. First of all we would like to thank the contributors for waiting patiently while we co-ordinated and edited the many essays on different aspects of audiovisual translation from all over the world. Our thanks also go to Jill Lake and Palgrave Macmillan for being supportive from inception through to production. We are also very happy to acknowledge our debt to Gillian James who has worked with us with diligence and attentiveness at all stages on the manuscript and its preparation. GUNILLA A NDERMAN University of Surrey JORGE DÍAZ CINTAS Imperial College London ix Notes on Contributors Gunilla Anderman was Professor of Translation Studies at the Centre for Translation Studies, University of Surrey. She lectured on the undergraduate and postgraduate programmes of translation including the MA in Audiovisual Translation. A professional translator of drama with translations of the plays staged in the UK, USA and South Africa, she was the author of Europe on Stage: Translation and Theatre (2005). Rachele Antonini is a researcher at the University of Bologna, Italy, where she is concerned with the perception of dubbing and subtitled humour, the teaching of spoken language mediation, and child language. She has been working as a freelance interpreter, translator and subtitler for the past ten years. Francesca Bartrina has a PhD in Literary Theory and Comparative Literature and is a lecturer in Translation Studies at the University of Vic, Spain. Publications include Caterina Albert/Víctor Català. La voluptuositat de l’escriptura (2001), Violència de gènere (2002, editor) and articles and chapters on Audiovisual Translation, Translation Theory and Gender Studies. She is series editor for the book series Capsa de Pandora from Eumo Editoral. Łukasz Bogucki is Associate Professor and Head of Translation Studies in the Department of English Language of Łód́ University. He obtained his doctorate in 1997 and published A Relevance Framework for Constraints on Cinema Subtitling in 2004. Łukasz teaches interpreting, translation and supervises MA theses in both fields. He is also the Polish editor of The Journal of Specialised Translation. Jonathan Burton is a surtitler at the Royal Opera House, Covent Garden, in London, UK. He has written English surtitles for over 90 operas, and provides subtitles for operas on DVD and the BBC’s Cardiff Singer of the World competition. He has taught Twentieth-century Music History, gives lectures on musical and operatic topics, and writes programme notes for operas and concerts. He studied Music at Cambridge and Birmingham Universities. Annamaria Caimi is Associate Professor of English Language and Linguistics at Pavia University, Italy. Her research interests include ESP x Notes on Contributors xi teaching methodology, text linguistics, pragmatics, corpus linguistics and subtitling. Her present research is concerned with screen translation and different forms of subtitling in language learning. She is the editor of Cinema: Paradiso delle Lingue. I Sottotitoli nell’Apprendimento Linguistico (2002). Susana Cañuelo Sarrión is completing her doctoral work on intercultural transfers between Spain and Germany through literature and film translation. Thanks to a scholarship from DAAD, she has been involved in research in Leipzig, Germany since 2002. She is also working as a freelance translator and university teacher. Publications include: Historia universal de la literatura (2002) and translations of Elfriede Jelineks novels Las amantes (2004) and Obsesión (2005). Delia Chiaro is Professor of English Linguistics at the University of Bologna’s Advanced School in Modern Languages for Interpreters and Translators at Forlì, Italy. She is best known for her work as a scholar in humour but she is also the author of numerous publications on dubbing and subtitling as well as on translation quality. She has lectured across Europe, in the USA and in Asia. Mario De Bortoli joined Euro RSCG 4D in 2000 and has been involved in the localisation of international online and offline cross-cultural campaigns for clients in a large number of industries. In 1998 he left Venice for London with a BA in Simultaneous Interpreting and a postgraduate degree in Corporate Communication. He is also involved in the organisation of the annual Best Global Website Award, promoting excellence in localisation. Lucile Desblache is Reader in French and Translation at Roehampton University, London, UK. She holds two doctorates, from Paris VII and Clermont-Ferrand. Her research interests span across translation and comparative literature. She is the editor of The Journal of Specialised Translation and has worked as a consultant and translator for a range of companies including British Rail. Her adaptation of Albert Herring’s libretto into French was performed for Opéra de Chambre de France. Jorge Díaz Cintas is Senior Lecturer in Audiovisual Translation at Imperial College London. He is the author of several books and articles on subtitling and has recently published Audiovisual Translation: Subtitling (2007), co-written with Aline Remael. Since 2002, he has been the president of the European Association for Studies in Screen Translation. xii Notes on Contributors Panayota Georgakopoulou has a PhD in Subtitling from the University of Surrey, UK. She has set up subtitling modules and taught specialised translation courses at university level, while at the same time working in the subtitling industry. She is now the Managing Director of the European Captioning Institute, a subtitling company specialising in multi-language subtitling for DVD. Andrew Holland trained as an audio describer in 1993 at the National Theatre, UK and has been part of a team of describers there ever since. In 1998 he helped to set up VocalEyes as a professional service to make touring theatre accessible to the blind and partially sighted. As current Head of Description, he has also helped in the creation of a visually-impaired tour on the British Museum ‘Compass’ website, and over the years, has trained a large number of professional and volunteer describers. He has also had a number of his own plays produced. Jesús Maroto Ortiz-Sotomayor attended the universities of Granada, St. Petersburg, Washington and London, where he studied Translation, IT and e-Marketing. Currently he is working on a doctorate in International Digital Marketing at the Universitat Rovira i Virgili, Tarragona, Spain. He has worked for the translation companies STAR, Lionbridge and SDL, as well as for the advertising agency Euro RSCG. He is a member of the IoL, TILP and the Intercultural Studies Group. Josélia Neves holds a PhD in Subtitling for the Deaf and Hard-ofHearing from Roehampton University, UK. She teaches Audiovisual Translation at the Polytechnic Institute of Leiria and at the University of Coimbra, in Portugal. Her main interests lie in teaching AVT and in developing action research projects in the field of accessibility to the media. Pilar Orero holds an MA in Translation by Universitat Autònoma de Barcelona (UAB), Spain, and a PhD in Translation from UMIST, UK. She lectures at UAB where she coordinates the Online Masters course in Audiovisual Translation. She is the co-editor of The Translator’s Dialogue (1997) and the editor of Topics in Audiovisual Translation (2004). She is the leader of the networks CEPACC and RIID-LLSS, consisting of 18 Spanish universities devoted to media accessibility research and quality training. Tomoyuki Shibahara obtained his MA in Interpreting and Translating from the University of Bath, UK in 1999. He is a freelance interpreter, translator and lecturer, who, having worked as a broadcasting interpreter for the BBC World TV in London, is now a part-time lecturer at Notes on Contributors xiii Aoyamagakuin University and Dokkyo University in Tokyo. He also provides broadcasting interpreting for NHK (Japan Broadcasting Corporation) and translation for Discovery Channel. Erik Skuggevik is a lecturer in Audiovisual Translation at the University of Surrey, UK and also teaches on the MA programme in Intercultural Communication. He also lectures on the University of Westminster’s MA in Translation. He has subtitled over 500 films and translated books, short stories and cartoons. Erik is working on his doctorate in Intercultural Theory and Translation at University College, London. Stavroula Sokoli is a tutor in Spanish at the Hellenic Open University, Greece and is a professional subtitler. She holds a BA in English Language and Literature from the Aristotle University of Thessaloniki and an MA in the Theory of Translation from the Universitat Autònoma de Barcelona. She is a member of Tradumàtica (Translation and New Technologies). Christopher Taylor is Professor of English Language and Translation at the University of Trieste, Italy, where he is Head of the Language Centre. He has written extensively on aspects of translation, and in recent years, has concentrated on the analysis of multimodal texts for translation purposes and on the language of film (Tradurre il Cinema, 2000). He is also the author of Language to Language (1998). Jan-Emil Tveit is Associate Professor at the Norwegian School of Economics and Business Administration in Bergen. He was the first Director of Translation and Subtitling at TV2, Norway’s biggest commercial television company. Jan-Emil has lectured at New York University and given presentations at conferences in Canada, USA, Spain and Britain. He has written several articles and published Translating for Television: A Handbook in Screen Translation (2005). This page intentionally left blank 1 Introduction Jorge Díaz Cintas and Gunilla Anderman The Wealth and Scope of Audiovisual Translation In the twenty-first century, the media is omnipresent: to inform, arguably sometimes to misinform, to sell, to entertain and to educate. A quick perusal of traditional television programmes or cinema guides will testify to the growth and importance of the media and the need for audiovisual translation (AVT) in most countries. The reasons are manifold: a larger number of television channels at all levels, international, national, regional and local, means a sharp increase in the quantity and range of programmes required to meet the needs of broadcasting schedules. With the steady decline of analogue technology, the arrival of the digital era has also contributed to the diversification of offerings provided by television. In a very short time, corporations such as the BBC and ITV in the UK have more than doubled their number of channels and similar developments have also seen a record boom in new television channels at European level with 277 new channels launched in Europe since 2004 and in excess of 200 in 2005 (Hamilton and Stevenson, 2005). As for the cinema, the film industry seems to have emerged from the lean years when the video appeared to pose a serious threat to its continued existence, and now the number of cinema-goers again seems healthy. The flourishing celebration of film festivals, with hundreds of them taking place in any given year in all corners of the globe also testifies to this positive outlook. Add to this the advent of the DVD and the fact that the Internet is firmly established in our society and the picture is virtually complete. There is also the theatre, the opera and other live events where translation may be required in the form of surtitles; and the rapid developments we are witnessing in the field of accessibility to the media for 1 2 Jorge Díaz Cintas & Gunilla Anderman people with sensory impairments. Traditionally ignored in academic exchanges, subtitling for the deaf and hard-of-hearing (SDH) and audio description for the blind and the partially sighted (AD) are becoming part of our daily audiovisual landscape and attracting the interest of many scholars and practitioners. Given the many ways in which viewers can access audiovisual material – DVD, television, cinema, Internet – it is difficult to quantify with precision the percentage of foreign-language programmes translated and screened in any given country. Statistics available tend to be concerned with the number of films exported and imported for cinema release only, forgetting crucially any other films or audiovisual programmes (sitcoms, documentaries, TV series, musical concerts, cartoons, etc.) that are broadcast by private and public television channels and distributed on DVD and the Internet. Taken in isolation, the sparse figures available are likely to give a somewhat skewed overview of the real situation. At the expense of European productions, and according to Yvane (1995), an extremely high percentage of audiovisual programmes originate in the USA; 90% in Denmark, 90% in France, 90% in Germany, 94% in Greece, 75% in Ireland, 80% in Italy, 92% in Luxembourg, 90% in the Netherlands, 70% in Portugal, 95% in Spain, and 88% in the United Kingdom. Now a decade old, these figures are likely today to be the same or slightly higher as much has happened since the mid-1990s, notably the exponential growth of television stations at international, national, regional and local levels. There has been, however, a trend in the opposite direction. New low production cost audiovisual genres have emerged that, emulating the format of similar programmes designed in other countries and for other audiences, can be produced in the language of other communities without the need for translation. Examples abound and British reality television programmes such as Big Brother; quizzes in the vein of Who Wants to Be a Millionaire and The Weakest Link, soap operas depicting the daily routines of next-door neighbours, talk shows and contests for ‘wannabes’ have proved popular in most European countries and beyond. In most cases, the only translation required is that of the name of the programme. However, these developments do not necessarily mean that the overall need for translation is lower since, as already mentioned previously, there are many more television channels broadcasting many more hours. All these changes are mainly concerned with television productions only and more studies in the transnational trading of audiovisual programmes at all levels are needed in order for an overall picture to emerge of the real expansion of the field. Introduction 3 Nevertheless, despite the fact that the number of programmes produced in national languages would seem to be on the increase, the situation in countries where English is not the official language is such that a large volume of audiovisual programmes still needs to be translated. While traditionally feature films, television series, cartoons, sitcoms, soap operas and documentaries have been ideal candidates for this intercultural journey, the current growth in the need to provide and supply more audiovisual material for new channels has made broadcasters re-examine and broaden the range of programmes suitable for interlingual transfer. Subjects ranging from cookery, travel, DIY, fashion, interviews, gardening and awards ceremonies to political speeches, have started to find their way, via translation, to television sets in the living rooms of other countries. To a large extent, these developments help to account for the so-called revolution experienced in the field of audiovisual translation during the last couple of decades (Díaz Cintas, 2003). The move from analogue to digital technology and the potential afforded by the digitisation of images has also opened up new avenues, radically changing the essence of the industry. Together with the ubiquitous presence of the computer and the Internet, the arrival of the DVD can be hailed as one of the most exciting and revolutionary developments in recent decades. In just a few years, the DVD has become the favoured mode for distribution and consumption of audiovisual programmes. Its increased memory capacity, when compared to the CD; its superior image definition compared to the traditional VHS tape; and its greater flexibility, allowing the viewer to watch the programme on the TV set, the computer screen or a portable DVD player, constitute some of the main features that make it a favourite with producers as well as distributors and viewers. This has, in turn, resulted in new working practices, had an impact on the design of dedicated software programs, facilitated the work of scholars researching the field, and altered the consumer’s view of audiovisual products. The rate at which some of these changes in working practice are taking place is perhaps most striking in the field of subtitling. In a relatively short period of time, the process of subtitling has gone through a substantial transformation; what a decade ago was common practice is now viewed as out of date. Changes are happening at all levels – technological, working routines, audience reception, emergence of new translation modes and approaches – all bringing in their wake advantages and disadvantages that are fully discussed in this volume. Also increasing in volume is the amount of translation required in the field of AVT. Films that in ‘dubbing countries’ have traditionally 4 Jorge Díaz Cintas & Gunilla Anderman been dubbed for both cinema and VHS releases as well as television broadcasting are now also being subtitled for distribution on DVD; and classic movies that were only dubbed when first released are nowadays also available in subtitled versions on DVD. In addition, the increased memory capacity of the DVD has made it possible to include more material also in need of translation: known in the industry as ‘VAM’, value added material, producer’s edits, false takes, interviews and other related bonus material often takes up more time and space than the actual film itself. TV series, sitcoms and cartoons that are normally dubbed when broadcast on television also end up on DVDs with subtitles. The music industry, too, seems to be slowly moving away from the CD to the DVD in order to promote their live concerts or video clips that tend to be subtitled. In addition, individual films and other audiovisual programmes are now released in different formats such as cinema, television, DVD, CD-ROM and Internet, which combined with the increasing number of media companies operating in the field, has resulted in many films and programmes being translated more than once to meet consumer demand. Since the early days of the cinema, in order to make these audiovisual programmes comprehensible to audiences unfamiliar with the language of the original, different forms of language transfer on the screen have been required. In the main, there are two basic approaches to the translation of the spoken language of the original programme: to retain it as spoken or to change it into written text. In the first instance the original dialogue is replaced by a new soundtrack in the target language in a process generally known as revoicing. The replacement may be total, whereby we do not hear the original, as in lip sync dubbing and narration, or partial, when the original soundtrack can still be heard in the background, as in voice-over and interpreting. All these modes are available to the profession and some of them are more suited to particular audiovisual genres than others. Lip sync dubbing, for instance, is mainly used in the translation of films and TV series and sitcoms, whereas narration and voice-over tend to be more used in the case of documentaries, interviews and programmes on current affairs. When the decision has been taken to keep the original soundtrack and to switch from the spoken to the written mode, by adding text to the screen, the technique is known as subtitling. Quicker and a lot cheaper than dubbing, it has more recently become the favoured translation mode in the media world and comes hand in hand with globalisation. Despite the historically strong polarisation between advocates and detractors of the two different approaches, nowadays it is Introduction 5 generally accepted that different translation approaches make their own individual demands while remaining equally acceptable. The choice of one method in preference to another will simply depend on factors such as habit and custom, financial constraints, programme genre, distribution format and audience profile – to mention just a few. In this volume the full range of approaches is discussed as applied to a number of different languages. For several years full access to audiovisual media for minority social groups such as the deaf and the blind has been an issue. Recent developments and studies show that the needs of these groups are increasingly being catered for and this field of expertise now holds an established position within audiovisual translation. Accessibility is a new key concept; an umbrella term that encompasses all associated new modes of translation. According to statistics (Neves, 2005: 79), between 1% and 5% of the population of any country, are deaf or hearing-impaired. The number of people in these categories is growing as more people are living into old age and they account for significant numbers on the continents of Europe and North America. According to Hay (1994: 55), around 30% of all Americans over 65 years of age have some degree of hearing loss. As for Europe, figures presented at the 2003 international conference Accessibility for All projected that by 2015 there will be over 90 million adults affected by a hearing loss (Neves, 2005: 79). These figures clearly call for a more consistent and systematic approach to making it possible for viewers with sensory impairment to gain access to television and other media. Since the mid 1970s, subtitling for the deaf and hard-of-hearing (SDH), also known as (closed) captioning in American English, has formed part of the audiovisual landscape in order to ensure greater democratic access to audiovisual programming. On television, these subtitles are broadcast by means of an independent signal, activated only by those interested, by accessing pages 888 or 777 of teletext in most European countries. In North America they are transmitted on what is known as line 21. The oral content of the actors’ dialogue is converted into written speech, which is presented in subtitles of up to three, or occasionally four, lines. They generally change colour on television depending on the person who is talking or the emphasis given to certain words within the same subtitle. Besides the dialogue, they also incorporate all paralinguistic information that contributes to the development of the plot or to the creation of atmosphere, which a deaf person cannot access from the soundtrack, e.g. a telephone ringing, laughter, applause, a knock on the door, and the like. 6 Jorge Díaz Cintas & Gunilla Anderman Visual impairment is one of the most age-related of disabilities and the vast majority of blind and partially sighted people are elderly. According to the European Blind Union (EBU, online) 7.4 million of Europeans – close to 2% of the total population – have a significant visual disability. As demographic trends show that the number of elderly people is on the increase, one can only expect that the percentage of people with visual impairment will also increase in the near future. In the field of accessibility, a more recent development has been audio description for the blind and the partially sighted (AD), a service which is rapidly gaining momentum and visibility. It can be defined as an additional narration that fits in the silences between dialogue and describes action, body language, facial expressions and anything that will help people with visual impairment follow what is happening on screen or on stage. Audio describing television programmes, films, plays, sporting events and even an exhibition of paintings is now technically relatively easy, especially with digital technology (Marriott and Vale, 2002). In the field of audio description, as well as in SDH, English-speaking countries such as the UK, the USA, Canada and Australia seem to be leading the rest of the world. However, coinciding with the European Year of People with Disabilities in 2003 (www.eypd2003.org) various actions were taken by the EU at international level and by individual European countries at national level, to raise awareness and to foster changes aimed at improving the lives of people with disabilities. Taking advantage of the situation, a number of countries in the European Union took the opportunity to launch initiatives leading to widening accessibility to the audiovisual media for all their citizens. In addition to increased general awareness, efforts already undertaken have also resulted in actions such as the passing of legislation making it compulsory to broadcast a minimum number of hours with SDH and AD,1 and the creation of national bodies responsible for monitoring developments in the field.2 A welcome step forward towards further increase in accessibility is the arrival of digital television, a development that is bound to change prevailing notions and attitudes. Although television is likely to remain a favoured medium for subtitling for the deaf and hard-of-hearing and for audio description for the blind and partially sighted, the future of these two services goes beyond the scope of television. SDH and AD are already gaining major importance in the DVD market and are also in regular use in the cinema. Intralingual subtitling is being complemented by interlingual subtitling for the deaf and hard-of-hearing and new modes of translation are emerging, such as Introduction 7 audio subtitling, in order to make subtitled programmes accessible to the blind in countries where a large percentage of the programmes are commercialised in a foreign language with subtitles. This flourish of activities in the field of AVT at industrial level has had a positive knock-on effect in the university world where AVT is now emerging as a thriving academic discipline for teaching and research. For many years, the skills of the trade were acquired in the workplace, away from educational establishments. Despite the importance of the role of AVT in our daily lives, on the whole universities have been slow in curriculum design and the development of new courses. This situation is, however, rapidly changing and there are now many different specialist courses in AVT on offer at universities worldwide. Audiovisual translation also has an important role to play in the classroom. Material and exercises may be drawn from the area of AVT and used in the process of second language acquisition. Subtitling can be a powerful training and teaching tool in the foreign language learning class. Hearing the original language while reading the dialogue in context provides a stimulating environment for students to consolidate what they are learning, enrich their vocabulary at the same time as they become familiar with and absorb the culture of a foreign language in an authentic setting. However, subtitles are also a potent force in language acquisition outside the classroom. In Singapore it is a significant factor in the life of young Singaporeans who have been brought up speaking English but in the presence of older generations find themselves watching programmes in Mandarin. Subtitled in English, Chinese drama on television is of great help in facilitating the understanding of the Mandarin they hear; the importance of the subtitles to the language acquisition process is considerable. And in some parts of Europe subtitled programmes serve as the most important means of acquiring mastery of English. As reported in Gottlieb (2004: 96), when asked why they wanted to learn how to read, 72 out of 75 first-grade Danish students declared that they wanted to be able to read the subtitles on television; the ability to read books no longer appeared to be a major incentive. For the citizens of a nation who reportedly listen to English on television and video for an average of one hour per day, this would hardly come as a surprise (Gottlieb, 1994: 153–7 and Gottlieb, 1997: 151–3). Reading subtitles in the foreign language while watching a foreign language programme also appears to be beneficial to vocabulary acquisition, as shown by Belgian studies (Van de Poel and d’Ydewalle, 2001). Further advantageous uses of subtitles include the learning of minority 8 Jorge Díaz Cintas & Gunilla Anderman languages through subtitled film sequences (Baldry, 2001 reported in Gottlieb, 2004: 88). More recently, audiovisual translation has evolved to the point where, as a discipline, it is now one of the most vibrant and vigorous fields within Translation Studies. With the celebration of several international specialised conferences and the publication of edited volumes and monographs on topics in AVT, research in the field has also gained visibility in a comparatively short period of time. Although at present, audiovisual translation is experiencing an unprecedented boom of interest and activity at all levels, a number of problematic issues remain to be addressed. The changes taking place in the profession are fast, not always allowing sufficient time for full adjustment. Old methods tend to compete with new techniques, and consistency is not always maintained. Subtitle styles tend to vary from country to country, even from company to company. In recent years, calls for a ‘Code of Best Practice in Audiovisual Translation’ have been recurrent, but would such a document be workable? Is it necessary or even feasible for all aspects to be harmonised? If so, should it be done at local, national or international level? In the audiovisual field, the global dominance of the English language in many spheres becomes even more of an issue. Production and distribution companies, run in most cases with the help of American capital, have US interests at their heart. Hollywood, the powerhouse of the western film industry, mirrors and exports cultural patterns that have an immediate impact on other languages and cultures. New developments in the industry mean that in addition to Los Angeles, London is also becoming a world nerve centre for AVT. Given this situation, can we really talk of an ‘exchange’ of cultural products? Can the term ‘intercultural’ be used to describe this process? Or are we simply faced with a one-dimensional flow of information and ideas from the metropolis to the ‘territories’, as countries are known in the profession’s jargon? Structure of the volume Audiovisual Translation: Language Transfer on Screen attempts to provide answers to some of these questions. Others, however, will remain open to debate but in an expanding field rapidly generating increasing interest at professional, educational and research levels. The seventeen contributions to the subject contained in this volume offer a detailed overview of most of the translation modes used in the audiovisual media. Part I of the book deals with Subtitling and Introduction 9 Surtitling, two techniques that respect the original soundtrack and add the translation in the form of short written texts. In Subtitling for the DVD Industry, Panayota Georgakopoulou takes a close look at the recent growth in the DVD market and the rise of subtitling as a rapidly expanding international industry. Georgakopoulou discusses some of the constraints (and possibilities) inherent in subtitling in general, then proceeds to examine how these are dealt with in the DVD subtitling industry. The focus is on subtitling from English into other languages, and all examples used come from the database of the European Captioning Institute (ECI), which is a UK subtitling company and one of the world leaders in multilingual DVD subtitling. Georgakopoulou’s contribution concludes by analysing the impact that centralisation is having on the profession and discussing the potential offered by the ‘template’, a file containing the master (sub)titles in English and used as the basis for translation into all languages required in a given project. In emphasising the value of norms as a heuristic tool in the field of Translation Studies and their major contribution to the evolution of Descriptive Translation Studies, Stavroula Sokoli proposes bringing in an evaluative element. In Subtitling Norms in Greece and Spain she explains that the mere description of translation behaviour for its own sake may not provide useful results, whereas the study of norms is bound to give insight to the intersubjective sense of what is ‘proper’, ‘correct’, or ‘appropriate’. Since norms are not directly observable, a possible approach to them is through their manifestations, whether textual or extra-textual. The focus in her contribution is on textual sources of norms, that is, regularities in the choices made by subtitlers, as manifested in the translated films themselves. The texts under study are the subtitled versions in Spanish and Greek of the films The English Patient (Anthony Minghella, 1996) and Notting Hill (Roger Michell, 1999). Given the paper’s qualitative and descriptive nature, rather than laws or absolute truths, possible explanations are proposed and more hypotheses are generated, which in turn become the basis for future research. Sokoli’s proposals can be used not only to explain and predict the way subtitles are done, but also to train future subtitlers. Łukasz Bogucki’s contribution, Amateur Subtitling on the Internet, looks at Polish subtitles of the film The Fellowship of the Ring in order to assess the criteria for quality assessment and their application to audiovisual translation. As a voice-over country, Poland has had little subtitling tradition to date, and only feature films are subtitled for their cinema release. However, recently a new kind of subtitling has developed. With the help of freeware computer programs and the Internet, amateur 10 Jorge Díaz Cintas & Gunilla Anderman subtitling is undertaken by non-professionals and governed by dramatically different constraints than professional subtitling. Often, the Source Text (ST) is digitised from a low-quality recording, and the end result – that is the subtitles in the Target Language (TL) – is conditioned by how much the subtitle producer has heard and understood from the original dialogue. As amply illustrated in Bogucki’s contribution, this approach is likely to result in a multitude of mistakes and misinterpretations. While subtitles are translated texts usually displayed below the image, as on a cinema or television screen, surtitles are most often displayed above the stage, in live opera and theatre performances. In his contribution entitled The Art and Craft of Opera Surtitling, Jonathan Burton highlights the fact that subtitling and surtitling differ substantially in their requirements and techniques, and proceeds to discuss surtitling as the most common approach to the translation of operatic texts. He offers a detailed analysis of the process, without forgetting the problems and pitfalls facing translators in this specialised field. Lucile Desblache’s contribution, Challenges and Rewards of Libretto Adaptation, also has music at its core. Translating for the stage means being torn between remaining faithful to the author’s intentions and providing a version of the text which takes into account a range of other factors. The original message, with its cultural references and context, must be communicated to the audience, but the transfer of non-semantic aspects of the ST, as well as some extra-textual aspects, is of vital importance to successful adaptation. These factors must all be taken into consideration in libretto adaptation, but in addition, the special nature of opera imposes musical restrictions on the text, as the music cannot be changed and dictates, to a certain degree, which words can be used. Surtitling has been in existence for over two decades and has now been adopted by most opera houses. Yet, in certain cases, the decision is still made to adapt libretti for live performances. Drawing from her own experience of writing the French version of Albert Herring, a comic chamber opera by Benjamin Britten, the author discusses the problems encountered in translation work of this kind. Three orders of difficulty are identified relating to cultural equivalence, humour and rhyming. Part II, Revoicing, is centred on the technique of replacing the soundtrack of the original audiovisual programme. Seen by many as the opposing strategy to subtitling, the arguments in favour or against each of these two approaches have been commonly debated in the existing literature. In Dubbing versus Subtitling: Old Battleground Revisited, Introduction 11 Jan-Emil Tveit attempts to assess the pros and cons of dubbing and subtitling, and to determine if either of the two is the ‘better’ option. Included among the disadvantages of subtitling are loss of information due to the transition from spoken to written mode as well as the frequently occurring failure to convey the dialectal and sociolectal features of spoken language. In addition, problems resulting from visual, spatial and decoding constraints may cause further difficulties if subtitling is the chosen method. However, considering the loss of authenticity and trans-national voice qualities in dubbing, not to mention the fact that it is more costly and time-consuming, the author concludes that, on the whole, the subtitling approach to audiovisual translation is to be preferred to dubbing. As mentioned above, revoicing can be carried out in two different ways: by completely erasing the voices of the source programme (dubbing) or by juxtaposing a new soundtrack to the original one (voice-over, interpreting, audio description). As is well known, Italy is commonly labelled a ‘dubbing country’ which, together with Austria, France, Germany and Spain, has adopted a tradition of dubbing rather than subtitling, the preferred mode of audiovisual translation in Greece, Portugal, Scandinavia and the UK. However, the situation in dubbing countries is rapidly changing as a choice between dubbing and subtitling is increasingly becoming available. In their contribution concerned with the level of audience tolerance to dubbing in Italy, The Perception of Dubbing by Italian Audiences, Rachele Antonini and Delia Chiaro discuss the results of a large-scale research project which set out to explore and assess the quality of dubbed television programmes. Based on a corpus of over three hundred hours of dubbed television programmes and by means of web technology, over five thousand Italian viewers were tested on their perception of dubbed Italian. After conducting an ANOVA multivariate analysis of the raw data, an index was created, accounting for the perception of linguistic naturalness in dubbed Italian by viewers. After analysing the responses of a robust sample of randomly selected viewers, the results seem to point towards a certain tolerance of what viewers recognise as being unnatural, an acceptance of linguistic features adopted on screen which are virtually non-existent both in autochthonous Italian fictional programmes and in established electronic corpora of naturally occurring Italian. The double transfer of a literary work to the screen and of the resulting film into a different culture, generates a series of highly interesting and mutually influential relationships between two cultures, involving three different types of transfer, namely: film adaptation, literary 12 Jorge Díaz Cintas & Gunilla Anderman translation and audiovisual translation. The aim of Susana Cañuelo Sarrión’s contribution, Transfer Norms for Film Adaptations in the Spanish– German Context, is two-fold. First she identifies the works from a corpus of Spanish films based on literary works and produced between 1975 and 2000 which have reached the German market and sets out to analyse them from a Polysystem Theory perspective, taking into account the characteristics of German cinematography. Second, a number of transfer norms for film adaptations is proposed on the basis of recurring patterns detected in adaptations of literary works in Spain and their subsequent distribution and reception in Germany. In an audiovisual market dominated by English-spoken productions, Cañuelo Sarrión’s contribution has the virtue of widening the scope of languages involved to German and Spanish. In Voice-over in Audiovisual Translation, Pilar Orero examines the many definitions and descriptions from the field of Translation Studies for the audiovisual translation mode known as voice-over, in which two soundtracks in two different languages are broadcast at the same time. She then describes the technique of translation for voice-over taking into consideration a distinction which, suggested by Luyken et al. (1991), has never been properly developed, namely the difference in the process depending on whether the translation takes place during the production or during the post-production phases. As an audiovisual translation mode, interpreting has received very little attention by academics or professionals in the past. In Broadcasting Interpreting: A Comparison between Japan and the UK, Tomoyuki Shibahara addresses the question of conformity to different house styles. The definition of the term ‘broadcasting interpreter’ and its history are discussed as are the many differences in broadcasting style between Japan and the UK. For example, broadcasting interpreting at the BBC Japanese Unit emphasises the quality of Japanese, while accuracy is the top priority at Japan Broadcasting Corporation (NHK). However, both broadcasting organisations require their interpreters to edit the information. Based on his own experience, the author describes the interpreting style, policy and quality control systems of the above two organisations. In addition, he provides a comparison of the employment style of both the BBC and NHK as well as an analysis of the pros and cons of working as a full-time member of staff and as a freelance interpreter. Part III of the volume contains contributions discussing Accessibility to the Media. In countries such as the UK and the USA, subtitling as a concept is usually understood as intralingual (English into English) Introduction 13 subtitling for the deaf and hard-of-hearing (SDH). However, in Portugal, Greece and the Scandinavian countries, for instance, it will be interlingual (English into Portuguese, Greek or any of the Scandinavian languages) open subtitling that comes to mind. In these countries, interlingual subtitling for the deaf and hard-of-hearing is rarely seen as a specific kind of subtitling. Even though some DVDs now carry the option of interlingual SDH, this is still not common practice, perhaps under the assumption that standard interlingual subtitles offer enough information for all, regardless of the fact that some viewers might have a hearing impairment. As discussed by Josélia Neves in Interlingual Subtitling for the Deaf and Hard-of-Hearing this particular group of viewers has special needs when it comes to gaining true access to audiovisual material. In some respects, standard interlingual subtitles offer more information than they are able to absorb; in others, an extra input needs to be added for the full semiotic message to be conveyed adequately. Looking into the future and taking advantage of television turning digital and interactive, Neves advocates the production of multiple solutions (dubbing, interlingual subtitles, intralingual and interlingual SDH, adapted subtitling) for each audiovisual programme so that they can best suit the specific needs of different audiences. Audio description is the other main technique aimed at widening accessibility to the media for viewers with sensory disabilities. After giving a brief introduction to what audio description is, Andrew Holland, in Audio Description in the Theatre and the Visual Arts: Images into Words, offers an informative account of the role of audio describers, professionals who act as the eyes of the visually impaired and transfer what they see into words. They express and bring alive, verbally, what they see at the same time as the action takes place, on stage or on the screen. Focusing on audio description for the theatre, Holland equates the importance of the audio describer’s sensitivity with the task of relating the plot and action of a theatre or screen performance. Providing an audience with the subtext or the unspoken message of a production is often as important as conveying facts, if not more so. Audio description represents a form of audiovisual translation that is rapidly gaining increasing attention as the needs of the visually impaired are beginning to receive the degree of concern more frequently granted to the hard-of-hearing in the past. The Internet has meant a revolution in our lives in general, and in the field of translation and business in particular. In their contribution, Usability and Website Localisation, Mario de Bortoli and Jesús Maroto 14 Jorge Díaz Cintas & Gunilla Anderman Ortiz-Sotomayor discuss the limited success as well as the problems encountered by big international companies when translating their English websites into the languages of the main foreign markets. According to the authors, for a site to be well received and successful in the twenty-first century it has to address the unconscious hidden aspects of what constitutes a community. Localisation needs to build in an in-depth knowledge of the local culture which in turn means that a multilingual website cannot be researched and developed in English and then simply sent off to be translated; rather, every aspect needs to be discussed and studied prior to development and subsequent implementation. Addressing these cultural differences requires the early attention of professionals with expertise in a variety of fields. Only through a synthesis of interdisciplinary expertise will there be a guarantee of the enormous benefits that globalisation can offer to truly international business. Part IV, Education and Training, is devoted to the teaching and training of future experts and professionals in the field of subtitling. The most distinctive feature of subtitling is the need for economy of translation. There is rarely enough space and time to fit all potentially transferable material in an audiovisual programme onto the stipulated number of lines and characters. This is where the subtitler’s sensitivity to plot, character and film narrative becomes crucial in order to determine the information to keep in and to leave out. But how can this sensitivity be defined? And how does it actually work? In his contribution, Teaching Screen Translation: The Role of Pragmatics in Subtitling, Erik Skuggevik looks at the issue of the sensitivity of the subtitler as linked to several translation strategies, providing students and trainees with underused tools to break down the constituents of both the dialogue of the original and the subtitles of the translated programme. Drawing from the Speech Act Theory developed by Grice (1975) and the six functions of communication identified by Jakobson (1960), the author proposes an analysis of communication that enables us to quantify the constituents of subtitling more closely. In his search for useful tools for the subtitler, Christopher Taylor draws from the potential offered by the multimodal transcription as devised by Thibault and Baldry (2000). In Pedagogical Tools for the Training of Subtitlers, he discusses how this methodology allows for the minute description of film texts, thereby enabling subtitlers to base their translation choices on the meaning already provided by other semiotic modalities contained in the text such as visual elements, music, colour, and camera positioning. This process, though time-consuming has Introduction 15 proved to be an extremely valid pedagogical instrument. A second stage in this research has led to the creation of transcriptions based on phasal analysis, following the ideas of Gregory (2002). Film texts are divided into phases and sub-phases based on the identification of coherent and harmonious sets of semiotic modalities working together to create meaning in recognisable chunks, rather in the manner of written text. Such texts are analysed in terms of their phasal construction, and also in terms of the transitions separating the phases. This further extension of the original tool provides the basis for a thorough analysis of any film text, and provides a useful addition to the pedagogy of film translation. As mentioned before, audiovisual translation is closely linked to technology and any new advances are bound to have a knock-on effect on the discipline, particularly on subtitling. The Internet has been one of such advances. Tutors face new challenges in the global era and perhaps the most revolutionary is the proliferation of online multimedia courses in higher education. As discussed in Francesca Bartrina’s contribution, Teaching Subtitling in a Virtual Environment, tutors need to transfer teaching skills from the face-to-face classroom to the virtual environment. This in turn gives rise to new trends in teaching methods and calls for the acquisition of new skills in the use of Information and Communication Technology. More specifically, in the case of online subtitling courses, it means that instructors need to stay abreast of innovations in digital subtitling technology. Bartrina’s contribution deals with the challenge posed by virtual courses to the teaching of subtitling and with the way skills of future professionals may be developed by online learning. The final contribution in this volume, by Annamaria Caimi, Subtitling: Language Learners’ Needs vs. Audiovisual Market Needs, highlights the importance of subtitling in language learning and suggests possible convergences between linguistic, educational and economic goals. Such an approach requires meeting different expectations and interests, which can be accomplished by encouraging the distribution of audiovisual programmes with subtitles, in order to introduce foreign languages through entertainment. The educational objectives, which consider subtitling as a linguistic product, are quality-oriented, whereas the economic objectives, whose main concern is the distribution of subtitled audiovisual programmes, are quantity-oriented. In order to satisfy both needs, the author tackles the issue from three different perspectives: (1) the linguistic perspective, which focuses on the accuracy and appropriateness of language transfer through the application of principles of translation theory; (2) the foreign language 16 Jorge Díaz Cintas & Gunilla Anderman teacher’s perspective, whose priority is to single out the most natural methods and techniques to facilitate foreign language acquisition; (3) the audiovisual marketer perspective, whose task is to meet the requirements of consumers in order to reach and maintain full demand for the product. It is hoped that these seventeen studies will together constitute a rounded vision of the many different ways in which audiovisual programmes cross linguistic barriers and frontiers and, in particular, of the importance of audiovisual translation at the present time. Notes 1. Many countries, for example, Belgium, Canada, the Netherlands, the UK and USA, have established targets to provide SDH for 100% of their television programmes (Remael, 2007). As for AD and sign language interpreting, the targets are lower, at between 10% and 20% (Orero, 2007). 2. The creation of the Centro Español de Subtitulado y Audiodescripción (CESyA) is a good example of the type of initiatives taken by some countries. More information on this centre can be found on www.cesya.es. References Baldry, A. (2001) Promoting Computerised Subtitling: A Multimodal Approach to the Learning of Minority Languages. Unpublished manuscript. University of Pavia. Díaz Cintas, J. (2003) ‘Audiovisual translation in the third millennium’. In G. Anderman and M. Rogers (eds) Translation Today. Trends and Perspectives (pp. 192–204). Clevedon: Multilingual Matters. EBU (online) The Future of Access to Television for Blind and Partially Sighted People in Europe. Paris: European Blind Union. www.euroblind.org/fichiersGB/accessTV.html#intro Gottlieb, H. (1994) Tekstning – synkron billedmedieoversættelse. Copenhagen: University of Copenhagen. Gottlieb, H. (1997) Subtitles, Translation and Idioms. Copenhagen: University of Copenhagen. Gottlieb, H. (2004) ‘Language-political implications of subtitling’. In P. Orero (ed.) Topics in Audiovisual Translation (pp. 83–100). Philadelphia and Amsterdam: John Benjamins. Gregory, M. (2002) ‘Phasal analysis within communication linguistics: two contrastive discourses’. In P. Fries et al. (eds) Relations and Functions within and around Language (pp. 316–45). London: Continuum. Grice, H.P. (1975) ‘Logic and conversation’. In L. Cole and J.L. Morgan (eds) Syntax and Semantics (pp. 41–58). New York: Academic Press. Hamilton, F. and Stevenson, D. (2005) ‘Sex and shopping drive new television channel launches as Europe sees a record boom in new TV channels’. Screen Digest, 13 September. www.screendigest.com/press/releases/FHAN-6G7BXGpressRelease.pdf Introduction 17 Hay, J. (1994) Hearing Loss: Questions You Have ... Answers You Need. Pennsylvania: People’s Medical Society. Jakobson, R. (1960) ‘Closing statement: linguistics and poetics’. In R. DeGeorge and F. DeGeorge (eds) (1972) The Structuralists: From Marx to Levi-Strauss (pp. 85–122). New York: Anchor Books. Luyken, G.M., Herbst, T., Langham-Brown, J., Reid, H. and Spinhof, H. (1991) Overcoming Language Barriers in Television: Dubbing and Subtitling for the European Audience. Manchester: European Institute for the Media. Marriott, J. and Vale, D. (2002) Get the Picture: Making Television Accessible to Blind and Partially Sighted People. London: Royal National Institute of the Blind, Campaign report 19. Neves, J. (2005) Audiovisual Translation: Subtitling for the Deaf and Hard-of-Hearing. London: Roehampton University. PhD Thesis. http://rrp. roehampton.ac.uk/artstheses/1 Orero, P. (2007) ‘Sampling audio description in Europe’. In J. Díaz Cintas, P. Orero and A. Remael (eds) Media for All: Subtitling for the Deaf, Audio Description and Sign Language (pp. 111–25). Amsterdam: Rodopi. Remael, A. (2007) ‘Sampling subtitling for the deaf and the hard-of-hearing in Europe’. In J. Díaz Cintas, P. Orero and A. Remael (eds) Media for All: Subtitling for the Deaf, Audio Description, and Sign Language (pp. 23–52). Amsterdam: Rodopi. Thibault, P. and Baldry, A. (2000) ‘The multimodal transcription of a television advertisement: theory and practice’. In A. Baldry (ed.) Multimodality and Multimediality in the Distance Learning Age (pp. 311–85). Campobasso: Palladino Editore. Van de Poel, M. and d’Ydewalle, G. (2001) ‘Incidental foreign-language acquisition by children watching subtitled television programs’. In Y. Gambier and H. Gottlieb (eds) (Multi)Media Translation. Concepts, Practices and Research (pp. 259–273). Amsterdam and Philadelphia: John Benjamins. Yvane, J. (1995) ‘Babel: un soutien actif aux transferts linguistiques’. Translatio, Nouvelles de la FIT-FIT Newsletter XIV (3–4): 451–60. This page intentionally left blank Part I Subtitling and Surtitling This page intentionally left blank 2 Subtitling for the DVD Industry Panayota Georgakopoulou We all know what subtitles are. Luyken et al. (1991: 31) define them as: ... condensed written translations of original dialogue which appear as lines of text, usually positioned towards the foot of the screen. Subtitles appear and disappear to coincide in time with the corresponding portion of the original dialogue and are almost always added to the screen image at a later date as a post-production activity. Interlingual subtitling is a type of language transfer in which the translation, that is the subtitles, do not replace the original Source Text (ST), but rather, both are present in synchrony in the subtitled version. Subtitles are said to be most successful when not noticed by the viewer. For this to be achieved, they need to comply with certain levels of readability and be as concise as necessary in order not to distract the viewer’s attention from the programme. So, what are the techniques used to make subtitles unobtrusive? And what is the subtitler’s role? The answers to these questions can be found if we take a closer look at the technical, textual and linguistic constraints of subtitling. Technical constraints The technical spatial and temporal constraints of audiovisual programmes relate directly to the format of subtitles. 21 22 ● ● ● Panayota Georgakopoulou Space. In the limited space allowed for a subtitle there is no room for long explanations. Two lines of text are usually the norm, and the number of characters per line depends on a number of factors, including the subtitling workstation used. Since readability of the text is of paramount importance, it has been suggested that an ideal subtitle is a sentence long, with the clauses of which it consists placed on separate lines (Díaz Cintas and Remael, 2007:172–80). Time. The length of a subtitle is directly related to its on-air time. Accurate in and out timing is very important and the text in the subtitles should always be in balance with the appropriate reading time setting. No matter how perfect a subtitle is in terms of format and content, it will always fail to be successful if viewers do not have enough time to read it. A lower word per minute (wpm) or character per minute (cpm) setting is applied, for example, when subtitling children’s programmes, as children cannot reach adult reading speeds.1 Presentation. Subtitles can take up to 20% of screen space. Important factors for their legibility are the size of the characters, their position on screen, as well as the technology used for the projection of subtitles in the cinema (DTS or Dolby), TV broadcast, DVD emulation, etc., as it affects their definition. In our digital age, most of these problems have been solved. In DVD subtitling, for instance, the choice of any font and font size supported by Windows is possible, unlike teletext subtitling for television, where this is not the case. These technical constraints determine subtitlers’ work practice and their linguistic choices. Textual constraints In subtitling, language transfer operates across two modes, from speech to writing, from the soundtrack to the written subtitles. This shift of mode creates a number of processing and cohesion issues that make it difficult to maintain the filmic illusion in the target product. Oral–aural processing Since in subtitling both source and target texts are present simultaneously (see Gottlieb, 1994: 265 for details on the four channels that compose the audiovisual text), the viewer of a subtitled Subtitling for the DVD Industry 23 programme has at least two different types of information on which to concentrate: the action on the screen, and the translation of the dialogue, that is the subtitles. This adds to the verbal information that might appear in the original programme in the form of inserts and which the viewers have to process through the visual channel, making it more difficult for them to relax and enjoy the programme. The situation becomes more difficult when the timing of the subtitles is not satisfactorily done. When a subtitle is continued over a shot change, for example, the viewer may think that it is a new subtitle and re-read it, losing precious viewing time. Also, the temporal succession of subtitles is quite different from the linear succession of sentences in a book; it does not allow the eye to move backwards or forwards to clarify misunderstandings, recapitulate the basic facts or see what will happen next. This can be done, however, when watching a film or a programme on video or DVD by using the rewind function. As a result, in order for the subtitled text to be successful, it needs to preserve the ‘sequence of speech acts [ ... ] in such a way as to relay the dynamics of communication’ (Mason, 1989: 15). A few rough rules are usually observed by subtitlers to help minimise the potentially negative effects of these extra processing demands made by the viewer: (a) When the visual dimension is crucial for the comprehension of a particular scene, subtitlers should offer only the most basic linguistic information, leaving the eyes of the viewers free to follow the images and the action. (b) Conversely, when important information is not in the images but in the soundtrack, subtitlers should produce the fullest subtitles possible, to ensure that the viewers are not left behind. (c) The presentation of the subtitles, the way in which the words of each subtitle are arranged on the screen, and on each subtitle line, can help enhance readability. This last point relates to the role of grammar and word order in subtitling. The simpler and more commonly used the syntactic structure of a subtitle, the least effort needed to decipher its meaning. For example, in all the subtitled versions of the following text from Hitchcock’s Psycho, the main and subordinate clauses of the sentence are placed in separate subtitle lines and the syntax is simplified 24 Panayota Georgakopoulou through a re-arrangement of the original phrase as shown in the English subtitle: Example 1 Original dialogue 01.19.21.04 – 01.19.24.22 One thing people never ought to be when they’re buying used cars is in a hurry. English subtitle French subtitle People never ought to be in a hurry On ne devrait pas être pressé when buying used cars. quand on achète une voiture d’occasion. Italian subtitle German subtitle Non si dovrebbe mai andare di fretta quando si compra una macchina. Beim Gebrauchtwagenkauf sollte man es nie eilig haben. Even appropriate line breaks within a single subtitle can facilitate comprehension and increase reading speed if segmentation is done into noun or verb phrases, rather than smaller units of a sentence or clause. In the example below from Weird Science it is possible, within the time allowed, to read the text of both subtitles, which are identical from a semantic point of view. However, of the two solutions in Example 2, the subtitle format in Option 2 is easier to read as the text is broken up after the noun phrase ‘your parents’. Example 2 Original dialogue 01.03.53.01 – 01.03.57.11 I don’t understand. How come your parents trust you all of a sudden? Option 1 Option 2 I don’t understand why your parents trust you all of a sudden. I don’t understand why your parents trust you all of a sudden. Finally, in this example from the audio-commentary of Hitchcock’s Vertigo, we can see how long sentences are split up into shorter ones in Subtitling for the DVD Industry 25 an obvious attempt to increase reading speed. As in this case, conjunctions usually provide a natural split for subtitle breaks. Example 3 Original dialogue English subtitles Option 1 Option 2 I met with him, talked to him, I’ve told 01.26.31.07 – 01.26.34.08 him the whole story and he started the Met with him, told him the whole story painting and one day he said he must and he started the painting. meet Vera Miles to keep on going. 01.26.35.06 – 01.26.38.23 One day he said he must meet Vera Miles to keep on going. Textuality issues Because of the limited space generally available for subtitles, certain elements of the soundtrack have to be omitted, and the obvious solution is to do away with redundant elements of speech. Redundancy helps participants in a conversation grasp its intended meaning more easily and its elimination from film dialogue may, therefore, weaken cohesion in the subtitled text. The question then is to what extent the predictability of discourse is affected by the systematic deletion of redundant features and the impact this may have on the viewers’ understanding of the narrative. There is, however, a strong link between the dialogue of a film and the context in which it takes place. Apart from linguistic redundancy in audiovisual programmes, there is also situational redundancy that usually works in favour of the translator. The visual information often helps viewers process the subtitles, and to a certain extent this compensates for the limited verbal information they contain. For example, aspects of interpersonal communication may be found in intonation, rhythm and the facial and kinesic movements that accompany the dialogue which are, to an extent, universal (Nir, 1984: 90). When, in Braveheart, Mel Gibson cries out ‘Hold! Hold! Hold! Now!’ at the moment when the British cavalry is about to attack the frontline of the Scottish army, the viewers can easily understand what is happening, even without a translation. This is also known as the ‘feedback effect’ of films (Nedergaard-Larsen, 1993: 214). 26 Panayota Georgakopoulou Change in mode The shift of mode from speech to writing presents the subtitler with yet more challenges. Characteristics of spontaneous speech, such as slips of the tongue, pauses, false starts, unfinished sentences, ungrammatical constructions, etc., are difficult to reproduce in writing. The same goes for dialectal, idiolectal and pronunciation features that contribute to the moulding of screen characters. The use of a pseudo-phonetic transcription to reproduce a regional or social dialect in the subtitles, for instance, would not be helpful as it would hinder the readability of the text by adding to the reading time of the subtitle, and also hinder the comprehension of the message by obscuring the style. What then is the subtitler to do? Certain spoken features may need to be rendered in the subtitles if their function is to promote the plot. But rather than reproducing mistakes in an uneducated character’s speech, a subtitler can make use of appropriate, usually simpler, vocabulary in order to indicate education, regional dialect or social class of the character. Or the decision may be taken not to reproduce the stuttering in a character’s speech since viewers have resource to this information already from the feedback effect of the soundtrack. For the subtitler, it is a matter of deciding in each case what priority needs to be given to certain features of each sequence of speech. Linguistic constraints The space and time constraints inherent in the subtitling process usually enhance traditional translation challenges, such as grammar and word order, as well as problems related to cross-cultural shifts. With an average 30% to 40% expansion rate when translating from English into most other European languages, reduction is obviously the most important strategy in subtitling. But what, then, are the elements of speech that are commonly omitted or edited in subtitles? According to Kovač ič (1991: 409), there is a three-level hierarchy of discourse elements in subtitling: ● ● ● The indispensable elements (that must be translated). The partly dispensable elements (that can be condensed). The dispensable elements (that can be omitted). The indispensable elements are all the plot-carrying elements of a film; they carry the experiential meaning without which the viewers would Subtitling for the DVD Industry 27 not be able to follow the action. In the two examples below of the Dutch and German subtitles of Hitchcock’s Vertigo, only the plot-carrying elements have been translated. Example 4 Original dialogue 01.07.37.13 01.07.41.23 – Duration: 4.10 He had no problem getting Harry Cohn to agree a loan-out. Dutch subtitle English back-translation Hij wist Harry Cohn over te halen tot een ruil. He got Harry Cohn to agree a loan-out. Example 5 Original dialogue 01.08.12.06 01.08.17.03 – Duration: 4.22 Samuel Taylor was the one who introduced the character of Midge. German subtitle English back-translation Samuel Taylor führte die Figur der Midge ein. Samuel Taylor introduced the character of Midge. There are also a number of linguistic elements that many subtitlers in the profession would omit even if the spatio-temporal constraints of subtitling did not apply, such as: (a) (b) (c) (d) (e) Repetitions. Names in appellative constructions. False starts and ungrammatical constructions. Internationally known words, such as ‘yes’, ‘no’, ‘OK’. Expressions followed by gestures to denote salutation, politeness, affirmation, negation, surprise, telephone responses, etc. (f) Exclamations, such as ‘oh’, ‘ah’, ‘wow’ and the like. (g) Instances of phatic communion and ‘padding’, often empty of semantic load, their presence being mostly functional speech embellishment aimed at maintaining the desired speech-flow. Among these, we can find expressions such as ‘you know’, ‘well’, 28 Panayota Georgakopoulou ‘naturally’, ‘of course’, ‘understandably’; prepositional phrases (‘in view of the fact that’); rhetorical flourishes; and phrases used for sound effect (‘ways and means’). Many of these linguistic elements are commonly deleted because they can be retrieved from the soundtrack (b, c, d, and f). Were they to be transcribed or translated, we would have a case of duplication, as the same information would be found both in the subtitles and in the soundtrack. English words such as ‘yes’, ‘no’ and ‘OK’ are known virtually throughout the world. This has partly to do with the global position of the English language today and its prominence in the media industry. The degree to which certain words or expressions can be easily recognised by target viewers also has to do with how closely the two languages involved are related, usually English and the particular target language into which a programme is subtitled. Elements such as repetitions, padding expressions or even ungrammatical constructions may at times be optionally condensed rather than omitted, as they may contribute to the textuality of the programme and the character development of the actors. In the following example, we find repetitive items in the second part of this dialogue from Weird Science. The three questions ‘Didn’t throw up?’, ‘No?’, and ‘Nothing?’ could easily be omitted from the subtitles, as the meaning is still obvious from the translation of only the last of the questions, ‘You didn’t see anything?’. The purpose of the repetition, however, is to show the character’s exasperation and sense of urgency in getting an answer. Thus, some repetition is still built into the subtitles in the form of ‘No?’ or ‘Nothing?’, the two most concise of the three phrases, as limited time is available. In the subtitles aimed at deaf and hard-of-hearing viewers, yet further repetition is built in through the addition of the phrase ‘I didn’t throw up?’ and modification of the timings appropriately. Example 6 Original dialogue 01.27.07.22 – 01.27.12.06 Duration: 4.14 All right, in your dream, did I get up in the middle of the night and yak in your sink? 01.27.13.05 – 01.27.15.17 Duration: 2.12 Didn’t throw up? No? Nothing? You didn’t see anything? Continued Subtitling for the DVD Industry 29 Example 6 Continued Danish subtitles German subtitles I din drøm, stod jeg så op midt om natten og knækkede mig i vasken? Stand ich in deinem Traum mitten in der Nacht auf und kotzte ins Waschbecken? Nej? Så du ikke noget? Nein? Du hast nichts gesehen? Greek subtitles Hungarian subtitles '  u,                   ; Felkeltem az álmodban, hogy belerókázzak a mosdóba? ‘;    ; Semmi? Nem láttál semmit? Spanish subtitles Russian subtitles En tu sueño, ¿me levanté por la noche a vomitar en el lavabo?  òâîåì ñíå ÿ âñòàâàë ïîñðåäè íî÷è, ÷òîáû ïîáëåâàòü â ðàêîâèíó? ¿Nada? ¿No viste nada? Íåò? Òû íè÷åãî íå âèäåë? English back-translation English SDH subtitles In your dream, did I get up in the middle In your dream, did I get up in the middle of the night and yak in your sink? of the night and yak in your sink? No?/Nothing? You didn’t see anything? I didn’t throw up? No? You didn’t see anything? Some observations on translation strategies There are numerous constraints in subtitling, and there is no systematic recipe to be followed, no prêt-à-porter solutions. To decide on the best translation strategy, a thorough analysis of each translation issue has to be made based on: ● ● ● ● ● Function (relevance to the plot). Connotation (implied information, if applicable). Target audience’s assumed knowledge of the language and culture of the source language programme. Feedback effect. Media related constraints. 30 Panayota Georgakopoulou Reduction, which is the most important and frequently used strategy in subtitling, is generally applied on the basis of the three functions of language as suggested by Halliday (1973). Whereas experiential meaning needs to be translated, aspects of interpersonal and textual meaning can be omitted, especially when these may be retrieved directly from the picture or the soundtrack. The task of the subtitler involves constant decision-making to ensure that the audiovisual programme is not bereft of its style, personality, clarity, and that the rhythm and its dramatic progression not hindered. The final aim is to retain and reflect in the subtitles the equilibrium between the image, sound and text of the original. DVD subtitling With the appearance of DVDs on the market, developments in the subtitling industry during the last part of the twentieth century have posed a challenge, creating the demand for multilingual subtitling in a large number of languages with very tight turnaround times. The response from the market has come in the form of centralisation, whereby large subtitling companies offer services for all languages required, minimising the costs of Hollywood studios since there is no longer the need to send master copies all over the world. But how does centralisation work? And what are the advantages and problems? The template In the past, the two main problems that subtitling companies had to address were: (a) The cost of subtitle production: in order to make their services attractive to large Hollywood production companies, prices needed to be controlled. (b) The management of subtitle files: numerous files are involved and their management had to be made easier in order to speed up the process and minimise the number of mistakes that might be made. The solution to both these problems has come in the form of universal template subtitle files in English (also referred to in the profession as the genesis file or the transfile), to be used as the basis for translation into all languages. In other words, the subtitling process has been split into two distinct tasks. The timing of a film or audiovisual programme is Subtitling for the DVD Industry 31 made by English native speakers who produce a unique timed subtitle file in English, that is, a file where all the in and out times have been decided. This file is then used as the basis for translation into other languages as required, with the translation carried out by native speakers of these languages. The advantages of this new working method are obvious. The costs involved in the timing of a film are reduced to producing one subtitle file only, rather than a different one for each of the languages required. To put it in financial terms, what the template has introduced into the subtitling industry is an economy of scale, whereby the greater the number of languages involved in a project, the larger the cost-savings to be made. From a linguistic point of view, certain problems can also be eliminated. When spotting and subtitling are carried out directly by a native speaker living in the target country and whose native language is not English, mis-hearings of the language of the original soundtrack, which is usually English, tend to be more frequent. When the mother tongue of the subtitler is not English, there can be potential problems in understanding the source language, especially slang and colloquialisms, which require an affinity with the spoken language that can only really be acquired by living in the country where the language is spoken. But if the subtitler used is one who has been living in an English-speaking country for some time, with near-native fluency of English, s/he might run the risk of not being 100% up-to-date with the changes in his/her own native tongue, languages being living organisms in constant change. The use of a template file makes it possible, therefore, to gain the best of both worlds: English native speakers are used to produce the original file in English, and the native speakers of the languages required in each project are then used to translate into their own mother tongues. Above all, the recruitment needs for subtitling companies catering for many languages can now be dealt with instantly as the pool of freelancers – in all the languages with which a company works – has similarly been expanded. It is no longer necessary to recruit subtitlers in all the languages required who were also able to do the technical task of spotting, a rare type of professional that, for some languages, may be hard to find. Translators with no subtitling training can be given a chance. As the issue of timing and cueing is removed from the equation, translators without subtitling experience may be given training in recognising mistakes and good practice in subtitling and in developing respect for its principles. Timing issues that emerge 32 Panayota Georgakopoulou can be reported and dealt with by the project managers, leaving the translators to concentrate on the translation of the dialogue. As all the subtitle files required for any given project or film are based on a single template in English, the resulting target files in all languages have now become identical in terms of timings and subtitle number. This facilitates their management and the number of checks they have to undergo before they are ready for DVD emulation. And this last part of the process does not require specific linguistic skills, other than knowledge of English, from the person performing checks, yet another cost-saving factor. The problems of centralisation When catering for worldwide subtitling needs and styles, the complexities of subtitling described above multiply. Decisions relating to issues such as when to use italics in the subtitles or whether to give priority to horizontal versus vertical subtitling (as in Japanese) are country-specific. The question has been raised as to how these different conventions can possibly be tackled within a single subtitle file that has to conform to each country’s conventions, while ensuring that the end-result is readily acceptable in the countries where the finished product (that is the DVD) is to be sold. These requirements clearly have to be kept in mind when creating a template file, making the effort required on the part of the subtitler greater. But the main principle that underlies the subtitling of a template file is much the same as the subtitling of a film or any other audiovisual programme into any language, there is one constant that never changes, which is the source file. The main difference between subtitling and other forms of translation is the fact that the translated version of the text does not replace the original: in the subtitled version both are present concurrently. In other words, what for subtitlers is one of the major translation challenges in their work – the original always remains present alongside their translation, limiting their choices and putting their solutions as the focus of criticism of audiences worldwide – can now be used as an advantage. It is true that over the years different countries have developed their own subtitling styles, a fact probably more true of, for example, Scandinavian countries with their longstanding tradition in subtitling, in contrast to primarily dubbing countries, such as Italy or Spain, where subtitling is more of a novelty. But as regards the first part of the process, that is the timing of a subtitle file, the choices are limited, as the Subtitling for the DVD Industry 33 actors on the screen will always open their mouths to utter a phrase at a certain point in a film or programme and close it at another point, and the timings of any subtitle file would need to reflect this. This may explain why the process of templating is not new to Scandinavian countries, where it was used long before the appearance of DVD, with companies creating subtitle files for VHS releases for all Scandinavian countries out of a single office (Georgakopoulou, 2003: 24). If template files are to be used successfully in the profession, greater effort needs to go into their creation, and more information included. At the European Captioning Institute, a template file always contains information such as: (a) Translation notes for unfamiliar or culturally-bound expressions, meant to help non-English native speakers produce a more accurate translation. Cases of irony, wordplay and related aspects are also brought to the attention of subtitlers. (b) Notes as to the points of the film or programme in which burnt-in text appears, that is written inserts on screen, in order to cater for information that, in certain languages, may or may not need to be translated. These notes help assure accurate positioning of the subtitles on screen, ensuring that the burnt-in text on the image is not obscured by the subtitles. (c) Further notes as to the choice of font, the use of italics, the treatment of songs, the treatment of titles, etc., in each language file, depending on the conventions of each language and/or the demands of the client. On the whole, the creation of a template file is far more complicated than the creation of any other subtitle file and the process involves an even greater amount of work because of the number of ancillary files produced. However, the rules and lists of notes necessary for the creation of template files are also the means whereby, at a later stage of the subsequent language files, the process of checking is facilitated to ensure that they all adhere to the conventions necessary for acceptance in the target country and clients’ specifications. Conclusion The advantages of template files and the flexibility that they offer to subtitling vendors cannot be overestimated; they form the basis of the 34 Panayota Georgakopoulou DVD subtitling industry. As this is a relatively young industry, it goes without saying that the underlying rules and additional constraints for creating template files make this subject wide open to further research on a global scale in order to refine current practices and produce even better quality subtitle files. Such research would inevitably require practitioners from all over the world to play a central role in order to ensure that the language files produced will conform to the appropriate subtitling style of each country. A prominent Greek subtitler, with over 40 years’ experience in the field, told me in an interview that a good subtitler can edit anything, even a book, down to one subtitle. In line with my true passion for the job as subtitler, and echoing the statement that ‘[t]he attempt to achieve perfect subtitling has some affinity to the search for the Holy Grail’ (Baker et al., 1984: 6), I cannot resist the temptation of editing down this chapter into a single subtitle: The template IS the Holy Grail of the DVD subtitling industry today. (Duration: 04.15) Note 1. Adult reading speed is calculated at ECI at 750 characters per minute, or approximately 180 words per minute for non double-byte languages, whereas children’s reading speed is set at considerably less and hovers between 120 and 140 words per minute. References Baker, R.G., Lambourne, A.D. and Rowston, G. (1984) Handbook for Television Subtitles. England: IBA. Díaz Cintas, J. and Remael, A. (2007) Audiovisual Translation: Subtitling. Manchester: St Jerome. Georgakopoulou, P. (2003) Redundancy Levels in Subtitling. DVD Subtitling: A Compromise of Trends. Guildford: University of Surrey. PhD Dissertation. Gottlieb, H. (1994) ‘Subtitling: People translating people’. In C. Dollerup and A. Lindegaard (eds) Teaching Translation and Interpreting 2: Insights, Aims, Visions (pp. 261–74). Amsterdam and Philadelphia: John Benjamins. Halliday, M.A.K. (1973) Explorations in the Functions of Language. London: Edward Arnold. Kovačič, I. (1991) ‘Subtitling and contemporary linguistic theories’. In M. Jovanovič (ed.) Translation, A Creative Profession: Proceedings XIIth World Congress of FIT – Belgrade 1990 (pp. 407–17). Beograd: Prevodilac. Subtitling for the DVD Industry 35 Luyken, G. M., Herbst, T. Langham-Brown, J., Reid, H. and Spinhof, H. (1991) Overcoming Language Barriers in Television: Dubbing and Subtitling for the European Audience. Manchester: European Institute for the Media. Mason, I. (1989) ‘Speaker meaning and reader meaning: Preserving coherence in screen translating’. In R. Kölmer and J. Payne (eds) Babel: The Cultural and Linguistic Barriers Between Nations (pp. 13–24). Aberdeen: Aberdeen University Press. Nedergaard-Larsen, B. (1993) ‘Culture-bound problems in subtitling’. Perspectives: Studies in Translatology 2: 207–41. Nir, R. (1984) ‘Linguistic and sociolinguistic problems in the translation of imported TV films in Israel’. The International Journal of the Sociology of Language 48: 81–97. 3 Subtitling Norms in Greece and Spain Stavroula Sokoli Introduction By simply watching subtitled films in Greece, traditionally considered a subtitling country, and Spain, a dubbing country, certain differences in this professional practice of audiovisual translation may easily be observed. In this chapter the aim is two–fold:first to look into the factors that guide translators’ decisions in the process of subtitling in these two countries, by finding indications of norms; second, closely related to the first, is to explore the relationship between the subtitles and the other elements of the audiovisual text. The main premises are that the use of descriptive tools can overcome the limitations posed by prescriptive approaches and that there are regularities in the practice of subtitling. Norm theory has contributed in a major way to the evolution of Descriptive Translation Studies by introducing an evaluative element. Indeed, the mere description of translation behaviour for its own sake may not provide useful results, whereas the study of norms is bound to give insight into the intersubjective sense of what is ‘proper’, ‘correct’, or ‘appropriate’, in other words, the content of norms (Hermans, 1999: 82). According to Toury (1995: 54), translators have to acquire a certain set of norms which will lead them towards adopting a suitable behaviour and help them manoeuvre among all the factors which may constrain it. In the case of subtitling, these factors concern mainly time and space constraints, which, until recently, have attracted primary attention in the discussions about subtitling and AVT. Here it is argued that a shift from a discussion of the constraints themselves to the factors that guide the translators in their work may prove useful. Since norms are not directly observable, a possible approach is a study of their manifestation, whether textual or extra-textual – two major 36 Subtitling Norms in Greece and Spain 37 sources of reconstruction of translational norms suggested by Toury (1995). The focus here is on textual sources of norms, that is, regularities in the choices made by subtitlers, as manifested in translated films. In addition, the findings will be compared to the ones encountered in extra-textual sources. The theoretical starting point of the discussion is based on the categorisations of norms made by Toury (1995) – initial, preliminary and operational (subdivided into matricial and textual) norms – and by Chesterman (1997) – expectancy and professional norms (subdivided into the accountability, the communication and the relation norms). The reason for the choice of both categorisations is that they cover the same areas but from a different perspective. In this chapter the focus is on matricial and relation norms. The former affect the textual segmentation of the linguistic material and its distribution, or, expressed in the terminology of the practice of subtitling, the spotting of the original script, that is its division into ‘chunks’ to be translated, and the cueing of the subtitles, that is the designation of their in and out times. Relation norms on the other hand stipulate that an appropriate relation of relevant similarity should be established and maintained between the source and the target text, where ‘equivalence’ or ‘optimal similarity’ is only one of the possible kinds of relation. Other parameters covered include the addition or omission of information, as well as the relation to accompanying channels, for example, synchronisation between speech and the emergence of subtitles on screen. Another aspect to be discussed concerns expectancy norms. Adapting Chesterman’s (1997) definition to the case of subtitling, they reflect the expectations that viewers of subtitled audiovisual programmes have with regard to what the subtitled product should be like. They are formed by the prevalent subtitling tradition in the target culture and by the previous viewing of subtitled films. In this sense, it can be argued that the expectation norms in Spain and Greece are bound to differ, since the first country is characterised by a tradition in dubbing and not subtitling. The nature of audiovisual texts Before proceeding to an analysis of the texts, it may be useful to examine the nature of the audiovisual text, and the features that distinguish it from other kinds of text.1 Characterised by its reception through two channels, the acoustic and the visual, its other distinctive feature is the importance of the nonverbal element. As pointed out by Zabalbeascoa 38 Stavroula Sokoli (1997) all texts contain some nonverbal elements, since the message cannot be delivered without some sort of physical support. In a film, however, the nonverbal elements, acoustic in the form of noises and music, or visual such as images, appear to a much greater extent than in written texts. Still, not all texts containing these characteristics are under discussion here, and more parameters have to be considered in order for any study to be defined with precision. One of these parameters is the medium: audiovisual texts appear on a screen whether big or small. The fact that nowadays films can be viewed not only at the cinema or on a television set but also on a computer screen, calls for another delimiting parameter which will define audiovisual texts as opposed to hypertexts, received through the same medium. The images in a hypertext, such as a web page, can be static or moving, whereas the audiovisual text always includes moving images. The difference between the two kinds of text when they include moving images is that the latter contains a predetermined succession of non-repetitive images in absolute synchronisation with the verbal elements. Another differentiating factor is interactivity: in the case of hypertexts the receivers decide the sequence of the elements, according to their needs, whereas the audiovisual text cannot be altered. The only possibility of intervention on the part of the receiver is the case of a videotape or a DVD where the viewer can backtrack or ‘move’ within the film. The features that distinguish the audiovisual text can be recapitulated as follows: ● ● ● ● ● Reception through two channels: acoustic and visual. Significant presence of nonverbal elements. Synchronisation between verbal and nonverbal elements. Appearance on screen – reproducable material. Predetermined succession of moving images – recorded material. These features condition the translation of the audiovisual text, and, as a result, their consideration is fundamental for its study. The combination of the acoustic and the visual channels, together with the verbal and the nonverbal elements, result in four basic elements of the audiovisual text: the acoustic verbal (dialogue), the acoustic nonverbal (score, sounds), the visual nonverbal (image) and the visual verbal element (subtitles).2 In this light, and for the purposes of the present analysis, the set of subtitles is viewed as forming part of the translated text, not as constituting a translation product by itself. In Subtitling Norms in Greece and Spain 39 Verbal Acoustic (dialogue) Visual Verbal (subtitles) Nonverbal Acoustic (music, sounds) Visual Nonverbal (image) Figure 3.1 Spatial and temporal relationships between the basic elements of the subtitled audiovisual text other words, the original, untranslated film is considered to be the source text and the subtitled film the target text. The spatio-temporal relationships between the four above-mentioned elements may be seen in the Figure 3.1, where the solid arrows represent existing relationships in an audiovisual text and the dashed arrows represent the relationships established by the subtitler.3 In the source text, the temporal relationship that exists between the images, the sounds and the dialogue is characterised by an inherent synchronisation. Logically, the same quality is bound to be required between the subtitles and the rest of the elements of the target text; subtitles will be expected to appear more or less when a character begins to speak and to disappear when s/he stops. In search of norms The texts under study comprise the subtitled versions in Spanish and Greek of the films The English Patient,4 the 162-minute film directed by Anthony Minghella in 1996, and Notting Hill,5 the 124-minute comedy directed by Roger Michell in 1999. A first quantitative and comparative analysis between the two sets of subtitles shows a significant difference in the number of subtitles, as shown in Table 3.1. Even though the above are the only films analysed in this chapter, there is further evidence regarding the higher number of subtitles in the Spanish versions compared to the Greek versions, as can be seen in the analysis of the movies The Perfect Storm,6 the 129-minute adventure film 40 Stavroula Sokoli Table 3.1 Comparison between the numbers of subtitles in the films The English Patient and Notting Hill Greek sub/s The English Patient Notting Hill Spanish sub/s Additional sub/s in Spanish 955 1358 42.1% 1052 1754 66.7% Table 3.2 Comparison between the numbers of subtitles in The Perfect Storm and Manhattan Murder Mystery The Perfect Storm Manhattan Murder Mystery Greek sub/s Spanish sub/s 915 1449 1396 1944 Additional sub/s in Spanish 52.5 % 34.1 % directed by Wolfgang Petersen in 2000, and Manhattan Murder Mystery,7 the 104 minute comedy directed by Woody Allen in 1993. This difference in number of subtitles can be attributed to two factors: a. Difference in dialogue omission: where there is omission in the translation of the acoustic verbal element in Greek there are subtitles in Spanish. b. Difference in distribution: one Greek subtitle consisting of two lines corresponds to two Spanish subtitles consisting of one line each. The question to be answered at this point is: why are there 403 ‘additional’ subtitles in the Spanish version of The English Patient and 702 in Notting Hill? In the first film, 48.9% (197 subtitles) of this difference is due to the first factor mentioned above, that is, these subtitles are present in Spanish but omitted in Greek. As for the remaining 51.1% of the difference (206 subtitles), it is caused by the second factor, that is, by the fact that two Spanish one-liners correspond to one Greek twoliner. Accordingly, in Notting Hill, 64.9% (456 subtitles) of the difference is attributed to the first factor and the rest 35.1% (246 subtitles) to the second factor. Indeed, the analysis of the number of subtitles consisting of one or two lines shows that there is a preference for two-liners in the Greek and for one-liners in the Spanish versions, as shown in Figure 3.2. Subtitling Norms in Greece and Spain The English Patient 41 Notting Hill 32% 33% Greek subtitles 68% 67% one-liners two-liners 37% 41% Spanish subtitles 63% Figure 3.2 59% Percentage of subtitles consisting of one-liners and two-liners The next step in the analysis is to find out the factors that determine the choices related to the distribution and the omission of subtitles in each language. Let us look at a couple of examples where a two-liner in Greek corresponds to two one-liners in Spanish. The first one is from The English Patient: Example 1 Somos vecinos. Vivimos ... a 2 manzanas de Montreal. Greek Spanish Caravaggio: Apparently, we’re neighbours. My house is two blocks from yours in Montreal.     .       . It can be observed that in the Spanish version, a second subtitle is introduced, even though both lines could fit in one subtitle. Assuming that this choice is not random, what seems to determine the specific decision by the subtitler is the existence of a cut, that is, a sudden change from one image (a close-up of Hanna) to another (showing both Hanna and Caravaggio). This norm operating in Spanish can also be found in extra-textual sources, more specifically in articles written by Spanish professionals like Castro Roig (2001: 280) who states that ‘whenever there is a cut, there must be a new subtitle’ (my translation), 42 Stavroula Sokoli a normative statement that is repeated by Leboreiro Enríquez and Poza Yagüe (2001). This is also the case in the example below from Notting Hill, where the second speaker follows almost immediately after the first one. In Greek there is one subtitle consisting of two lines, whereas the Spanish subtitler has chosen to create two separate subtitles in order to respect the visual nonverbal element, that is, the shot change: Example 2 Vale, se acabó. No, no vale la pena. Greek Spanish Will: Right, that’s it. Sorry. Anna: No, there’s really no point. -     . - ‘.      In many cases, the difference in lexical distribution is determined by the influence (or absence of influence) of the acoustic nonverbal elements, namely pauses in the characters’ speech. This can be seen in the following example from Notting Hill where the dialogue is characterised not only by many pauses in speech, but also by a high number of false starts: Example 3 Anna: Tempting, but ... no. Thank you. ... pero ... ... no. Greek Spanish Tentador, ...     ... ... ‘,  . Gracias. In the above example, we can see that there are four Spanish subtitles when the Greek subtitler has decided to use only one. The distribution of subtitles in Spanish seems to be determined by the pauses in Anna’s speech, whereas in Greek it appears to be the direct consequence of the norm which stipulates that, where possible, each subtitle must have a complete meaning in itself. The extra-textual source for this norm consists of a questionnaire directed to Greek professional subtitlers (Sokoli, 2000). One of the outcomes of the analysis of the questionnaires Subtitling Norms in Greece and Spain 43 indicates that completeness of meaning in each subtitle is considered one of the main characteristics of good subtitling. A technique often used in order to indicate the existence of pauses is the use of triple dots, as in the above example. An observation that has to be made concerning the Spanish versions is that there are quite a few examples where there is both a pause and a cut and the subtitler has to choose between cueing the subtitle according to one of the two. It has been found that in all these cases, priority is given to the norm which requires synchronisation with the acoustic element; in other words, the cueing in of the subtitle will be done when the next part of the utterance begins and not when there is a cut. The next step in the analysis is to examine the difference in the choice of omissions, more specifically, the cases where an acoustic verbal element has an equivalent subtitle in the Spanish but no subtitle in the Greek version. The following example is from Notting Hill: Example 4 Hola. - Hola. - Desapareciste. Sí, sí. Tuve que irme, no quería molestar. Greek Spanish Anna: Hi. William: Hello. Anna: You disappeared. William: Yeah ... yeah. I had to leave. I didn’t want to disturb you.    .   . ‘      '   . In the Greek version, even though there are no time or space constraints, the greeting between the two characters and the almost inaudible ‘yeah ... yeah’ are omitted in the subtitles. These choices do not seem to be random and the norm which possibly governs them can be found in the results of the previously mentioned questionnaire: a high percentage of the Greek subtitlers involved stated that they omit utterances that they consider either easily recognisable by the Greek audience, (such as ‘OK’, ‘hello’, ‘yes’, ‘no’, etc., names, repetitions), or not relevant to the plot, (hospital or airport announcements, 44 Stavroula Sokoli songs, etc.). Moreover, it was pointed out that, despite the absence of time or space constraints, such utterances are often omitted in order for the viewers to have time to enjoy the image. These are utterances which can be recovered by other elements of the audiovisual text: ● ● ● ● ● The acoustic verbal (recognisable utterances, names, etc.). The acoustic nonverbal (phatic elements, exclamations, etc.). The visual verbal (repetition found in other subtitles, etc.). The visual nonverbal (objects in the image, etc.). A combination of all or some of the above elements. It is clearly difficult, if not impossible, to mark a clear-cut distinction between the elements of the audiovisual text, since these are closely interrelated. The argument here is that the omitted utterances are mainly and not only recoverable from these elements. Another example comes from The English Patient: Example 5 ¡Estamos aquí! !! ¡Aquí! - ¡Pare!     ! ¡Madox! Greek Spanish Katharine: Stop! Here! Over here! Stop! Madox! Almasy: Madox! Madox! - ¡Madox! - ¡Madox! - In the Spanish version, there seems to be a norm stipulating that there must be as few omissions as possible, which is also extra-textually corroborated by Díaz Cintas (1997: 281). In an analysis of the subtitled version of Manhattan Murder Mystery in Spanish he finds that there is a tendency for what he calls sobretraducción [over-translation]. According to this scholar, the specific phenomenon is explained by ‘the intention that the viewer can have the feeling of not being cheated, and of having all the information contained in the original version’ (my translation). He provides examples similar to the ones above – such as Sí, sí, ¿Helen?, ¡Jack! – and considers these subtitles unnecessary for the comprehension of the plot since they are of a purely phatic or vocative Subtitling Norms in Greece and Spain Table 3.3 45 The 13 possible relationships (Allen, 1983) Relation Symbol Symbol for inverse Pictorial example x before y # $ xxxx x equal y " " xxxx yyyy x meets y m mi xxxxyyyy x overlaps y o oi xxxx yyyy x during y d di xxxx yyyyyyyy x starts y s si xxxx yyyyyy x finishes y f fi xxxx yyyyyy yyyy nature. Moreover, according to the answers given by Spanish subtitlers to the questionnaire discussed above, the spectator must not be left to feel that there is any missing information, and consequently a subtitle must appear every time an utterance is heard. In order to study the relation between the acoustic and the visual verbal elements, I have borrowed a relevant categorisation from the discipline of Artificial Intelligence, relating to the basic possible relationships between two intervals. According to Allen (1983) there are 13 such relationships, as indicated in Table 3.3. Let us now see which are the most frequent temporal relationships between the utterances heard and the subtitles. In the Greek versions of the films under analysis it has been found that the subtitles appear some deciseconds (one-tenths of second) after the start of the utterance. In the Spanish versions, on the other hand, the subtitles appear exactly when the utterance begins and finish either when the utterance finishes or some deciseconds after it has finished. If Allen’s categorisation is used, where x represents the duration of the utterance and y the duration of the subtitle, we have the following most frequent temporal relations between the two: ● ● Greek: ‘x overlaps y’ (o) Spanish: ‘x equal y’ (") and ‘x starts y’ (s) As far as the Spanish versions are concerned, the choice for the cueing of the subtitles seems to be governed by the requirement for 46 Stavroula Sokoli synchronisation. Naturally, even though it is almost always possible to cue the subtitle in at the beginning of the utterance, the same does not hold for the cueing out, because the specific duration might not be enough for the subtitle to be read. This may explain why both relationships are frequent in Spanish. A similar assumption of the norm cannot be made for the Greek versions. This is because I have not encountered any extra-textual sources verifying the existence of a norm requiring that subtitles must be inserted some frames after the utterance has begun, despite the fact that the specific relationship is the one most frequently met. The regularity of this relationship could be attributed to certain technical limitations that existed in the past, such as the necessity for manual insertion of subtitles (Sokoli, 2000). Despite the fact that these technical limitations have now been overcome, the lack of a norm requiring absolute synchronisation seems to have led to a confirmation of this practice. Thus, it could possibly be included among the cases that Chesterman (1993: 4) describes as ‘behavioural regularities [which] are accepted (in a given community) as being models or standards of desired behaviour’. Conclusion The above quantitative and qualitative analysis shows that there are indications of the presence of norms in the subtitling practices in Greece, as well as Spain. In particular, there are indications of the operation of the following norms: ● ● Matricial norms: in Spain, the distribution of subtitles is determined by the acoustic nonverbal (pauses) and the visual nonverbal (cuts) elements. When there is a conflict between the requirements for synchronisation with the acoustic and the synchronisation with the visual element, synchronisation with the acoustic element prevails. In Greece, the distribution of subtitles is determined by the requirement for completeness of meaning within the same subtitle and the preference for subtitles consisting of two lines rather than one. Relation norms: in Spain, the choice dictating omission is determined by the requirement of ‘equality’ between the acoustic verbal and the visual verbal element. As a result, the omitted utterances are as few as possible. Moreover, there is an effort to avoid disproportion between the duration of the dialogue and the subtitles. In Subtitling Norms in Greece and Spain 47 Greece, the choice for omission is determined by the utterance’s recoverability from the other elements of the audiovisual text. Recoverable elements are often omitted even in the absence of time and space constraints. Again, it has to be stressed that these are only indications of the presence of norms and that more films have to be analysed in order for them to be verified. Future research could include the comparison of each set of subtitles with the original dialogue, with the aim to find the degree of language compression, since the only comparison made in the present study has been between the two subtitled versions of each film. Moreover, other films to be studied could belong to different genres (for example action films, documentaries) or have stricter time and space constraints (for example films with fast dialogue), with the aim to explore the way the above norms are adapted. Another possible line of research involves investigation of the reception of subtitled audiovisual programmes. For example, a subtitled film translated according to the Spanish norms could be shown to a Greek audience, or vice versa, and their response could be examined in order for expectancy norms to be discovered. Given the qualitative and descriptive nature of the present discussion in the absence of laws or absolute truths, possible explanations have been proposed and more hypotheses have been put forward, which in turn may serve as a basis for future research. The results could be used not only for the explanation and prediction of the way subtitles are manifested, but also in a programme of training for subtitlers. Notes 1. The definition of text used here is the one established by de Beaugrande and Dressler (1981), according to which a text has to meet seven standards of textuality: cohesion, coherence, intentionality, acceptability, informativity, situationality and intertextuality. Hence, texts are not only written utterances, in spite of the connotation this word has in everyday language. This definition also accounts for spoken utterances, as well as for television and cinema programmes, our present object of study. 2. Delabastita (1990: 101–2) refers to these as ‘four types of film sign: verbal signs transmitted acoustically (dialogue), nonverbal signs transmitted acoustically (background noise, music), verbal signs transmitted visually (credits, letters, documents shown on the screen), nonverbal signs transmitted visually’. 3. The term ‘subtitler’, as used here, does not necessarily refer to one person but may represent the team of people involved in the subtitling process. 48 4. 5. 6. 7. Stavroula Sokoli Depending on the practice, the same person is responsible sometimes for all the steps, whereas in other cases, the spotting and the translation are done by different people (see Georgakopoulou, this volume). The set of Greek subtitles broadcast on TV is the same as the one appearing in the VHS version (no information was available about the cinema or DVD version). As for the Spanish version, only the VHS was available. In Greece, this film had the same set of subtitles for TV, VHS and DVD (no information was available about the cinema version). In Spain, interestingly enough, there are four different sets of subtitles for each medium and the one used for this analysis is the TV version. The TV version has been used for both languages. The VHS version has been used for both languages. See Díaz Cintas (1997) for more details about the subtitling of this film into Spanish. References Allen, J. (1983) ‘Maintaining knowledge about temporal intervals’. Communications of the ACM 26(11): 832–43. de Beaugrande, R. and Dressler, W. (1981) Introduction to Text Linguistics. London: Longman. Castro Roig, X. (2001) ‘El traductor de películas’. In M. Duro (ed.) La traducción para el doblaje y la subtitulación (pp. 267–98). Madrid: Cátedra. Chesterman, A. (1993) ‘From “is” to “ought”: Laws, norms and strategies in Translation Studies’. Target 5(1): 1–20. Chesterman, A. (1997) Memes of Translation. The Spread of Ideas in Translation Theory. Amsterdam and Philadelphia: John Benjamins. Delabastita, D. (1990) ‘Translation and the mass media’. In S. Bassnett and A. Lefevere (eds) Translation, History and Culture (pp. 97–109). London: Pinter. Díaz Cintas, J. (1997) El subtitulado en tanto que modalidad de traducción fílmica dentro del marco teórico de los Estudios sobre Traducción (Misterioso asesinato en Manhattan, Woody Allen, 1993). Valencia: Universidad de Valencia. PhD Thesis. Hermans, T. (1999) Translation in Systems: Descriptive and Systemic Approaches Explained. Manchester: St Jerome. Leboreiro Enríquez, F. and Poza Yagüe, J. (2001) ‘Subtitular: toda una ciencia ... y todo un arte’. In M. Duro (ed.) La traducción para el doblaje y la subtitulación (pp. 315–23). Madrid: Cátedra. Sokoli, S. (2000) Research Issues in Audiovisual Translation: Aspects of Subtitling in Greece. Barcelona: Universitat Autònoma de Barcelona. MA Dissertation. Toury, G. (1995) Descriptive Translation Studies and Beyond. Amsterdam and Philadelphia: John Benjamins. Zabalbeascoa, P. (1997) ‘Dubbing and the nonverbal dimension of translation’. In F. Poyatos (ed.) Nonverbal Communication and Translation (pp. 327–42). Amsterdam and Philadelphia: John Benjamins. 4 Amateur Subtitling on the Internet Łukasz Bogucki Introduction The traditional understanding of the audiovisual translation mode known as subtitling is that it is intended primarily for cinema and television use, with the help of a visual component in the form of a (video) recording and the final programme script of the original, except perhaps for instantaneous, live subtitles (Kilborn, 1993). The ubiquity of the Internet, however, has given rise to a new kind of AVT which I refer to here as ‘amateur subtitling’. Amateur subtitling is not unrelated to fansubs (www.fansubs.net/fsw/general), subtitles of various Japanese anime productions made unofficially by fans for non-Japanese viewers. Despite their dubious legal status, fansubs have been in existence since the late 1980s (O’Hagan, 2003). The rationale behind the decision to undertake the translation in the form of fansubs and amateur subtitling is largely the same: to make a contribution in an area of particular interest and to popularise it in other countries, making it accessible to a broader range of viewers/readers, who belong to different linguistic communities. Amateur subtitle producers, a term used here in place of ‘translator’ or ‘subtitler’ as the product under discussion does not qualify as fullyfledged subtitling, typically work with a recording of the original, but they have no access to the post-production script. They rarely work with classics, as the intention behind their work is to make local viewers familiar with recent film productions. The quality of their product is thus conditioned by how much they understand of the original. Comprehension is essential in any kind of translation, yet understanding the original for amateur subtitling purposes differs from the kind of source text analysis performed in ‘paper’ translation. The common 49 50 Łukasz Bogucki ground is that the original can be listened to as many times as the subtitle producer deems necessary, but certain fragments may be unclear, made unintelligible by background noises, or otherwise irretrievable as in, for example, interpreting. Amateur subtitlers have at their disposal a range of freeware or shareware computer programs to create subtitles (e.g. Sub Station Alpha, www.topdownloads.net/software/view.php?id=9740) and to superimpose them on the film (e.g. Virtual Dub, www.virtualdub.org). Inserting subtitles in the film or programme is very easy even for relatively inexperienced users. In amateur subtitling there is no strict limit as to the number of lines per subtitle, but experienced subtitle producers realise that human perception is not boundless; as cinema-viewers themselves, they subconsciously tend to apply the conventional limit of two lines per subtitle. Three-line subtitles are used in rare cases where the pace of the dialogue is so fast that two-liners would have to be displayed very quickly one after the other. Nor are there any strict restrictions as regards the number of characters per line.1 There are programs that trim the number of characters per line to a certain limit, but we have yet to see a standard emerge.2 As font size is a user-defined variable, longer subtitles sometimes occupy too much screen space if large fonts are selected. As a result, subtitle producers will adjust their work to the individual requirements of a given computer application. Amateur subtitling is also subject to technical constraints, but it has to be stressed that the limitations at work as regards unprofessional amateur subtitling are unlike the ones that traditionally apply to professional subtitling (see Karamitroglou, 1998 for a description of subtitling standards). There is far more freedom and, as a consequence, technical constraints do not normally impinge on the resulting product to the extent that strategies of curtailment such as decimation, condensation, etc. (Gottlieb, 1992; Bogucki, 2004) have to be resorted to. The problem with amateur subtitling lies not so much in squeezing the gist of what the original characters say into 30 or so characters per line to enable the audience to appreciate the filmic message without too much effort; the problem, it seems, lies mostly in the quality of the source material and the competence and expertise of the translator. Text files with amateur subtitles for many recent cinema releases are relatively easy to find on the Internet. 3 The examples in this section come from an amateur translation made by Mirosław Jaworski, with permission from the subtitle producer. The film under discussion is Peter Jackson’s The Fellowship of the Ring (2001). Jaworski made great efforts to become the first amateur to subtitle this particular release. Amateur Subtitling on the Internet 51 He posted his translation in the form of a text file on the WWW about six weeks before the film was actually shown in cinemas in Poland, and approximately at the same time as it was released in the USA. Not only was he pressed for time not to be outdone, but also his ‘source text’ (the inverted commas being very much in place here) was what he had managed to take down from a camcorder recording of the original cinema release. As a result, he had to cope with very poor quality and the translation required the meticulous effort of listening to the original scores of times. He was left to his own devices, as most amateur subtitle producers are, and had to rely on his own language competence. Let us now look at some illustrative excerpts to see how comprehension problems can impinge negatively on translation quality. Error analysis As we shall see presently, the term ‘error analysis’, avoided wherever it can be helped in literature (quality assessment, after all, is not merely about error-hunting), can be used to discuss amateur subtitling quality. (A) ST: You know Bilbo – he’s got the whole place in an uproar. TT: -Znasz Bilba, całe miejsce zamienił w altan̜. [You know Bilbo. He’s turned the whole place into an arbour.] The visual image in this instance is a rural, peaceful landscape. (B) ST: All I did was give your uncle a little nudge out of the door. TT: Wszystko, co zrobiłem, to dałem twojemu wujowi par̜ porad u drzwi. [All I did was give your uncle some advice at the door.] The problematic words in these two examples are ‘uproar’ and ‘nudge’. Since, from the point of view of English learners, and Jaworski is a confessed self-learner, these terms are normally learned at later stages of linguistic development, they may have been new to the subtitle producer. Had he had the opportunity of seeing them in writing, he would no doubt have looked them up and may have come up with a more accurate version. Let us look at another example: (C) ST: The Shadow does not hold sway yet. TT: Cié nie ustabilizował si̜ jeszcze. [The Shadow has not set in yet.] 52 Łukasz Bogucki The expression ‘to hold sway’ may be said to have similar features to ‘nudge’ or ‘uproar’ in that it is likely to be unfamiliar to a speaker of English as a second language with a limited lexicon; these words do not typically occur in textbooks for beginners or intermediate students of English. However, in this case it can actually be seen rather than heard, as the original sentence appears as an on-screen insert in Elvish, and the example quoted above is the translation into English that appears as a subtitle. In all probability, the subtitle producer looked up the unfamiliar word in a dictionary and was able to offer an accurate translation. Another type of error results from inadequate linguistic competence coupled with the specificity of cross-medium translation, that is, having to go from the spoken of the original to the written of the translation: (D) ST: ... friends of old. TT: ... przyjaciele Starych. [ ... friends of (the) Old.] This short excerpt comes from the Council of Elrond where noble representatives of different races meet to discuss a serious issue. The original is merely a polite form of address suggesting that the characters have known each other for some time. The Polish translation may be implying that the speaker is addressing the friends of the old race of Elves. The mistake might never have been made but for the change in medium from spoken to written. If the subtitle producer had seen the dialogue in writing, he would have noticed the lower case where he was expecting a capital letter, and would have thought twice about the appropriateness of his rendition. Similarly: (E) ST: ... festering, stinking marshlands as far as the eye can see. TT: -Obrzydliwe, ́mierdz̜ce ... Tam, dok̜d si̜ga Oko! [ ... disgusting, stinking ... Where the Eye reaches!] ‘The Eye’ is used throughout the film as a metonym for Sauron, the Great Enemy. It would make sense in the context to interpret this fragment as ‘as far as the Eye can see’, as Sauron could scan the territory around his tower, looking for enemies. Again, had the subtitle producer had access to the written script, he might not have made the mistake. Amateur Subtitling on the Internet 53 In (F) the producer may have heard only the key word (‘arrangements’) and produced a subtitle where the message is not entirely off the mark, but the version nevertheless remains little more than a lexical addition in terms of translation. (F) ST: All the arrangements are made. TT: Tylko kilka umówionych spotká. [Just a few appointments.] In natural speech the beginnings of sentences are often unstressed, sometimes hardly heard. In the following example, had the subtitle producer heard the subject of the sentence and not, apparently, mistaken ‘he’s’ for ‘it’s’, the subtitle might have been nearer the mark. As it is, the meaning is completely changed: (G) ST: (He’s) very fond of you. TT: To bardzo naiwne z twojej strony. [It’s very silly of you.] Interestingly, Jaworski may have been aware of this kind of ellipsis. In the example below, he clearly appears to have thought Gandalf said ‘a few’ rather than ‘few’, and the reason why he heard the latter may have been due to the fact that in natural speech short words can easily be missed: (H) ST: (- I can’t read it.) - There are few who can. TT: S̜ tacy którzy umiej̜. [There are some who can.] On more than one occasion, the subtitle producer must have misheard the original, due to a combination of inadequate language skills and poor recording quality. As a result, his translations are often wide of the mark: (I) ST: You will keep an eye on Frodo, won’t you? Two eyes. As often as I can spare them. TT: B̜dziesz miał oko na Froda? Nawet par̜. Tak cz̜sto jak tylko b̜d̜ je mógł wyczarowac ́. [( ... ) as often as I can conjure them up.] 54 Łukasz Bogucki Here Jaworski must have misheard the verb ‘spare’ thinking it was ‘spell’, and given that the sender of the message is a wizard, interpreted ‘spell’ as wyczarowac ́ [conjure up]. The same situation can be found in the three examples below: (J) ST: Your love of the Halflings’ leaf has clearly slowed your mind. TT: Twoje uwielbienie ż ycia hobbitów spowolniło twój umysł. [Your love of the Halflings’ lifestyle has slowed your mind – ‘leaf’ understood as ‘life’.] (K) ST: The friendship of Saruman is not lightly thrown aside. TT: Przyjá́ Sarumana nie jest już po naszej stronie. [The friendship of Saruman is no longer on our side – ‘aside’ for ‘side’.] (L) ST: One ill turn deserves another. It is over! (understood as ‘the ring returns to serve another’.) TT: Kiedy Pieŕcié powróci by służ yc ́ innemu, wszystko b̜dzie skóczone! [When the Ring returns to serve another, everything will be over!] Below is another interesting error. During the conversation that takes place between Gandalf and Saruman, the subtitle producer may have misheard the name uttered (‘Sauron’) thinking the wizard was addressing his interlocutor: (M) ST: But we still have time, time enough to counter Sauron if we act quickly. TT: Mamy jeszcze czas, Sarumanie, jeż eli b̜dziemy działac ́ szybko. [We still have time, Saruman, if we act quickly.] This also shows that amateur subtitle producers may be unfamiliar with subtitling standards developed for professional translation purposes (Karamitroglou, 1998). When time and space limitations are stringent, if a character addresses another one by name in the original dialogue (and both are seen on screen), the subtitler leaves out the name, as the context makes it clear who is being addressed. Had Jaworski known this rule of thumb, he would probably have ignored the name (we are assuming here that he thought it was ‘Saruman’, not ‘Sauron’). However, unwarranted reliance on visual stimuli or excessive context-dependence may result in translations which are impossible to Amateur Subtitling on the Internet 55 understand. In the following example, Galadriel, Queen of the Elves, greets Frodo at her abode with the following words: (N) ST: Welcome, Frodo of Shire – one who has seen The Eye. Before she mentions ‘The Eye’, which in the context refers to Sauron, the Great Enemy depicted throughout the film as an evil eye, the camera shows her own eyes in close-up. This visual context may have misled the translator: TT: Witaj Frodo z Shire’u ... Co widzisz w moich oczach? [Welcome Frodo of Shire ... What do you see in my eyes?] On another occasion, two characters are talking about Gandalf, the fellowship’s leader, who has fallen to his death in the mines of Moria: (O) ST: For me the grief is still too near. TT: Nie powinnímy go jeszcze opłakiwac .́ [We shouldn’t mourn him yet.] The subtitle producer might have opted for another translation had he not been over-influenced by the book. In Part Two (which Jaworski could not have seen on screen at the time of the TT production, but might had read in book form) Gandalf returns; it turns out that he has survived his misadventure. This is what the translation implies but in fact it is inappropriate, as the original makes no such implication. The intended message of the original is merely that Gandalf’s fall is still a fresh memory for the fellowship. On balance, it can be deduced that the process of non-professional amateur subtitling by someone with a limited knowledge of the source language, tends to result in erroneous, infelicitous or, at best, chance decisions due to the following factors: ● ● ● ● ● Inability to identify less commonly used words that fall outside the domain of basic vocabulary (examples A and B). Failure to comprehend complete utterances; taking translational decisions on the basis of minimally heard and understood information (example F). Misinterpreting ellipsis (examples G and H). Misunderstanding single lexemes (examples I, J, K) and longer stretches of connected speech (example L). Excessive reliance on context (examples N and O). 56 Łukasz Bogucki Conclusion As illustrated by this discussion the quality of the translation of films differs greatly between amateur and professional subtitling. While in both cases the outcome, to a substantial extent, is conditioned by other elements of the filmic message, the differences are too significant to attempt to apply whatever conclusions seem valid for one type to the other, especially as the product of amateur subtitling tends to be marred by translational error due to the translator’s lack of linguistic competence in the SL, incomplete source texts, or both. Is there then a future for this type of home-made subtitling? As freeware computer programs are widely available, technically speaking anyone owning a PC and having access to the Internet and a copy of a film can make their own subtitles and post them on the web. But why should they do so in the first place? In the case of the translation discussed here, the rationale behind the decision to translate the film was two-fold: to make a name by being second to none to subtitle a release as popular as The Lord of the Rings, and to enable the Internet community to watch (illegally copied) films in a language other than English. Leaving aside the – otherwise important – question of the legal aspect, the subject is discussed here with an eye to the future; if Internet subtitlers were to get access to official scripts rather than what they manage to take down from poor quality recordings, and if the subtitled films were to be made legally and officially available (possibly for a fee), the resulting data could be compared and contrasted with professional cinema productions. It might then be interesting to research the linguistic implications of inhouse regulations issued by film and translation companies vis-à-vis the technical restrictions of computer video players. Do amateur subtitlers pose a threat for professional subtitlers? This question cannot be fully addressed until we have a working answer to the question of whether movies available on the Internet can compete against the big screen and DVDs. It seems, however, that there will always be a market for both cinema productions and digitally recorded copies of films for home use. Hence, it would seem justified to assume that cinema/DVD professional subtitling will continue to exist alongside ‘home-made’ captions. The latter, however, will have to become more professional in every respect. If amateur subtitling ceases to be ‘amateur’, in other words if the source text is not taken down from a poor quality recording but made available to the subtitle producer prior to producing the subtitles, the resulting target text can be subject to translation quality assessment and compared to professional cinema Amateur Subtitling on the Internet 57 subtitling. Then – and only then – can it be studied by academics and scholars. If, however, amateur subtitling continues to be done on the basis of incomplete information, it will necessarily also be imperfect and not available to academic study due to its high degree of unpredictability. As a result, this contribution is merely a prolegomenon to a more thorough study of amateur subtitling, which can only be undertaken if amateur captions are produced under conditions comparable with those of professional subtitling. Notes 1. Some computer programs (movie players capable of showing subtitles and amateur subtitling software) limit the number of characters per line of subtitle, but as the limit can be 60,000 or more characters, for the purpose of this discussion it can safely be ignored, as it does not impinge on the translation process, no quantitative constraints being applicable. 2. Mirosław Jaworski, personal communication. 3. For Polish subtitles, see www.napisy.prv.pl. Films cited in my examples are The Lord of the Rings: The Fellowship of the Ring, Peter Jackson, New Zealand/ United States, Warner Bros 2001, and Władca Pieŕ cieni: Druż yna Pier ́ cienia, Warner Bros Polska, 2002, Polish subtitles by Elż bieta Gał̜ zka-Salamon. References Bogucki, Ł. (2004) A Relevance Framework for Constraints on Cinema Subtitling. Łód́: Wydawnictwo Uniwersytetu Łódzkiego. Gottlieb, H. (1992) ‘Subtitling – a new university discipline’. In C. Dollerup and A. Loddegaard (eds) Teaching Translation and Interpreting (pp. 161–70). Amsterdam and Philadelphia: John Benjamins. Karamitroglou, F. (1998) ‘A proposed set of subtitling standards in Europe’. Translation Journal 2 (2). http://accurapid.com/journal/04stndrd.htm Kilborn, R. (1993) ‘Speak my language: Current attitudes to television subtitling and dubbing’. Media, Culture & Society 15(4): 641–60. O’Hagan, M. (2003) ‘Middle Earth poses challenges to Japanese subtitling’. LISA Newsletter XII. www.lisa.org/archive_domain/newsletters/2003/1.5/ ohagan.html 5 The Art and Craft of Opera Surtitling Jonathan Burton Introduction While subtitles are translated text displayed below the image, as on a cinema or television screen, surtitles are displayed above the stage, in live opera or theatre performances (some opera companies refer to these as ‘supertitles’). Subtitling and surtitling involve differing requirements and techniques. In this chapter both approaches to the translation of operatic texts will be discussed. History of subtitling and surtitling for opera Subtitles for opera on film have been around almost as long as cinema itself, since the early years of the twentieth century (Ivarsson, 2002). On television, the first subtitles for opera in the early 1970s consisted of a series of caption boards placed in front of a camera and superimposed on the television picture. This cumbersome arrangement was superseded by experiments with automated electric typewriters, and eventually by the familiar electronic systems in use today, including teletext, which was used on the short-lived LaserDisc format. For video and DVD, titles can be cued to timecode with an accuracy of a single frame (1/25 of a second), and software can provide useful information, such as whether a title is flashed up too quickly to be read at a specified reading speed; this is important in opera, as the pace of sung text can be much faster, or slower, than the speed of normal conversation. The history of surtitles is not well documented. Reputedly, the first live titles in an opera house were in Hong Kong in the early 1980s but did not fit the definition of either sub- or surtitles, as they were in Chinese and therefore displayed vertically at the side of the stage. 58 Art and Craft of Opera Surtitling 59 English surtitles were first used in Canada in about 1984. In Britain, they appeared experimentally at the Royal Opera House (ROH), Covent Garden, in 1986, and were soon taken up by other organisations such as the Welsh National Opera, Glyndebourne Festival Opera, Scottish Opera and various touring companies, Opera North in Leeds being a recent convert. English National Opera resisted the use of surtitles for many years but decided to introduce them during the 2005–6 season in an attempt to bring opera to the widest audience.1 Hardware and software for surtitles Various systems for surtitling are currently in use, with no widely or internationally agreed ‘standard’ system; each type of surtitling package has advantages and disadvantages. Titles can be projected on a screen, either from slides or electronically via a digital projector, or with presentation software such as PowerPoint. There are various types of LED screens, that is, a matrix of illuminated dots, some of which have refinements such as variable brightness, fading, and colour options. Most recently there are ‘seatback’ screens, like those found in airliners, which display the titles on a small screen below eye-level on the back of the seat in front of the user. These were pioneered at the Metropolitan Opera in New York, and are now used in Vienna, Barcelona and several other opera houses around the world. They have the advantages of multi-language possibility, directionality so that the titles can be visible only to the individual user, and, perhaps most importantly, the option to be switched off if the audience member does not wish to have titles at all. The Royal Opera House in London has used a variety of systems, starting with Kodak carousels each containing a limited number of 35mm slides. Some of the disadvantages of this system are that the slides are prepared in advance by an outside supplier, so re-editing of the text is impossible, and that even with three projectors and three carousels, only a limited number of titles can be used in each act of an opera, before the carousels have to be changed. Another setback is that with such a cumbersome system, the possibility of something going wrong in performance is considerable. This system was superseded by a FOCON LED matrix screen, which meant a vast improvement in simplicity and reliability of use, flexibility of editing, and the capacity for an unlimited number of titles, though font style and text layout are restricted. FOCON screens are no longer manufactured, but they are still widely used, particularly by touring companies. The Royal Opera House now uses a custom-tailored digital system (Diamond Credit ROH, by Courtyard 60 Jonathan Burton Electronics Limited), that allows the projection of electronically-generated titles via a digital video projector on to a suspended screen above the proscenium. This is backed up by a limited number of seatback screens, relaying the titles to areas at the back and sides of the auditorium which do not have a view of the main surtitle screen. Texts are prepared and edited on PCs in the Surtitle Office elsewhere in the building, and then downloaded via the company’s internal IT network to the computer in the surtitle operators’ box, situated behind a window at the back of the balcony. During the performance the surtitle operator cues the titles from a marked-up musical score. The output signal is relayed to the digital projector, high up at the back of the amphitheatre, and via a separate computerised system to the seatback screens. The Diamond Credit software has great flexibility in the programming of parameters such as luminance and speed of fade (in and out) for each title, and individual letter spacing (kerning) for legibility – an important consideration when the projected letters are 30 cm high. However, at present, there is no facility for coloured text and the system can only show white lettering on a dark screen. There is of course no multi-language capability and only one language can be projected. If the seatback installation were extended to cover all the seats in the auditorium, this would become a possibility, although the logistical problems would be great. Why surtitles? In the early days of surtitling, there was much debate as to whether surtitles were necessary or desirable. Opera critics and stage directors tended to be opposed to them, with audiences mostly in favour. That battle has now largely been won, with only a few critics and directors resolutely against the idea. In the eighteenth century – before electric lighting with dimmers – the more well-heeled members of the opera audience would have invested in a printed libretto with translation, which they would follow during the performance with the house lights up; this is perhaps the nearest equivalent to modern surtitling. An early argument in favour of some form of translation as an aid to comprehension was amusingly set out by Joseph Addison, writing in The Spectator around 1712 (quoted in Marek, 1957: 567–9): It is my design in this paper to deliver down to posterity a faithful account of the Italian Opera, and of the gradual progress which it has made upon the English stage: For there is no question but our Art and Craft of Opera Surtitling 61 great-grandchildren will be very curious to know the reason why their forefathers used to sit together like an audience of foreigners in their own country, and to hear whole plays acted before them in a tongue which they did not understand [ ... ]. We no longer understand the language of our own stage; insomuch that I have often been afraid, when I have seen our Italian performers chattering in the vehemence of action, that they have been calling us names, and abusing us among themselves; but I hope, since we do put such an entire confidence in them, they will not talk against us before our faces, though they may do it with the same safety as if it were behind our backs. In the meantime I cannot forbear thinking how naturally an historian, who writes two or three hundred years hence, and does not know the taste of his wise forefathers, will make the following reflection, ‘In the beginning of the eighteenth century, the Italian tongue was so well understood in England, that operas were acted on the public stage in that language’. Our approach to watching opera has changed in recent decades and we are no longer content just to appreciate the lovely sound of the voices and let the opera wash over us. We are now a text-dominated society, and audiences expect to know in detail what words are being sung, as they would for the dialogue in a subtitled foreign film. No longer do we sit in the dark for hours at a time, listening to whole acts of Wagner or Richard Strauss with only the dimmest idea of what is actually going on. Surtitles are now largely a necessity and there are likely to be complaints if they are not provided. The approach of opera directors has changed too. With surtitles, directors know that we will be aware of the meaning of what is being sung, and they will no longer be tempted to fill the stage with superfluous action or comic elements just to keep our attention from wandering. David Pountney, former Director of Productions at English National Opera, one notable director still opposed to surtitling, has famously compared surtitles to ‘a prophylactic between the opera and the audience’. But what are the alternatives? We cannot now sit in the dark and try to follow a printed libretto. Critics hope that we will do our homework, and study the opera text in advance before coming to the theatre; but it would hardly be possible to remember all the words in detail at each point in the action. One alternative is to sing the opera in translation, in the language of the audience, a viable policy followed at English National Opera and until recently at least in many German and Italian opera houses. However, it raises many problems, specifically the 62 Jonathan Burton difficulty of arriving at a singing translation which follows the composer’s musical line and is comfortable to sing with, for instance, appropriate vowels on high notes, while remaining an accurate rendering of the original text. These questions are fascinating, but beyond the scope of this chapter. One point of contention is whether or not to surtitle an opera in the language in which it is being sung. The Royal Opera as a matter of company policy, now provides English surtitles for operas sung in English, for the benefit of members of the audience who rely on surtitles because of hearing problems, or for whom English is a foreign language; but there is still some opposition to this, particularly from singers and directors, who feel that it is an insult to the singers and their clarity of diction. In the case of Britten’s The Turn of the Screw, performed at the ROH in 2002, Deborah Warner, the director, requested that English surtitles should not be used at the first few performances. On the opening night there were 47 complaints from the audience about the lack of surtitles, but none at all when these were subsequently reinstated. Conversely, some composers and directors specifically ask for English surtitles and two examples are Harrison Birtwistle for his opera Gawain and John Adams for El Niño, both performed at the ROH in 2000. Even for older works in English, such as Handel’s Semele, it may be a good idea to provide surtitles, since the poetic, archaic and sometimes convoluted language of the libretto may be difficult to disentangle by ear but comprehensible when laid out on the screen. The aim of surtitles The aim of surtitles is to convey the meaning of what is being sung, not necessarily the manner in which it is being sung. Interjections such as ‘Oh!’, ‘Ah!’ or ‘Ye Gods!’, and musical repetitions of words and phrases, need not be included. Flowery or poetic turns of phrase may be simplified. Punctuation can be kept simple or omitted, in particular exclamation marks; we can see the singers ‘emoting’ on stage, so the titles do not have to do it for us. For example, a fictitious line of Italian operatic text might be literally translated as:2 Ah! How this poor heart is bursting within my bosom! For surtitles, a simplified rendering is preferable: My heart is breaking Art and Craft of Opera Surtitling 63 It may be necessary to clarify the plot, for example tactfully adding a character’s name if it is not clear who is referred to. Sometimes it may help to expand details of the action that may not be clear to a watching audience such as ‘I am putting poison into this jug of water’. However, the surtitler must remember that the audience has come to see the opera, not the surtitles and the titles should be discreet and not distracting. Titles may be omitted where the action is self-explanatory: Beloved Alfredo! Oh joy! – My Violetta! Oh joy! Take this letter which I am giving you What happened next? – Listen and I shall explain to you – Speak then, I am listening ... (etc.) There is nothing worse than sitting through an entire scene of an opera and realising at the end that one does not know what the singers looked like, as one has been too busy reading the titles. This problem is difficult to avoid in ‘wordy’ operas with a great deal of text and little repetition, for example Wagner’s Tristan und Isolde and Parsifal, Richard Strauss’s Die Frau ohne Schatten and Capriccio. The subtitler’s aim should be transparency, or even invisibility. As with a subtitled film, the ideal critical reaction should be along the following lines: ‘Did you go to the opera last night? What did you think of the surtitles?’ – ‘Surtitles? I didn’t notice there were any’. Many opera libretti are based on existing plays or novels, so it may help to consult the original source works – Beaumarchais, Pushkin, Shakespeare, Schiller – though there may be little of the original left. For instance, Ambroise Thomas’s opera Hamlet does contain the line Être ou ne pas être? but almost nothing else from Shakespeare’s text. Sometimes the motivations or relationships of the characters may be clearer in the source work than in the libretto as set by the composer, thus aiding the surtitler’s task. It is also vital to obtain the correct version of the musical score in use for a particular production. Many operas exist in more than one original version, or in conflicting modern editions; Bizet’s Carmen is notorious in this respect. Sometimes a director or conductor may exhume an unfamiliar version, or an extra aria, even for a well-known Mozart or Verdi opera. There may be variant readings in different versions of the score, in which the alteration of even a single word may change the sense of a line; or some directors may have taken things into their own hands, and changed the sung text where it conflicts with the production ‘concept’. 64 Jonathan Burton Each opera company evolves its own ‘house style’ for surtitles, but there is general agreement on some points. a. Each caption contains either one or two lines of text, to a maximum of about 40 characters per line as any more would be too much information for the viewer to take in. The text will normally be centred, and in an unobtrusive font such as Arial or Helvetica. b. Dashes are used if two speakers appear on the same caption, either singing simultaneously, in duet, or in rapid dialogue: – Dear Figaro, look at my hat – Yes, my darling, it’s lovely However, care must be taken not to ‘give away’ dramatic punch-lines too early by putting two lines on the same caption, for example, ‘Who am I?’ [pause ...] ‘You are my long-lost son’. In opera there may be trios, quartets or larger ensembles and given that surtitle systems can only display a maximum of two lines of text, some judicious selection is necessary. c. Italics signify that the voice is offstage, or that the text is a letter or a ‘song’, sometimes a difficult concept in opera, when everything is sung. Italics may also be used in normal fashion, for emphasis. d. Quotation marks signify reported speech. The convention at The Royal Opera is to place quotes in double quotation marks and italics, for clarity: She said “I am his mother” e. Brackets or parentheses signify an ‘aside’, i.e. a line that is not intended to be heard by other characters on stage: (I mustn’t let him see me hiding here) f. Left and right: some systems allow the possibility of splitting titles on either side of the screen, which will help to indicate which character is singing in a large ensemble. This is effective only with short lines of text, and can be distracting. It is therefore more suitable to comedy than to serious opera. Left/right titles may be successive: He thinks I cannot see him She thinks I cannot see her Art and Craft of Opera Surtitling 65 or even simultaneous, although in this case they must be used with due care so as not to confuse the audience: I’m hiding in this alcove Figaro is spying on me Pacing and linguistic flavour For live surtitles in the theatre, the pacing should be kept slow and simple, to avoid distracting the audience. If a text is much repeated, the title can be left up for a long period if necessary. Fast exchanges should be combined, simplified, or omitted for if they are too quick to cue, they will be too quick to read. For the subtitling of a live television broadcast, the pacing can be faster. It will be necessary to consult the TV director’s camera script and edit the titles to fit the shots he or she has chosen, avoiding if possible going over a cut. Close-ups may give the opportunity to title lines which would be impossible to catch at the slower pace of surtitling. Generally, a title should not be held on the TV screen for longer than it takes to read, provided that the effect does not jar musically, and a maximum of about 10 seconds is acceptable. For video and DVD, ‘fine’ editing is possible, so that tiny details of text may perhaps be titled, although anything less than one and a half seconds is probably too short to read. As already noted, opera libretti tend towards the flowery, archaic and poetic in their vocabulary and grammatical formulations. This is particularly a problem in the operas of Richard Wagner, with their long and convoluted German sentence structures. The titler should try to reduce subordinate clauses to simple sentences, and simplify grammar and vocabulary, keeping to clear, modern vernacular unless there are exceptional circumstances. Slang, expletives and colourful language should be treated with care; for example, nineteenth-century Italian opera is notoriously well endowed with words for ‘bad man’ or ‘transgressor’ (empio, traditore, infido, barbaro, misero, cattivo, rio, mostro, sciagurato ... ), for which English is hard pressed for equivalents. It is better to play safe and just use ‘he’, rather than alarm or puzzle the spectator with variations on ‘wretch’, ‘villain’ or even ‘bastard’. Problems of translation: two tiny examples Despite all these considerations, the actual translation of the text should aim to be as accurate as possible, but this may not be so easy. To take a 66 Jonathan Burton famous example from Puccini’s La Bohème, Rodolfo’s aria in Act I, addressed to his next-door neighbour, Mimì: Che gelida manina! Se la lasci riscaldar. Cercar che giova? Al buio no si trova. Anyone who has seen the opera will know roughly what he is saying: ‘Your hand’s cold, let me warm it up. It’s no use looking for your key in the dark’. And any English opera-goer of the last hundred years will know the familiar sung translation in English: ‘Your tiny hand is frozen’. But what exactly do these Italian words mean? Che is ‘what’; gelida is not precisely ‘cold’ (which would be fredda), or ‘frozen’ (gelata), but ‘icy’ or ‘freezing’ (the archaic English word ‘gelid’ could hardly be used in a subtitle!); and manina is not simply ‘hand’ (la mano) but a diminutive, ‘little hand’. Interestingly, Rodolfo does not say ‘your little hand’ – la tua manina – as this would have committed him to using the intimate tu form to a girl he has just met. In fact they have been addressing each other with verbs in the familiar second person singular ever since she came in, while carefully avoiding actually using the pronouns tu or te. Therefore an accurate translation would be something like ‘What an icy little hand’, or even ‘Your tiny hand is frozen’, which turns out to be not so far off the mark after all. She is suffering from tuberculosis, one of the symptoms of which is coldness of the extremities due to poor circulation; but we cannot have explanatory footnotes in surtitling. So we can see that even with a familiar text we cannot take a single word for granted. After Rodolfo has told Mimì all about himself, she replies: Mi chiamano Mimì, ma il mio nome è Lucia ... ‘They call me Mimì, but my name is ... ’ – what? Lucia? But we are translating into English: shouldn’t she be Lucy? And we are in a Parisian garret, so surely she should have a French name: Lucile? If we look at the original novel on which Puccini’s librettists, Giuseppe Giacosa and Luigi Illica, based their text, Henri Murger’s Scènes de la vie de Bohème (1947), we discover that she is indeed called Lucile – except that the librettists have altered the story, and the girl with the cold hands is not Lucile but Francine, since Giacosa and Illica combined features of various characters in Murger’s tale to arrive at the operatic Mimì. Even more confusingly, if we research the life of Murger himself, we will discover that he had a predilection for a succession of pale, doomed tubercular girls, each of whom he would Art and Craft of Opera Surtitling 67 nickname ‘Mimì’. As it happens, the translation of the name ‘Lucia’ is not a serious problem, since it is never referred to in the opera again; but what do we do about the other characters? Should Rodolfo be ‘Rodolphe’, as in the novel? Or ‘Rudolph’, as our translation is in English? The standard convention, or compromise, is to stick to the version of the names as they appear in the opera, illogical as this may sometimes seem: thus ‘Lucia’ and ‘Rodolfo’. The same problem arises in many other operas, for example Verdi’s La traviata – also with a Parisian setting – in which the hero and heroine are Italianised as Alfredo and Violetta rather than Alfred and Violette or, in English, Alfred and Violet – a most unromantic-sounding couple. In Dumas’ original autobiographical novel and play (1986), they are not in fact called Violette and Alfred but Marguerite and Armand; names that have been reinstated in some opera productions as well as a famous ballet. One could even go back to the true story behind Dumas’ re-telling and be tempted to call them Alexandre Dumas and Marie Duplessis, although this may be stretching a point. Pitfalls A problem with live surtitles is that an audience of perhaps two thousand opera-goers will be quick to laugh at anything they find amusing, whether intentional or not. Solitary TV or video/DVD viewers similarly should not be distracted by inappropriate reactions. One must guard against ‘howlers’. For example, in Rossini’s Guglielmo Tell (William Tell), the line La gioia egli vedria d’Elvezia intera that was initially translated as: How happy he would have been to see Switzerland united was pointed out as suggesting a football team (‘Switzerland United’), and had to be hastily changed to ‘a united Switzerland’. Another notorious case is provided by the surtitles for Puccini’s Tosca at an American opera house. When the jealous Tosca is asking Cavaradossi to change the colour of the eyes in a portrait he is painting, Ma fa gli occhi neri, the request is usually tactfully translated as ‘But make the eyes dark’. On this occasion, however, it was rendered too literally as ‘Give her black eyes’, to the great amusement of the audience. When jokes occur in the original text and therefore are intentional, the cueing of the surtitles must be timed so that the audience does not laugh too soon; nothing upsets singers more than a laugh which comes before they have sung the relevant punch-line. The surtitler may attempt 68 Jonathan Burton to split the line into two titles, to delay the laugh. This, however, might distract the audience’s attention, resulting in the stage action being missed altogether. The addition of jokes not found in the original is not advisable, although most titlers are at times severely tempted. The ideal is a joke which is suggested by the libretto and as a result does not contradict the spirit of the original. A classic example is found in Rossini’s comedy La Cenerentola. Dictating a decree, Don Magnifico turns to his eager scribes with the words ‘No, put it in capitals’. So, the next surtitle obligingly appears as: I, DON MAGNIFICO, DUKE, BARON, ETC., ETC., HEREBY DECREE ... which never fails to raise a huge laugh, although it is implied in the libretto. The need for further research Some operas have exotic or obscure settings which require a certain amount of historical research in order to establish the correct spelling of unfamiliar names and places, or even to determine what exactly is going on; the titler needs to understand the text to be translated, or the audience certainly will not. A classic example is Adriana Lecouvreur, an opera by the late nineteenth-century Italian composer Francesco Cilèa, which has a libretto derived from a French play originally written for Sarah Bernhardt. The characters are based on historical figures: the actress Adrienne Lecouvreur, and her lover, a Polish nobleman named Maurice, Count of Saxony (or Maurice Saxe), who appears in the opera as Maurizio, Conte di Sassonia. At one point in the opera, Maurizio is asked to recount his exploits in Curlandia and ‘the attack on Mitau’. In pre-Internet days many hours spent sitting in libraries delving in encyclopaedias finally revealed that Courland, Curlandia, is present-day Lithuania, whose ancient capital was indeed not Vilnius but Mitau. The original Italian text begins as follows: Il russo Mèncikoff riceve l’ordine di cormi in trappola nel mio palagio ... Era un esercito contro un manipolo, un contro quindici ... Ma, come a Bèndera Carlo duodecimo, nemici o soci contar non so ... Art and Craft of Opera Surtitling 69 Already, in the four allusions we have a sheaf of problems: who was Mèncikoff, and how is he spelt in English? Is un esercito contro un manipolo some kind of proverbial saying? Who was Carlo duodecimo and where was Benderà, and what was he doing there? Charles XII turns out to be a Swedish king [1682–1718], held captive in exile at Bender or Benderey in Bessarabian Turkey in 1708. He defeated not only a Russian army but also Augustus II of Poland and Saxony, so Maurizio, the ‘Count of Saxony’, would have known what he was talking about. Is nemici o soci contra non so another famous historical quotation? And so it goes on. The final subtitle rendition of Maurizio’s recitation goes as follows: The Russian, Menshikov, was ordered to trap me in my palace It was an army against a handful, fifteen to one But like Charles XII at Benderey, I could count neither friends nor enemies ... All these problems are very minor, and should not be evident to the audience. However, failure to deal with them may result in inaccuracies or ‘howlers’ in the titles. Conclusion The subtitling of opera on television, video and DVD, and latterly the surtitling of live opera in the theatre, are disciplines which have come into being only in the last few decades and they have developed rapidly both in sophistication of hardware and software, and in subtlety of application. A correct and sensitive approach to the translating of operatic texts will involve considerations and problems not found in other forms of translation. Finding successful solutions can be hard work, but can also be satisfying. The titling of opera is not only a craft, but also an art. Notes 1. More information is available on: www.eno.org/src/eno_introduce_surtitles. pdf 2. All surtitle translations quoted are either fictitious examples, for the purposes of illustration, or are from titles by the author. 70 Jonathan Burton References Dumas, A. fils (1986 [1848]) La Dame aux Camélias, D. Coward, (trans.) Oxford: OUP. Ivarsson, J. (2002) A Short Technical History of Subtitles in Europe. www.transedit.st/ history.htm Marek, G.R. (ed.) (1957) The World Treasury of Grand Opera: Its Triumphs, Trials and Great Personalities. New York: Harper & Brothers. Murger, H. (1947 [1847–49]) Scènes de la vie de Bohème. Paris: SEPE. 6 Challenges and Rewards of Libretto Adaptation Lucile Desblache Introduction During the interval of one of his recitals the Irish tenor John McCormack was told that W. B. Yeats was attending his performance. The singer was keen to meet the poet, but the latter was disappointed with last–minute changes in the programme, where popular ballads replaced some of the more highbrow art songs originally announced and is reported to have said quite abruptly that he did not enjoy the concert. Desperately trying to win the poet over, John Mc McCormack fumbled:1 But what about my diction? People usually comment on the clarity of my words.’ ‘Ah!’ is reported to have said Yeats, disappointed not to have heard poems set to music, ‘the clarity of the words, the damnable clarity of the words! In this chapter, I am first going to discuss libretti in general, historically and linguistically. This is followed by a case study in which I was personally involved as a translator; the transfer from English to French of Benjamin Britten’s and Eric Crozier’s Albert Herring. Libretti – past and present In opera, far from complaining about the clarity of the words, the public has only been able to relax and understand the plot since surtitling was adopted by opera houses in the late 1980s. Before then, only those who were familiar with the repertoire or took time to read the story ahead of the performance could understand what was happening 71 72 Lucile Desblache on stage (see Burton, this volume). As far as content is concerned, there rarely seems to be a happy medium in opera. Texts tend to be either over-simplistic or repetitive or based on fiendishly complicated historical plots. In either case, jealous baritones generally scheme the murder of amorous tenors while (not always seemingly) feeble sopranos die of unrequited love singing top Cs. In operatic music and art songs, attitudes to languages have always been different from those in other performance art forms. In straight theatre, one of the priorities of the translator is to convey the specificities of the source culture to the target audience. As a quintessential product of the Renaissance, opera, at least originally, was fuelled by a desire to convey a universal message, beyond local or national significance. Although this trend is not always followed in contemporary compositions, reinterpretation of myths has remained widespread throughout the last century. Italian was adopted as the universal musical language at the start, in the seventeenth century. French opera was soon also established as a school (paradoxically, initially by an Italian, Jean-Baptiste Lully, then by a German, Giacomo Meyerbeer) and composers often wrote operas in collaboration with Italian or French librettists to be performed for audiences, usually aristocratic audiences, whose native tongue was not Italian: Glück, Handel and Mozart, the founders of modern opera, all composed using texts not written in their native tongue. Opera and art songs, as artistic expressions of high culture, were written for an educated audience who expected to hear Italian, French and, to a lesser degree, German. Handel’s Vauxhall audiences were handed out English translations of the Italian libretto performed, but nevertheless, expected operas in a foreign language as an unquestionable tradition. Several versions of operas, such as Glück’s Orfeo, were composed to please audiences – and patrons – of different nationalities. This not only resulted in linguistic transfer but also in substantial musical or production modifications. Yet, while opera was intended for the aristocracy, it appeared only in French, Italian and German. Intercultural awareness spoke the language of the rich and powerful. The situation changed in the nineteenth century. The main operatic and song repertoire was mostly adapted to the language of the country in which they were performed from the mid-nineteenth to the midtwentieth centuries. This was due to an expansion in popularity of opera and comic opera as well as a growth of compositions in a wide range of languages. Major operas were composed in Russian, Czech, Hungarian, Norwegian and other ‘minor’ languages. A two-tier level Challenges and Rewards of Libretto Adaptation 73 of opera performance emerged, of which traces are still visible today, particularly in Britain, where Sadler’s Wells Opera, which evolved into the English National Opera, has a tradition of opera performance in English established in 1931. World-class opera houses, such as the New York Metropolitan Opera always performed works in their original language, whereas provincial and less prestigious houses adapted original pieces. While world-class theatres and opera festivals relied on international audiences, smaller houses, in a spirit of popularisation, adapted operas to entice a local public. Resentment from the music intelligentsia about such transfers was not uncommon. Many musicians have defended the position of orchestral and instrumental music as a pure signifier, as they feel that music expresses a message other than semantic. While operatic and vocal music played with the dual use of signified and signifier, words could transcend their semantic meaning in a musical or poetic context and be pure sound, removed from their linguistic sense. In 1919, Gustav Kobbé, the American opera historian, was already writing: Any speaker before an English-speaking audience can always elicit prolonged applause by maintaining that in English-speaking countries, opera should be sung in English. But, in point of fact, and even disregarding the atrocities that masquerade as translations of opera into English, opera should be sung in the language in which it is written. For language unconsciously affects, I might even say determines, the structure of the melody. (Kobbé, 1933: 2 [1st edn 1919]) The advent of surtitling, with its relatively unobtrusive ways of conveying the semantic message of operatic works, made it possible to watch and hear operas in their original language while conveying the libretto’s message. In most cases, this method of transfer works extremely well and surtitles allow the audience to understand the gist of the plot. In the case of light opera though, direct transfers may still be preferred. In music terminology, comic opera refers to lyrical works which integrate spoken dialogue. However, it can also be used for works written in a lighter vein and this is how I refer to it here. Tragic opera, opera seria, is often based on mythology to some degree known to the audience (Orpheus, Bluebeard, Faust, Don Giovanni), or structured on a narrative made obvious by the action taking place on stage (Carmen, Madam Butterfly, Tosca). Complex historical plots, which at times would obscure intelligibility, can now be summarised very efficiently with surtitles, and often are, even in the case of operas performed in the native tongue 74 Lucile Desblache of the audience, as opera singers are notoriously difficult to understand. In light opera, words may be dependent on action, their meaning may be enhanced by the production, but they can also express humour through puns and purely semantic means. Besides, they have often been composed with a more intimate setting in mind than grand opera and require full contact with the audience. The statement made by Kobbé nearly a century ago that words are not only semantically indispensable to the understanding of an opera but also vital to its musical texture and structure, still stands. Yet in light opera, the need to have direct contact with an audience through words is also vital. Summarising text, as must be done in surtitling, and conveying the message through an intermediary screen may damage the immediacy of communication needed in humour, where timing is supremely important and semantic subtleties are not always transferable. Light opera does not have the high status of grand opera, is usually not conceived on the same scale (using a smaller number of musicians and singers) and perhaps also for these reasons, is often more acceptably transferred into another language. For example, in some ways, the idea of setting a performance of Aida in English is old-fashioned whereas an adaptation of a French operetta such as Orpheus in the Underworld, especially if peppered with contemporary references and completely reset in an British context can be very attractive to the public. This corresponds to the tradition of using mythology and history as a setting for opere serie’s topics from the past, whilst comic operas and operettas are conceived in a contemporary context. Benjamin Britten’s and Eric Crozier’s Albert Herring The reasons why comic opera is less highly regarded than grand opera fall beyond the scope of this discussion, but from a translator’s point of view, lighter music is more challenging, because of the type of musical forms favoured (many recitatives where each note corresponds to a syllable, fewer lengthy arias, where long passages can be sung on one vowel, hence the term ‘vocalise’) and because of the frequent presence of humour. As with any score, restrictions are imposed on the text by the music. Strong beats always correspond to stressed syllables in a word, and for a transfer into a language like French, which has no stresses, this poses many problems and requires varied strategies. To a certain degree, the music dictates which words should be used. In light music the purpose of the translation is slightly different in that the plot is often deliberately ridiculous: the informative function of the text is Challenges and Rewards of Libretto Adaptation 75 therefore played down to the profit of appellative and persuasive functions. Textual parody is very frequent (for example, there are many instances of legal discourse being sent up in comic opera, from Mozart to Donizetti). Register and intonation are also vital elements. Composers emphasise certain words through musical means, be they repeated or not, and it is important not to miss their prominence. Libretto translators usually work from a vocal score, showing piano and vocal parts. They may also have access to the orchestral score and to a published libretto. A libretto is extremely useful, as some additional information (foreword, stage directions, capitalisation of words for emphasis) may be included. It also shows clearly the text forms used, line rhymes and breaks and any structural aspect of the text which may be easily overlooked when considered in its musical context. In the case of Albert Herring, all three forms of publication were available to me. The libretto proved very useful, as rhyming is used with subtlety and might have been missed if not laid out as verse. The example below, an excerpt from Lady Billows’ first aria, is a good illustration of how original rhymes and structures may be overlooked when looking exclusively at a musical score, as they are not textually emphasised (Crozier, 1981: 9): Is this the town where I Have lived and toiled? A Sodom and Gomorrah Ripe to be despoiled? O spawning ground of horror! Shame to Loxford: - sty The female sex has soiled! Eric Crozier started to work with Benjamin Britten as a stage producer and Albert Herring was the first libretto he wrote for Britten. The whole opera was composed between October 1946 and February 1947 ‘to launch the first independent season of the English Opera Group, in summer 1947’ (Crozier, 1947d: 5). Britten’s and Crozier’s collaboration was very close and Britten even asked Crozier to stay with him for a few weeks, so that they could work together, sometimes at the same desk (quoted in Newman, 1947): One of the secrets of writing a good opera is the working together of poet and composer. [ ... T]he musician will have many ideas that may stimulate and influence the poet. Similarly when the libretto is written and the composer is working on the music, possible alterations 76 Lucile Desblache may be suggested by the flow of the music and the libretto altered accordingly [ ... ] The composer and poet should at all stages be working in the closest contact. Britten’s intimate relationship with the singer Peter Pears also made him very aware of issues regarding text and ‘singability’ in opera and songs. It may also have intensified his need to write for the voice and to set words to music. His collaboration with Peter Pears was in fact criticised at the time, as Peter Gammond’s humorous remark shows: ‘He wrote two kinds of works: vocal, which all sound as if they were written for Peter Pears to sing (and were), and non vocal, which all sound as if they were written for Peter Pears to sing (and may have been, but he was busy at the time)’ (quoted in Surrans, 1988: 40). The collaboration with Crozier was very successful and after the completion of Albert Herring, the librettist notes in a letter to his wife-to-be: ‘The opera is all finished except for rewrites. [ ... ] We celebrated with a glass of Genever [ ... ] and Ben said, would I devote the rest of my life to writing libretti for him?’ (Crozier, 1947b). This collaboration was not as fruitful as it promised to be, but the two artists did work together on several pieces, operatic and sacred. Albert Herring is an adaptation of a de Maupassant short story, Le Rosier de Madame Husson, originally written in 1887. Crozier (1981: iii) explains in the foreword to the second edition of the libretto how the story was chosen: [W]e settled on Guy de Maupassant’s brilliant short story Le Rosier de Madame Husson [ ... ]. We transferred the scene of the story to Suffolk, and centered it upon a small market-town familiar to both of us. This brought immediate changes in the treatment of the French story and its characters. Working from our knowledge of English people on one hand and from the qualities of particular singers on the other, we made our first sketch for a lyrical comedy with twelve characters, in five scenes and three acts. As translator, I was therefore finding myself in the unusual situation of having to operate a cultural transfer into a French context, since the opera was going to be performed for a French audience, but having to keep the English essence that Britten and Crozier had imparted to the work, even though the original story was French. Many original allusions to English social behaviour were necessarily going to be lost. The purpose of translating the libretto was mainly to allow this gentle satire Challenges and Rewards of Libretto Adaptation 77 to be understood, in order to allow the French audience to laugh as well as enjoy the music. The question of the (in)visibility of the translator is particularly acute in such work, as this professional must be as unobtrusive as the original librettist was to the music. In Crozier’s words (1947d: 9), the librettist’s ‘highest ambition is to serve the composer’s intention sincerely, neatly and well’. The validity of the concept of the invisible translator has long been put in question (Venuti, 1995), but with libretto transfer, one has to be not only faithful to the original text but also to the music. The translator of such a work may not be required to be invisible, particularly in light genres where s/he may be expected to provide a contemporary version of the libretto, but s/he certainly has to write words that not only serve their purpose in communicating a message, but also serve the music. I was very privileged to have met and developed a relationship with Crozier such that I was able to discuss any issue I felt uncomfortable with. This is a rare privilege indeed. Clarification on some minor points of the original libretto was very useful. But his general advice was not to be too close to the source text. The first edition of the score was originally published bilingually in English and German. Crozier’s advice (1947c) to the German translator was the same: I am especially pleased that you would like to make the German translation. [ ... ] Of course, I needn’t tell you that from my point of view as a librettist, I should wish you to have an entirely free hand and to adapt the text into a German equivalent – it can’t be translated in any literal way.2 The opera is set in 1902 in East Suffolk and relates the coming of age of a simple young working-class boy, Albert Herring, who is dominated by his mother. As they cannot find a pure enough May Queen to be elected, the well-to-do society of Loxford, an imaginary town remarkably similar to Aldeburgh, decides to crown a May King. Albert is chosen but against all expectations rebels and walks off after the ceremony to lead an independent life. I was faced with many problems, three in particular are noteworthy: cultural equivalence, humour and rhyming. The first two issues were often linked, although not always. Although I read the well-known de Maupassant story again before embarking upon the translation, it was more interesting than useful. Albert Herring had become an English story and had to remain one. Another main issue was the general tone and register of the piece. The opera was a satire, but stressing the humour too heavily would destroy the spirit of the piece. It was also difficult to 78 Lucile Desblache convey the obsession of the English with their social system while keeping the lightness of the piece. One of Crozier’s letters (1947a) to Nancy Evans points out very clearly that the light vein of the opera must be preserved: A long letter from Carl Ebert3 tonight – quite stupid – talking about Act 1 of Albert and how the production must express the ‘social criticism’ of the comedy and the ‘mendacious prudery’ of the characters. O God ... ! All so off the point ... Now I must write a long letter back in words of one syllable, explaining that this isn’t an Expressionist or Trotskyist attack on the upper classes of a decadent England, but a simple lyrical comedy. Finally, there was the problem of rhyming. Crozier (1964) took a very clear stand about rhyming: One difficulty that faces anyone who writes a comic opera in English is the question of rhyme. Some form of rhyming is almost indispensable: The common cormorant or shag Lays eggs inside a paper bag. Take away the rhyme, substitute ‘box’ for ‘bag’ and the whole effect is destroyed. Yet this device has been so fully and finally exploited by W. S. Gilbert in the Savoy Operas that exact rhyming (especially if the rhymes are double or triple ones) is liable to sound as obsolete nowadays as the Victorian pun. I tried to avoid this danger in my libretto by using a form of near-rhyming. Near-rhyming is not used as readily in French as in English, but the square model of French rhyming used in opérettes and other light works was hardly suited to the music of Benjamin Britten, with its many disjointed effects and syncopations. Besides, it was desirable to mirror the diversity of strategies used by Crozier. A consistent decision would have to be made, which would allow a contrast between free prose, free rhyming and strict rhyming, as this contrast is essential to the original work. Cultural and linguistic solutions To illustrate solutions chosen as regards cultural equivalence, humour and rhyming, some examples have been selected, from the vocal score Challenges and Rewards of Libretto Adaptation 79 (Britten, 1948a). In Albert Herring, cultural references are mostly linked to social representations (allusion to social class issues, ways of celebrating events). The problem of cultural transfer was not so much that of finding an exact equivalent, since the translator is not tied to finding a literal semantic counterpart. The difficulty, given the music constraints, consisted in finding an equally English concept which a French audience would understand. The example chosen here is taken from the beginning of Act II. The schoolteacher, Miss Wordsworth, has organised the children to sing a festive song in honour of Albert’s crowning. As they gather for a last rehearsal, the children cannot contain their excitement, faced with the food display laid out for the ceremony. An enumeration of different English dishes and sweets, interspersed with the schoolteacher’s last recommendations follows. Delicacies include jelly, pink blancmange, seedy cake with icing on, treacle tart, sausagey rolls (sic) and trifle. Most of these had to be transferred with French equivalents as few English sweets and dishes are well known in France. Musically, this passage is very fast, mirroring the excitement of the children. Short and snappy words have to be used as the musical quality of the language is as important as the equivalent meaning. Party food names chosen for their snappiness were therefore chosen: tartes, bonbons, biscuits, petits fours fitted in well musically, even if the English reference was lost. Cultural dimensions are always more challenging to the translator when they are linked to humour. In the case of Albert Herring, the original English text could sometimes be kept. I decided to leave untouched the exchange of ‘Good mornings’ taking place at the beginning of the opera; it helps to assert the conventional relationships of a provincial town and the Englishness of the opera. Equally, all names were pronounced with an English accent in the French version. Yet, both satire and Englishness are not always so easily conveyed to a French audience. In her congratulatory speech to Albert, Lady Billows, Albert’s patron, uses more and more clichés and set phrases and her vocalisations become quite incoherent. The parody of a stiff upper lip and conventional speech is created through the build up of standard English phrases which are funny through the semi-incoherence of their accumulation. They could not all be transferred into French as it was important to keep to set phrases. Some expressions such as ‘Britons, Rule the deep’ could be transferred adequately (Anglais, Régnons sur les flots), but concise phrases such as ‘King and country’, transferred as Honneur et patrie, lost some of their flavour. The problem of rhyming and near-rhyming was also a recurrent issue in this comic opera. In many ways, this was the most difficult problem 80 Lucile Desblache to solve. In very structured musical pieces which mirror established forms (threnody, hymns), rhyming was relatively easy to transfer. Yet, near-rhyming was used in many passages which were written in recitative form, needed to keep a flexible musical line and to convey semantic ideas. In some cases, rhyming/near-rhyming could not be achieved because of these restrictions: writing a text which gave a clear message and which flowed with the music were the two priorities. For example, in the first act of the opera, each member of the ‘May Queen committee’ suggests a possible young girl to be elected. Singers are very exposed at this stage, as the priority is to communicate the text to the audience. Through the speech and the music they express their social character, their personality as well as the basic message that has been composed in verse to be sung. In spite of many attempts to keep all these elements, I sometimes decided not to use verse in French, as the fluidity of the text would suffer, particularly in the less structured musical passages. I have thus rendered the vicar’s suggestions as follows: The first suggestion on my list Is a charming local girl Who takes Communion And never missed A Sunday – Jennifer Searl La première sur ma liste Est une charmante jeune fille. Elle communie et Vient toujours à l’office: Jennifer Searl. Conclusion In her discussion of Reiss’s definition of audio-medial and multimedial texts, Snell-Hornby (1997: 280) questions the general principle according to which text-types dictate the translation method used: As a general principle, this [the fact that text-types determine the translation method] may be debatable, as most texts are in fact hybrid forms, multidimensional structures with a blend of sometimes seemingly conflicting features, but in the case of the audio-medial text, it is certainly arguable that the translation strategy should depend on possibilities of expression inherent in the human voice. This can develop into an issue where the principles of rhetoric and speakability conflict with culture-specific expectations and language-specific text type conventions. Although Snell-Hornby does not address the specific problem of libretto transfer, she emphasises here both the key elements in libretti which Challenges and Rewards of Libretto Adaptation 81 determine their translation: their hybrid, textual nature and the absolute priority needed to make them singable. So many transfer problems are inherent in libretto adaptation that some translators find it a more frustrating than rewarding experience. Indeed, some of the worst translations published have been texts linked to music and we have all laughed at some dated nineteenth-century translations of art songs. In this sense, the screen translation offered through surtitling may be more satisfactory. Yet, in such works as Albert Herring, based on humour and direct communication with an audience, and which are generally performed in small theatres where sung words can be understood much more easily than in average opera houses, direct contact with the text is, in my view, preferable. It is always difficult to make a judgement of one’s work, and many of my own solutions were not fully to my satisfaction. Nevertheless, it is ultimately very gratifying to know that you are collaborating with a first-class musician and to hear your audience laugh. It was also an unsurpassable experience to meet and get to know Eric Crozier and his wife Nancy Evans, who were so generous in sharing their experience and providing encouragement. I would like this contribution to be a tribute to them. Notes 1. This anecdote was reported to me by an opera singer. 2. In the end, the opera was not translated by Hans Oppenheim, to whom this letter was addressed, but by Fritz Schröder. 3. Carl Ebert was the artistic director of the Glyndebourne opera at that time. References Britten, B. (1948a) Albert Herring. Vocal score. London: Boosey and Hawkes. Britten, B. (1948b) Albert Herring. Orchestral score. London: Boosey and Hawkes. Crozier, E. (1947a) Unpublished letter of Eric Crozier to Nancy Evans, 10th February. Crozier, E. (1947b) Unpublished letter of Eric Crozier to Nancy Evans, 12th February. Crozier, E. (1947c) Unpublished letter to Hans Oppenheim, 21st June. Crozier, E. (1947d) Foreword to the Libretto Albert Herring. London: Boosey and Hawkes. Crozier, E. (1964) Foreword to the Decca recording of Albert Herring. London: Decca. Crozier, E. (1981) Albert Herring, Libretto (2nd edn). London: Boosey and Hawkes. 82 Lucile Desblache Kobbé, G. (1933 [1919]) Complete Opera Book. London and New York: Putnam’s and Sons. Newman, E. (1947) Albert Herring. The Sunday Times, 6th July. Snell-Hornby, M. (1997) ‘Written to be spoken: The audio-medial text in translation’. In A. Trosborg (ed.) Text Typology and Translation (pp. 277–90). Amsterdam and Philadelphia: John Benjamins. Surrans, A. (ed.) (1988) 100 compositeurs de A à Z, Jeux de massacre. Paris: Editions Bernard Coutaz. Venuti, L. (1995) The Translator’s Invisibility. A History of Translation. London and New York: Routledge. Part II Revoicing This page intentionally left blank 7 Dubbing versus Subtitling: Old Battleground Revisited Jan-Emil Tveit Introduction Early in the twentieth century the new film medium transcended all national and cultural borders, but with the arrival of the talkies, the film industry faced a translation problem since only a small percentage of the world’s population understood English. As a result, there was a growing need to find appropriate screen translation approaches. In Europe, France became a forerunner experimenting with both dubbing and subtitling. It did not take long, however, to find out that both approaches had their disadvantages: it was even claimed that translating a film ruined it. To solve the problem a third approach was tried out in the form of multiple versions, which meant that films were shot in several languages instead of one. But the different versions were not on an equal footing, and it did not take long before it became obvious that the TL-versions were suffering. A main problem was that their linguistic quality was not up to par. As time went by, however, French audiences grew increasingly dissatisfied with subtitling and dubbing gained considerable territory. Along with other countries like Italy, Spain and Germany, France gradually developed into a dubbing stronghold, while the Scandinavian countries and the Netherlands, on the other hand, opted for subtitling. During the first few years, this was largely a question of money. But much has changed since the arrival of the talkies and the advent of television. Which one of these two translation solutions is the best option for today? In order to answer this question, I will compare the two approaches paying special attention to their constraining factors. 85 86 Jan-Emil Tveit Constraining factors of subtitling Whereas it may be argued that subtitling is the only intelligent solution (Reid, 1978), it has also been claimed that this approach is only the lesser of two evils (Marleau, 1982). And there is no doubt that subtitling has its constraints. An important aspect of the subtitling process is the filtering of potential loss of information: for the purposes of expressing nuances the written word cannot possibly compete with speech. Hence a large number of lexical items tend to be required in order to match what is conveyed by stress, rhythm and intonation. Normally, however, the subtitler does not have room for wordy formulations or complex structures: in order to enhance readability, brevity is the essence. And if the subtitles are to remain on the screen long enough for audiences to read them, contraction is a must, which in turn can result in a regrettable loss of lexical meaning. Often it is not easy to decide what to leave out. Although there are redundant linguistic features in speech, sometimes even slight omissions may bring about significant changes in meaning. It is difficult to generalise when it comes to reading speed and rate of standardised presentation. According to Luyken et al. (1991), the reading speed of adult viewers hovers around 150 to 180 words per minute. This is, however, subject to extensive variation and depends on the complexity of the linguistic and factual information that the subtitles contain. If lexical density is high, accessibility to the information tends to be low, which calls for added subtitle exposure time. Furthermore, readability is said to be affected by film genre. This is how Minchinton (1993: 14–15) comments on love stories: ‘Viewers need not read many of the titles; they know the story, they guess the dialogue, they blink down at the subtitles for information, they photograph them rather than read them’. Crime stories, according to Minchinton, may give translators and viewers a harder time: if the action is to be understood the subtitles have to be read. A further point made by Minchinton concerns reading speed that may be affected by the viewers’ attitude to the subject matter of films or programmes. He suggests that if viewers find a story exciting they are able to read the subtitles faster. On the other hand, it may be argued that the more interesting audiences find a film, the less inclined they are to spend the time reading. As far as television subtitling is concerned, condensation levels vary between countries. If we compare the Scandinavian countries, stricter rules have traditionally applied in Sweden than in Norway and Denmark. Dubbing vs Subtitling 87 In Sweden the duration of a full double-line subtitle is supposed to be 6–7 seconds, compared to 5 seconds in Denmark. The position of the Norwegian Broadcasting Corporation (NRK) is that a full double-line subtitle should remain on the screen for at least 6 seconds. Except for some research carried out in Sweden a couple of decades ago, the definition of readability is determined more by common sense than on research results. By way of obtaining what I felt was necessary empirical evidence, I decided to test whether the exposure time of Norwegian subtitles could be cut down without significantly reducing readability and comprehension. My samples were drawn from pupils/ students at nine Norwegian lower and upper secondary schools. There were 508 respondents of between 13 and 20 years of age, and the response rate turned out to be as high as 95%. The samples would appear to be sufficiently random to constitute a representative cross-section of Norwegians belonging to these age groups (Tveit, 2005). My results showed that the retention of textual information was only marginally reduced when the exposure time of each subtitle was cut by 1 second. Hence readability was not dramatically affected when the duration of a full double-line subtitle decreased from 6 to 5 seconds. When the exposure time was cut by a further second, however, the situation changed significantly, most respondents losing out on a considerable amount of information. Cohesive devices are often considered omittable. But although they may not have obvious semantic functions, these still play an important role in making relationships and events explicit. A text that does not contain words of this kind may be difficult to access, and omitting cohesive devices in order to boost readability can therefore prove counterproductive. It may, indeed, reduce readability. Admittedly, Gottlieb (1997) has a point when he emphasises the fact that subtitling is additive by nature, that is verbal material is added to the original programme and nothing is removed from it. The usefulness of this addition, however, depends on the viewers’ comprehension of the original dialogue. It is true that tone of voice, stress and intonation may contribute to conveying information across language barriers, but if source and target languages are poles apart in terms of lexis, the value of keeping the original soundtrack may be rather limited. Very few Scandinavian viewers benefit much from dialogue in Russian or Greek being retained in order to further understanding of lexical meaning. They may benefit greatly, however, if the dialogue is in a language similar to their own, and therefore, comprehension is markedly facilitated for Norwegian television audiences when Swedish interviewees 88 Jan-Emil Tveit appear in, for instance, news bulletins as, in addition to being subtitled into Norwegian, they continue to speak in Swedish. Although my research indicates that it is possible to reduce exposure time by a second, it is difficult to generalise in terms of reading speed and rates of presentation of information. As mentioned, the complexity of the linguistic and factual information has to be taken into account. Traditionally, a distinction has been made between television and video on the one hand and cinema on the other. It has been assumed that identical subtitles are easier to read on the cinema than on the television screen. The reason for this has never been satisfactorily investigated, but it is assumed to be related to the size of the screen, the font used for the letters, and the better image resolution of the cinema. Spoken versus written language Another constraining factor of subtitling results from the spoken word containing dialectal and sociolectal features which are extremely difficult to account for in writing. Whereas spoken language tends to contain unfinished sentences along with redundant speech and interruptions, writing has a higher lexical density and a greater economy of expression. In addition, written translations of spoken language often display a tendency toward nominalisation, whereby verbal elements are turned into nouns. Hence it is difficult to retain the flavour of the spoken mode in subtitles. When it comes to keeping the register and appropriateness of the SL-version dubbing can undoubtedly be at an advantage. Previously it was rather common for informal stylistic features of speech to be replaced by more formal and inappropriate language in subtitles, but in this respect a marked improvement can be observed in recent years. It could well be that the orientation away from a traditional ‘bookish’ approach is related to the increasing number of American productions that are shown on television in Norway. This reflects a trend away from a formal register of language in American culture. Sitcoms like Friends, Frasier, Seinfeld and Ally McBeal are far from easy to handle for screen translators and seem to have contributed to making subtitlers more aware of the need to handle register at both the lexical and syntactic level. The translation of such programmes into other languages requires a high degree of communicative competence. Dubbing vs Subtitling 89 Traditionally, Norwegian television has handled four-letter words with conspicuous caution. Since the advent of television, taboo words have been used increasingly in foreign language programmes that the Norwegian Broadcasting Corporation has bought from abroad, but they have not always been translated into equally strong language. This has to do with the assumption that the effect of swear words is reinforced when they are printed. Not uniformly accepted in a country like Norway with its deep-rooted puritan traditions, it now seems strange that it took such a long time for Norwegian translators to challenge the assumption that the printed swear word is stronger than the spoken one. If anything, it ought to be the other way around; a printed representation is probably only a weaker version of a word pronounced with stress and intensity. Literature is full of four-letter words and it seems hardly likely that their force or effect would be reduced in films based on the books. As for the handling of four-letter words in translation, the most difficult task I have encountered was the translation of a Channel Four production portraying Graham Taylor who had been sacked as England manager following the failure to qualify for the 1994 World Cup in soccer. The programme provided me with a number of challenges, not least because of its continuous context-dependent changes of style and register. It ranged from the elevated and analytical language of Channel Four reporters to Graham Taylor giving his players a dressing down for being unable to produce their best results. The most difficult part to transfer to Norwegian was Taylor’s verbal behaviour during matches. As the ball did not roll England’s way that season, four-letter words were fired in rapid succession. Pretty much in line with Norwegian subtitling tradition and what I had been trained to do, I tried my best to tone down the expressive force of the four-letter words, in addition to leaving many of them out. But on the day that the programme was to be broadcast, Norwegian morning newspapers focused on the reception the programme had got in the UK, where Graham Taylor’s strong language had created a stir. As a result, I felt I had to change a number of my translation solutions, the neutralisation strategy having turned out to be inappropriate. Fortunately, I still had a couple of hours to bring the register of the target version in line with the rather colourful original. When the programme was aired in the late evening, the subtitled Norwegian version contained a rather high number of four-letter words; there were no viewer reactions in the days and weeks to follow. 90 Jan-Emil Tveit Visual, spatial and decoding restraints In addition to linguistic restraints there is the visual factor. Trying their best to read everything that has been translated, viewers are often unable to concentrate adequately on other important visual information and sometimes also on oral information. This is regrettable since audiovisual programmes combine words and images, and the translation should observe the interrelation between the way a plot is told and the manner in which it is shown. Subtitles should synchronise not only with speech, but also with image. Twenty years ago, cinema and television films were cut less fast than they tend to be now. This is undoubtedly an important reason why the history of subtitling does not contain much discussion of editing and camera manipulation techniques. Visual aspects are now more an issue, which reflects the rapid development of the last couple of decades in camera manipulation, sequence construction and programme editing. This is not only due to the number of cameras involved and the way they are used, but also to what takes place when the filming of a programme is complete, when the different sections, with their shots, scenes and sequences, are assembled and edited. It is increasingly important for the subtitles to be integrated with the film and to fall in with the rhythm of the visual information on the screen. This is far from easy to achieve when the editing of the film is gaining momentum and the flow of the shots is adjusted to the demands of the narrative and the impact on the audience. With respect to spatial constraints, in the Nordic countries, one- and two-line subtitles are used. The length of the line varies, but normally does not exceed 38 characters. The main limiting factors are the size of the television screen and the size of the font. When the subtitles follow each other in rapid succession, reading one two-liner appears to be less strenuous than reading two one-liners containing the same amount of information. Although precise timing has usually been defined in terms of subtitle–speech synchronisation, research carried out by Baker (1982) in Great Britain indicates that subtitles overrunning shot changes cause perceptual confusion. Interesting as these results may appear, they do not seem to have had too much of an impact on subtitling territory in the Netherlands and the Scandinavian countries and the great majority of Norwegian screen translators still favour two-line subtitles. Sometimes, however, two lines occupy too much space and interfere substantially with the visual information and composition of the Dubbing vs Subtitling 91 picture. A case in point is the animation series South Park, where the bodies and faces of the characters fill up most of the screen. This series can be difficult to handle for subtitlers. Also in news bulletins, faces often dominate the screen as the camera focuses on the so-called ‘talking head’ but since there is limited activity in the picture, the viewers are able to concentrate on the translation. News, current affairs and documentary programmes frequently contain captions or inserts. These are texts added to the film after shooting. They may be names of people taking part, or names of places visited or shown. As captions often appear at the bottom of the screen, there is the risk that they may collide with the subtitles, a situation that should obviously be avoided. Decoding may present translators with a difficult task due to the presence of ambiguities. Normally a translator has the means of following up unclear linguistic or factual points; for the subtitler, however, this is not necessarily the case. One reason may be the lack of a manuscript or dialogue list, often the situation when interviews are made in news programmes. However, as the reporter tends to know enough about the context, the subtitler may be helped with potential decoding problems. Frequently the subtitler does not have the time to obtain adequate knowledge of the context. In addition, the huge number of varieties of English – and the fact that the interviewees’ command of the language may be far from satisfactory – often make translation a rather difficult process. A mistake may occur after the translator has spent considerable time trying to decode a difficult word. But sometimes we simply misinterpret what somebody is saying without realising it until we are confronted with our mistake at a later stage. The following mistake belongs to the latter category. It was made by Swedish television in a translation of Bill Clinton’s 1996 re-election campaign. The words of the president were: ‘We’ve created nine and a half million jobs, more than half of them high-wage jobs’. The translator, who does not appear to have had access to a manuscript, produced the following translation: Vi har skapat 9 ½ milj.arbetsplatser, mer ãn hälften inom motorvägsarbete. [We have created nine and a half million jobs, more than half of them highway jobs.] Mishearing phonemes is a common type of decoding mistake, and seems to be what occurred here. The translator must have been convinced that s/he heard ‘highway jobs’ (motorvägsarbete) and not ‘high-wage jobs’. In 92 Jan-Emil Tveit this particular case there is no doubt that decoding was extremely difficult. The two constructions contained very similar sounds. Nevertheless, common sense combined with adequate knowledge of US society and politics would probably have saved the translator from dropping this brick. After all, it would have been rather extraordinary if half of all new jobs had been created in the field of road construction – even in the homeland of Henry Ford. Stringent deadlines in combination with decoding problems often present screen translators with difficulties and lead to misunderstandings, errors and inaccuracies. They represent considerable decoding constraints and are bound to have a negative impact on the quality of the translation. And when the audience has access to the Source Language dialogue, a high degree of translation accuracy is required. The constraints of dubbing and lip synchronisation When we experience state-of-the-art lip synchronisation, it is not difficult to understand why this method is the favoured screen translation approach in large parts of the world. However, the constraining factors of this approach are very obvious. One important consideration is the loss of authenticity. An essential part of a character’s personality is their voice, which is closely linked to facial expressions, gestures and body language. Authenticity is undeniably sacrificed when a character is deprived of their voice and instead the audience hears the voice of somebody else. At the Cannes Film Festival in 2003, I interviewed 25 people working in the film industry about their screen translation preferences. All but two said they favoured subtitling. When asked why, most of them answered that they regarded subtitling as the most intelligent and authentic option. For actors to experience what their contribution has been turned into in different Target Language versions must be somewhat startling, and it may seem a bit strange that directors do not put their foot down more often than they do, for example, in cases where dubbing has been favoured and subtitling would have been the preferred option. When such linguistic transplantation takes place, it is not only authenticity that is sacrificed but, in addition, credibility, which may be particularly problematic in news and current-affairs programmes when voice-over is used. Reporters seem to be increasingly preoccupied with the dramatic effect subtitled interviews are reputed to have. This probably has to do with the transnational qualities of the human voice. Even if we do not Dubbing vs Subtitling 93 understand the words of a foreign language, the voice itself may convey a great deal of information. Although intonation patterns often vary from language to language, universal features such as the expression of pain, grief and joy should not be ignored: linked to pitch, stress, rhythm and volume, they contribute considerably to conveying information not only about the speakers, but also about the context of which they form a part. Voices reflect the mood and atmosphere of a situation, whether it is at a major sports event, the scene of an accident or the convention of a political party. The effect of a persuasive speech during a presidential or parliamentary election campaign is probably significantly reduced in a voice-over. Since many politicians capitalise on their voices, sound is an important part of their public image. The Northern Ireland politician Gerry Adams is a good example of the significance of the impact of a political voice. For a long period of time the British people did not have the opportunity to hear his voice on radio or television. Whenever Adams made a statement, it could only be broadcast when somebody else read his words, revealing the importance the British authorities attached to the voice of the charismatic politician. It also underlines the significance of letting the audience have access to the original soundtrack. There is little doubt in my mind that subtitling has an important educational value. Visitors to the Scandinavian countries are often impressed by the standard of English of the people they meet, most of whom have never lived in an English-speaking country. Rather than reflecting the superior result of differing language teaching standards, it may in fact be the inherent pedagogical value of having access to the original English language soundtrack that has brought this to bear. In 1987 I undertook a study that had as its objective to monitor the needs, preferences and perceptions of the linguistic standard of 4,200 students of English from the following nine European countries: Norway, Sweden, Denmark, Finland, Germany, France, Italy, Austria and the Netherlands (Tveit, 1987). An interesting result of the study was that listening comprehension was perceived as significantly more difficult by students from ‘dubbing countries’ than by students from ‘subtitling countries’, the former category showing a much stronger need for an increased knowledge of vocabulary when communicating in English. Dubbing is both expensive and time-consuming A further factor to be considered is cost; dubbing is a lot more expensive than subtitling. The fact that figures vary considerably might be linked to the relationship between supply and demand. Since trained actors 94 Jan-Emil Tveit have traditionally been in short supply in small countries like Norway, the cost of using their services has been high, which has contributed to making subtitling a much cheaper alternative. Although the difference in cost has started to even out, dubbing remains 5 to 10 times more expensive. It might therefore seem odd that an increasing number of foreign films and television productions are dubbed for the Scandinavian markets. A case in point would be the family film (children under 7 are not admitted) Charlie and the Chocolate Factory (Tim Burton, 2005). It was dubbed into Norwegian, something that would hardly have happened a few years ago. Other productions that have been dubbed for the Scandinavian audiences in recent years are the animation films Shrek (Andrew Adamson and Vicky Jenson, 2001) and The Polar Express (Robert Zemeckis, 2004). In France, cinema audiences have been able to choose between dubbed and subtitled versions of major productions for a number of years. Although it is still only the thin end of the wedge, a similar development can now be seen in Scandinavia. What is it then that makes dubbing a viable alternative when it is so expensive? The argument seems to be that costs do not matter too much if revenues are big enough. If lip synchronisation can attract bigger audiences, increased translation costs would not present too much of a problem. The television business is increasingly preoccupied with ratings. If a network is able to get an edge on its competitors, costs are not likely to hold it back. In 1997, an interesting screen translation experiment on subtitling took place in Scandinavia. TV 2, Norway’s biggest commercial television channel, decided to test viewer reactions to lip synchronisation and the first three episodes of The Gregory Hines’ Show were dubbed into Norwegian. Some of the country’s most renowned actors were hired, and the show was given prime viewing time on Friday evenings. After three weeks the viewers were asked for their verdict. The result showed that 85% of the viewers wanted to see subtitled versions of the remaining episodes, a somewhat baffling reaction, considering The Gregory Hines Show is a family programme, aired at a time when parents tend to watch television together with their children, many of whom have not yet learned to read. Film aficionados looked at the result as proof of subtitling being the best solution – and far superior to dubbing. But there is clearly a lot more to it than that, and tradition is a key word. One should not forget that preferences depend on what viewers are accustomed to – and Norwegians are used to subtitling. In a country lacking the tradition, it takes time to develop state-of-the-art dubbing. As for TV 2, the result of Dubbing vs Subtitling 95 their market research seemed to convince the powers that be that dubbed versions did not attract big enough audiences. Normally the dubbing process takes considerable time. In the case of news bulletins, it is obviously not possible to invite actors to play the parts of interviewees. Instead, the big US and European networks make use of voice-over. It is a reasonably speedy process, but it tends to distract viewers who tend to concentrate on the initial voice to the extent that they lose out on parts of the voice-over. In most cases, the subtitling of news reports can be done swiftly, and often takes place minutes before the news bulletin starts. The subtitling of films and series can also be done surprisingly quickly, as illustrated by the following example dating back to the 1990s. A couple of hours after US television audiences had been treated to the solution to the murder mystery, the final episode of Murder One was on its way to Europe. It arrived in Norway 11 hours before the ending to Steve Bocho’s drama was to be shown on Norwegian television. Nine hours later the target version had been brought to completion and the programme was ready for transmission. In dubbing countries such as Germany and Spain, the linguistic adaptation process would probably have taken weeks rather than hours, involving a large number of people. For a start, the dialogue list has to be translated. Then the chosen actors have to be given the time to study and rehearse their parts. The recording sessions also tend to take time. From this it would seem evident that as far as meeting stringent deadlines is concerned the usefulness of this screen translation approach is rather limited. Conclusion Based on the premises outlined above, my conclusion is that subtitling is normally a better approach to screen translation than is dubbing. This does not follow logically from counting the number of constraining factors of the two approaches, but has to do with the fact that some of the constraining factors are easier to get around or compensate for than others. In my opinion the loss of authenticity in dubbing, since important aspects of a character’s personality are revealed through the use of their voice, is the biggest problem of all. There are, however, cases where the voice does not form an integral part of a character, it simply belongs to the off-screen commentator. This tends to be the case in documentary programmes, which normally lend themselves more easily to revoicing than subtitling. Here, if the latter method is chosen, it may lead to extensive loss of information. 96 Jan-Emil Tveit Programmes that are cut extremely fast and have rapid speech rates should not necessarily be subtitled. If readability requirements are to be met, the sheer levels of condensation needed result in too great a loss of information. In spite of all its disadvantages, in such cases dubbing may be the lesser of two evils. Moreover, films and programmes aimed at small children have to be dubbed for the simple reason that the target audiences have not yet learned to read. References Baker, R. (1982) Monitoring Eye-movements while Watching Subtitled Television Programmes. London: Independent Broadcasting Authority. Gottlieb, H. (1997) Subtitles, Translation & Idioms. Copenhagen: University of Copenhagen. PhD Thesis. Luyken, G. M., Herbst, T., Langham-Brown, J., Reid, H. and Spinhof, H. (1991) Overcoming Language Barriers in Television: Dubbing and Subtitling for the European Audience. Manchester: European Institute for the Media. Marleau, L. (1982) ‘Les sous-titres ... un mal nécessaire’. Meta 27(3): 271–85. Minchinton, J. (1993) Sub-Titling. Hertfordshire, England. Manuscript. Reid, H. (1978) ‘Sub-titling: the intelligent solution’. In P.A. Horguelin (ed.) Translating: A Profession. Proceedings of the VII World Congress of the International Federation of Translators (pp. 420–8). Ottawa: Council of Translators and Interpreters of Ottawa. Tveit, J.-E. (1987) Orientation Towards the Learner. Stavanger: Stavanger lærerhøgskole. Tveit, J.-E. (2005) Translating for Television. A Handbook in Screen Translation. Bergen: JK Publishing. 8 The Perception of Dubbing by Italian Audiences Rachele Antonini and Delia Chiaro Screen translations, and especially those for the small screen, are consumed by millions of people in Europe, and of course all over the non-English speaking world, at every moment of the day.1 In this respect, Italy is representative of the many nations which require evergrowing amounts of interlingual mediation for both big and small screen. As is well known, Italy is commonly labelled a ‘dubbing country’ which, together with Austria, France, Germany and Spain, has adopted a tradition of dubbing rather than subtitling, the preferred mode of screen translation in Greece, Portugal, Scandinavia and the UK. However, the situation in dubbing countries is rapidly changing as a choice between dubbing and subtitling is now available on satellite stations and, owing to the fact that subtitling is the more cost-effective choice of the two, for the colossal DVD market. 2 However, despite the fact that many younger people who are more proficient in English than their parents may well prefer subtitling to be more widely available, dubbing is bound to remain the chief form of linguistic mediation for some years. Perhaps it is worthwhile remembering that attitudes and preferences are largely an issue of habit and thus a person who has been used to almost a lifetime of dubbing is unlikely to be persuaded to change to a different mode of screen translation: ‘viewers are creatures of habit’ (Ivarsson, 1992: 66) and preferences depend on ‘what the audience is used to rather than rational arguments’ (ibid.: 20). The Eurobarometer survey conducted in 2001 on the language skills of European citizens and their attitudes towards language learning supports Ivarsson’s claims. In fact, 60% of respondents declared that they prefer to see foreign films and programmes dubbed into their own language. In Italy, more than 70% of respondents expressed support for 97 98 Rachele Antonini & Delia Chiaro 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% Germany Spain Non-US Imports France UK Films Fiction Films Fiction Films Fiction Films Fiction Films Fiction 0% Italy US and US Co-production Figure 8.1 Volume of fiction/films broadcast on television channels in the five principal European markets, 2001 Source: European Audiovisual Observatory; Press Release 28 January 2003. dubbing as their preferred form of audiovisual translation (AVT) while countries comprising the other block confirmed their strong support for subtitled products. Television programmes which undergo the process of dubbing in Italy consist of all imported genres of fictional products from films and TV series to sitcoms and soap operas for a total of over 40 hours airing on weekdays and, as illustrated in Figure 8.1, most products are translated from American English. Now of course, while the UK is the chief importer of American programmes, unlike non-English speaking countries, these products require no linguistic negotiation and are thus cheaper by default as they are purchased ‘ready-for-use’. In the other countries reported in Figure 8.1 the same imports require time and investment before they are ready for broadcasting. With globalisation dictating that films should be premièred simultaneously on both sides of the Atlantic, many operators working in Italy’s dubbing sector who are feeling the strain of such a great workload, poor wages and working conditions have begun to take issue with Perception of Dubbing by Italian Audiences 99 their employers. In fact, in February 2004, operators working in the Italian dubbing sector took industrial action in protest at the fact that they do not possess the fundamental rights of any worker in a democratic country, namely a national contract. From dubbing translators to actors, workers withdrew their labour thus threatening the substantial revenues expected at box offices and from TV advertising. In a totally anarchic market, a film which once required three weeks to dub from start to finish, in the new millennium necessitates the same task to be completed in five days or else it runs the risk of being sent to a small make-do ‘do it yourself’ company willing to comply with a lower fee and a quick and dirty translation. And what of quality? According to Toni Biocca, a representative of AIDAC, Italy’s association of dubbing translators, executives at RAI and MEDIASET are frequently heard to say della qualità non importa a nessuno [who cares about quality] (quoted in Gallozzi, 2004). Needless to say, the key function of translation is to allow people to be privy to texts in languages with which they are not familiar, yet the emphasis of research in audiovisual translation, and translation studies in general, has been on actual texts, the translations themselves and their translators rather than on readers and viewers. In other words, the people who essentially make use of these translations seem to have been largely ignored by researchers. The literature on the processes involved in the work of a screen translator is generally based on studies of a descriptive and prescriptive nature (Gambier, 1998; Gambier and Gottlieb, 2001). Case studies concerned with dubbed texts are abundant and are normally based on contrastive analyses of audiovisual texts in their original form and their dubbed version(s), and it can safely be said that the ins and outs of the various compromises involved in the process of dubbing have been explored in a variety of language combinations by several scholars (Baccolini et al., 1994; Heiss and Bollettieri Bosinelli, 1996; Bollettieri Bosinelli et al., 2000). However, while uncovering and analysing all the possible translating techniques and choices applied to AVT and the point of view and the experience of the translator, such approaches tell us nothing at all about end-users’ perception of AVT and its quality. In fact, media translation is an area in which quality control becomes extremely difficult due to the complexity of several semiotic systems functioning simultaneously. Yet, although interest in this issue is growing both amongst practitioners and academics, regrettably, it would appear that attention stops short at interest. Rare exceptions to this state of affairs are represented by the study on reception of screen translations in 100 Rachele Antonini & Delia Chiaro Greece (Karamitroglou, 2000), work on reception of dubbing in Spain (Fuentes Luque, 2001), and perception of subtitled humour in Italy (Antonini et al., 2003). The present investigation is an attempt to overturn the traditional standpoint of research and look at AVT from the point of view of its recipients rather than the translator or the text in an attempt to identify their attitudes, opinions and above all the perception of what they watch. Interlingual translations are textually unique in that they exist as a version of a pre-existing duplicate in another language. By the same score, it can be safely assumed that recipients of translations, that is to say, the consumers of these translations, are mainly those who are not au fait with the Source Language and that such viewers perhaps would rather not perceive translations as translations. Presumably most recipients would like to hear a smooth, easily digestible text in their mother tongue and, in the case of AVT, although what appears on the screen may be unfamiliar, what is perceived through the ears (in the case of dubbing) should be identifiable and free of turbulence of any sort. This research sets out to explore exactly what audiences perceive when they watch and hear a dubbed product. Thus, by examining the reaction of Italian audiences to dubbed programmes, the research shifts its emphasis from the text and translator to the recipient in an attempt to explore new ground. This preamble contains an underlying implication. Surely smooth, Italian-sounding TV products, the contents of which are easily comprehensible to the average viewer, correlate with good quality products in terms of linguistic mediation? Due to the total inter-dependence of several semiotic systems functioning simultaneously, the verbal elements in AVT are strongly compromised, thus what we have labelled lingua-cultural drops in voltage may well be inevitable. Lingua-cultural drops in voltage refer to viewer perception of lingua-cultural uneasiness or turbulence, such as a cultural reference which is not completely understood, an unnaturalsounding utterance, an odd-sounding idiom or a joke which falls flat. Lingua-cultural short circuits or worse still, lingua-cultural power cuts (and the metaphor is, hopefully, quite clear) which suggest significant faults or even a total breakdown in which communication fails, are outside the scope of this research. On the other hand, audience misconceptions, however slight, are exactly within the bounds of this study. Several translational choices in Italian AVT contain a number of compromises that either filter through to audiences unnoticed, or else are instinctively accepted. Many such translational choices may well have been norm-driven, however, there is no valid reason why a Perception of Dubbing by Italian Audiences 101 norm should not be adjusted if necessary. Thus, in devising the experimental design, the research questions we asked ourselves were the following: ● ● ● ● Are viewers aware that many mediated linguistic forms which occur in dubbed products are unusual in naturally occurring Italian? Do viewers who are aware of lingua-cultural drops in voltage find them acceptable? Who exactly perceives what when watching a linguistically mediated text by means of dubbing, in terms of age, gender, education, etc.? If a programme is successful, do such drops in voltage matter anyway? The focal point of these research questions are viewers and their perception of what they see and hear, rather than our own perception as researchers. However, for reasons of space, the results pertaining to language-specific features alone will be presented in this chapter. The experimental design of the project The experimental design of this project was based on a web questionnaire and a corpus made up of 300 hours of Italian-dubbed fictional programmes. Respondents involved in the study were asked to visit a website, log in, watch four short video clips from the corpus and answer one question regarding each clip. Web technology provided the means to carry out the investigation on a large scale rapidly, cheaply and accurately. Responses were stored directly in an ACCESS database and subsequently elaborated statistically. What follows is a brief explanation of the creation of both corpus and questionnaire. The corpus Over 300 hours of dubbed television programmes were recorded over a period of eight weeks during February and April 2002. A total of 30 programmes were selected and recorded in order to form a representative sample of the choice of dubbed products that Italian viewers had at their disposal on the three state-owned channels (Raiuno, Raidue, Raitre), on the three privately owned channels (Canale5, Rete4, Italia1), and a private channel (La7) which first began broadcasting just before recordings for the corpus commenced in January 2002. Dubbed fictional programmes were recorded on a daily basis between 8am and midnight. Feature films and TV films were deliberately excluded from the corpus.3 102 Rachele Antonini & Delia Chiaro During this period, approximately 133 hours of dubbed programmes were aired each week during the 8am to 12pm time span. Interestingly, the Mediaset group (Canale5, Rete4, Italia1) accounted for 60% of the total amount of broadcast TV programmes. Thus, together with the dubbed programmes aired by the other privately-owned channel, La7, we obtain a total percentage of 76% (79 hours), while the state owned RAI group share only averaged a little less than 33 hours (24% of the total amount). The selection of the programmes included in the corpus took explicit criteria into consideration. Care was taken to include all TV genres which undergo the process of dubbing (series, serials, sitcoms, soap operas, telenovelas and cartoons). Furthermore, time spans and target audiences were also identified and recordings reflected all possible viewing times i.e. mornings when soaps and telenovelas targeted at housewives are aired; afternoons which comprise mainly soaps, cartoons and series aimed at teenagers; prime time evening TV which airs telenovelas and series/serials for a more general audience. The corpus thus included examples of programmes aimed at all specific audiences and age-groups as well as programmes from all Source Languages aired in dubbed versions during the two month recording period i.e. US, UK, Canadian and Australian English; Standard German, Austrian German, Brazilian Portuguese, and various Latin American varieties of Spanish. Editing the corpus All examples of highly specific cultural references and instances of what were considered to be examples of Italian dubbese (Alfieri, 1994; D’Amico, 1996; Pavesi, 1994 and 1996) contained in the corpus were identified. Scores of examples were reduced to what were considered the best in terms of technical clarity and subsequently packaged into 170 clips (.mpg files averaging between 8 to 15 seconds in length). These clips were then grouped according to four categories of cultural and/or linguistic drops in voltage, classified as follows: (1) Culture-specific references (e.g. references to food and drink; weights and measures; place names; institutions, etc.). (2) ‘Lingua-cultural drops in translational voltage’ (e.g. rhymes; songs; proverbs; puns, etc.). (3) Language-specific features. (4) Purely visual cultural specificity. Perception of Dubbing by Italian Audiences 103 Finally, a short introduction was prepared for each clip in html. The aim of these short texts was to contextualise each clip. Selecting language-specific features As this chapter is concerned particularly with viewers’ perception of language-specific features, what follows is an account of the methodology adopted for the part of the questionnaire which dealt expressly with these features. When choosing clips from the complete recorded corpus which were representative of ‘turbulent’ languagespecific features (Category 3) and that adhered to norms that were highly discernible to translators and researchers, the final choice fell on a number of elements. These included the norms for the following features: (1) Terms of address. Unlike the other Source Languages in the corpus, English does not express politeness, courtesy and familiarity via its second person pronoun like Italian, which markedly connotes familiarity and/or social distance. Consequently such difficulty necessitates firm translational strategies regarding not only personal pronoun forms, but also the substitution of a large range of terms of endearment, titles, names, etc. (Pavesi, 1994). (2) Taboo language. Swear words were explicitly selected from the corpus owing to the fact that Italian tends both to under-translate such words and insert fewer than occur in the original (Pavesi and Malinverno, 2000). (3) Words left spoken in the original language. (4) Written scripts left in the original language: signs, letters, notes, newspapers, etc. (5) Exclamations. (6) Affirmatives. The English affirmatives ‘yes’, ‘yeah’ and German ja create problems of lip synchronisation if left translated with the customary Italian word for asserting sì and are thus normally translated with the term già. The questionnaire A questionnaire-style test, in which questions were based on a random selection of the video clips, was devised and divided into six sections: the first four each containing one question referring to one clip from each of the four categories under investigation, and the last two comprising general questions on dubbing and subtitling and socio-demographic 104 Rachele Antonini & Delia Chiaro classification questions respectively. One single question was devised for each category of clip in which respondents were asked to: ● ● ● ● Rate their understanding of the first clip concerning culture-specific references – Category 1. Rate their understanding of the second clip regarding lingua-cultural drops in voltage – Category 2. For clips concerning language-specific features, rate the likelihood of such language arising in naturally occurring Italian – Category 3. Rate their understanding of the situation for the clips concerning purely visual cultural specificity – Category 4. Thus each respondent answered four questions regarding four video clips. All questions were of the closed type. In fact, respondents were required to assess each video clip according to a 1–10 point graphic rating scale by simply dragging an arrow and dropping it at the exact point on the scale which most closely corresponded to their evaluation. Having done that, respondents also had the opportunity of explaining what they had understood in their own words. This extra qualitative information proved to be quite precious in interpreting the results of the investigation. Both the questionnaire and the clips with their introductory texts were placed on a website which was developed using ASP and Java script programming languages.4 The site was programmed in such a way that each respondent who agreed to complete the questionnaire was automatically given four clips with their matching question from the categories (listed on page 103). Although the four questions were always identical, the questionnaire was programmed in such a way as to ensure that no respondents would receive the same combination of clips. In fact, questions and clips were automatically selected at random from four separate electronic folders. Thus, whenever a visitor entered the site and agreed to take part in the investigation by completing the questionnaire, programming was such that: (1) They were presented with a short introduction which contextualised each clip they were about to watch. (2) They watched four clips picked at random by the program. (3) They were invited to rate their appreciation of the clips on a 10-point graphic rating scale aimed at assessing their self-reported understanding of the content of the clip. Perception of Dubbing by Italian Audiences 105 (4) They were asked to explain in their own words (in an electronic notepad) what they had understood from each clip. With regards to the question on language-specific features, Category 3, respondents randomly received a clip regarding any one of the six features described above (terms of address, exclamations, etc.). Elaborating raw data regarding language-specific features What follows are the results regarding the sample’s perception of language-specific features. The raw data was elaborated statistically with SPSS (Statistical Package for the Social Sciences).5 In order to measure the Perception of Likelihood of Occurrence (GPLO) of the items in naturally occurring Italian, the respondents’ scores were first analysed in general and subsequently in more detail after collapsing the scores into a dichotomic variable coded as follows: (1) Scores from 0 to 5 " Generally perceived as more towards unlikely to occur. (2) Scores from 5.01 to 10 " generally perceived as more towards likely to occur. Administering the questionnaire The questionnaire was promoted on one of the largest and most popular Italian web providers: Virgilio. An advertisement announcing the questionnaire and inviting surfers to participate was placed as a popunder 6 which appeared on the provider’s home page and also on pages dedicated to cinema, TV, music and books. Between 16th January and 29th February 2004 over 3,000 people visited the website and 2,478 submitted a questionnaire. From these submissions, a total of 195 questionnaires were considered to be complete and hence valid for data elaboration. The reason why so many questionnaires were considered invalid was due to the fact that many people did not complete them with their socio-demographic details. In fact, blocks eliciting attitude towards dubbing and socio-demographic information were positioned last on the questionnaire and presumably many respondents simply watched the clips, answered the questions and then logged off without giving any personal information. Of course, this partial information was useless for the purposes of data elaboration. 106 Rachele Antonini & Delia Chiaro Results The sample The final block of the questionnaire required respondents to give socio-demographic information about themselves and their lifestyles. Respondents were mostly male; 123 males compared to 72 females. This proportion is predictable if we accept the commonplace that the typical user of the Internet in Italy is indeed male; 63.1% males compared to 36.9% females. On the other hand, the majority of questionnaires were actually completed during office hours, thus reflecting the Italian working reality that more men than women are employed outside the home. This too reflects CNEL and EURISKO data which shows that 39.8% of employed Italians use the Internet at the workplace.7 Furthermore, over half the sample were under the age of 30, with 106 of the respondents aged below 50. Now, although this data reflects what we know about Internet use and confirms the researchers’ forecasts regarding the kind of people who were expected to respond, the sample also includes 23 senior citizens. So, even if the sample is skewed in terms of Internet users only, the number of respondents is high enough to be fairly balanced in terms of age. As far as social background is concerned, most respondents were educated at least to secondary school level and above and represented a wide variety of occupations. Finally, the vast majority of respondents were familiar with English, having learnt it either at school or college, or else were self-taught. 150 respondents also claimed to know at least one other of the languages from which the clips in our corpus were translated. Forty respondents had visited the USA. The people in the sample maintained that they watched an average of 4.6 hours of TV a day and that they had a preference for films, sitcoms, sci-fi and cartoons. In terms of quality of Italian dubbing, 34.9% of the sample considered dubbed products to be good, 30.8% judged them to be satisfactory and 11.8% excellent. 15.9% of the sample found their quality to be mediocre and 6.7% considered them to be poor. Furthermore, 40.5% of respondents favoured having a choice between dubbing and subtitling, a choice which nowadays is only available in Italy on pay TV and some programmes broadcast by satellite channels. The general perception of language-specific features Likelihood versus unlikelihood The distribution of the general score of all the features taken together is quite normal around the central point of the proposed scale of Perception of Dubbing by Italian Audiences 107 measurement (0–10) with a mean of 5.19 and a standard deviation of 2.69. After having collapsed this variable into a dichotomic one, results show that 43% of the sample perceive the language features in question ‘more towards unlikely to occur’ in natural Italian while 56.9% considered them to be ‘more towards likely to occur’. Crossing this variable with socio-demographic data, it was found that the former view is strongly correlated with age and education as can be observed in Tables 8.1 and 8.2. Respondents above the age of 50 appear to be more oriented towards unlikelihood and the same pattern was observed for the more highly educated members of the sample. That older people found the language more ‘unlikely’ (69.6%) than younger members of the sample may well be due to the fact that they are less likely to have undergone the process of internationalisation (see Table 8.1). In other words, younger people are more likely to have Crosstabulation of Age and GPLO Table 8.1 GPLO AGE Below 30 Between 30 and 50 Above 50 Total Table 8.2 Count % within Age Count % within Age Count % within Age Count % within Age Unlikely Likely Total 42 39.6% 26 39.4% 16 69.6% 84 43.1% 64 60.4% 40 60.6% 7 30.4% 111 56.9% 106 100% 66 100% 23 100% 195 100% Crosstabulation of Education and GPLO GPLO EDUCATION Non university educated University educated Total Count % with in Education Count % with in Education Count % with in Education Unlikely Likely Total 55 38.2% 89 61.8% 144 100% 29 56.9% 22 43.1% 51 100% 84 43.1% 111 56.9% 195 100% 108 Rachele Antonini & Delia Chiaro been influenced by having grown up in a more globalised context, surrounded by the lyrics of international music and the slogans of global advertising. Furthermore, they are likely to have learnt at least one foreign language at school. Thus, having been more exposed to foreign linguistic forms, younger people are likely to be linguistically aware and may well be able to recognise foreign forms seeping through into Italian forms. The fact that the more educated members of the sample find Italian dubbese is self-evident if we presume that education correlates with more linguistic competence and awareness. Perception of specific language features Elaboration of data revealed that four of the features had a mean which was greater than 5 (già, exclamations, bad language and utterances left in the original language) and two features had a mean which was less than 5 (terms of address and written scripts left in the original language). These findings appear to indicate that the former cluster of items are judged as leaning more towards likeliness of occurrence, while the latter two of the features are judged as leaning more towards unlikeliness of occurrence (Figure 8.2). Number of cases 10 34 35 41 34 25 34 Mean scores 8 5.47 5.37 6 4.99 5.45 5.17 4.59 4 2 0 Già Original language Terms of Bad language Non Exclamations address translated written Figure 8.2 Comparison of mean scores and standard deviations* of viewer perception of lingua specific translational uneasiness * The vertical line inserted within each bar represents the standard deviation. Perception of Dubbing by Italian Audiences 109 The features considered more unlikely to occur are: Untranslated written scripts: These features obtained the lowest mean score of all those examined, averaging 4.59. Predictably, signs, letters, newspapers etc. in a foreign language were not seen as much of a likely occurrence in everyday Italian life. This information is not always translated into Italian and comments in the notepads revealed a very low understanding of instances of writing in the clips. The problem is that some inserts contain vital information which is denied to viewers unfamiliar with the Source Language. Thus, with regard to audiences’ rights to be able to be fully privy to a TV product, those who do not know foreign languages appear to be somewhat short-changed. Terms of address: The translational norms for adjusting English screen texts for Italian, a language which connotes social rules regarding distance and familiarity through the use of the pronouns tu and Lei are complex and haphazard (Pavesi, 1994). For example, Italian dubbese frequently adopts the use of Lei (polite form) with a person’s first name, thus jarring both socially and linguistically. Questions on clips exemplifying mismatches in social distance between interlocutors and terms of address achieved a mean score of 4.99 in terms of likelihood of arising in naturally occurring Italian. So, they were perceived as being on the borderline between likeliness and unlikeliness of occurrence. However, the qualitative data supplied by respondents in the notepads provided significant input. For example, several respondents commented on the unusual occurrences of the terms amico [friend] which is the norm for translating the US ‘guy’; Signore [Sir]; figliuolo [son] and even cara suocera [dear mother in law]. Thus, the scores reveal that respondents seemed perfectly aware of the fact that amico, Signore and figliuolo are not used in naturally occurring Italian. The features considered more likely to occur are: Exclamations: Respondents were questioned on their perception of likeliness of occurrence in Italian of three English exclamations which remain untranslated in dubbed programmes: ‘wow’, ‘hey’, and ‘oh oh’ (in the sense of ‘Oh no, something’s wrong’). No incidence of the expressions were found in home-made fictional TV programmes, although the term ‘wow’ does indeed occur in institutionalised TV talk such as talk shows, reality TV, advertisements and so on. Furthermore, the LIP (Lingua Italiana Parlata) corpus gave us zero occurrences of ‘wow’ and ‘oh oh’ and only five concordances for ‘hey’ (spelt ehi), two of which occurred in non-standard Italian (Neapolitan dialect). Respondents 110 Rachele Antonini & Delia Chiaro gave exclamations a mean score of 5.17 thus deeming them moderately likely to occur in natural Italian. As far as the exclamation ‘wow’ is concerned, several older respondents informed us that although it was not part of their own linguistic repertoire, it was certainly used by their children. Untranslated spoken forms: These forms achieved a mean score of 5.37. Terms which remained untranslated (‘condom’, ‘take away’) appear to have been generally understood by recipients. Several commented on the use of the term ‘condom’, claiming that they understood it even if it was not an Italian word and thus did not find the term unusual in any way. On the other hand acronyms created a certain amount of perplexity amongst respondents. For example, a clip from the sitcom Willy the Fresh Prince of Bel Air featured Will Smith asking for a BLT sandwich. Despite plenty of clues from the context no respondents understood that the acronym refers to ‘Bacon, Lettuce and Tomato’. Taboo language: The sum of the score of respondents’ perception of taboo words translated from English resulted in a mean of 5.45. In a certain sense the score is extremely generous if we consider that none of the translated taboo words actually exist either in everyday Italian or ‘screen’ Italian; this is because the Italian norm is firstly, mainly to tone down the force of swear words and vulgarities and secondly, to adopt invented conventions (Pavesi and Malinverno, 2000). For example, the stock translation for the insult ‘fucking bastard’ would be the literal fottuto bastardo, a form of abuse which is not part of naturally occurring Italian usage. However, although respondents gave a somewhat lukewarm response, comments in the notepad clearly leant towards the unusualness of translational choices. Già: This feature gained the highest score with a mean of 5.47. In other words it was fairly well accepted as a naturally occurring form. In Italian dubbese, già, an adverb normally used to mean ‘already’, substitutes for UK English ‘yes’, US English ‘yeah’ and German ja. For obvious reasons of lip synchronisation the use of sì, pronounced with spread lips, would be out of the question as a substitute for the pronunciation of unrounded, jaw lowered ‘yeah’ and ja. Thus, già has become the conventional substitution. Such is the force of this norm that già appears to be just as frequent in programmes dubbed from Hispanic languages in which ‘yes’ is sí. In an examination of a parallel corpus of home-made Italian programmes, già is used almost exclusively as an intra-sentential adverb meaning ‘already’ and is rarely used independently. A query on the LIP corpus resulted in 641 concordances with only five occurrences of già Perception of Dubbing by Italian Audiences 111 used alone in assertion of the previous utterance. Furthermore, the word never stands alone, in all five instances it is either preceded by e as in e già or ah as in ah già. It would appear from our dubbed corpus that it is just in Italian dubbese that già stands alone. In addition, unquantifiable observation reveals that the form may be slowly seeping through to the language of home-produced Italian TV fiction. In fact, while going into print the authors noticed an increasing occurrence of independently used già in affirming function in several Italian-produced series, serials and soaps (Orgoglio a mini-series screened on RAI1 and Un posto al sole an Italian soap opera).8 Conclusion The two features which respondents thought were most likely to occur in naturally occurring Italian were già and taboo words. Although these scores were interpreted as an acceptance of likelihood of occurrence in naturally occurring Italian, at the same time they are hardly awe-inspiring. In fact, the scores in all four more ‘positive’ categories are rather middle-of-the-road. Had the scores been lower, then translations could be discarded as being unnatural and consequently of poor quality. Conversely, if scores had been higher they would have pointed towards excellence. However, as scores stand they point towards average acceptance. This automatically raises the question: does this imply average quality? However, it was decided to precede the terms ‘likelihood’ and ‘unlikelihood’ with the phrase ‘more towards’ which modulates the strength of perceived likelihood or unlikelihood of occurrence. Furthermore, the qualitative data revealed a marked awareness of the fact that what respondents were watching was ‘television language’ and thus acceptable as such even if it was somewhat removed from the reality of everyday speech. Although none of the features were actually rejected, in other words respondents generally gave all elements a pass mark in terms of likelihood of occurrence, they were hardly convinced of their Italianness. If Italian dubbese is to be regarded in terms of excellence, then operators are slightly off target. However, if numeric data is checked against qualitative data, what emerges is that a significant number of viewers who are perfectly aware of the fact that, for example, già is a convention, are willing to accept it on screen but admit to not using it themselves. Our data reveal that the Italian dubbing industry appears to be producing a nation of viewers who are suffering from a syndrome of 112 Rachele Antonini & Delia Chiaro linguistic bipolarity. On the one hand they are aware that TV dubbese is unlike real Italian, on the other hand they are willing to accept it but as long as it remains on screen. Respondents allege that they do not use these forms yet there may well be a mismatch between what they think they do and what they actually do. However, the fact that they notice their children using expressions like ‘wow’ coupled with an awareness that they do not and would not use the form themselves is a clear signal of some kind of language awareness. Finally, operators in the Italian dubbing industry should pay heed to these results and consider investing in human resources who are able to mediate between cultures in such a way as to render Italian dubbese more similar to naturally occurring Italian. If this were possible, the future repetition of such an experiment could then budge scores slightly upwards so as to seriously point towards excellence in the long run. Notes 1. The Andersen report prepared on behalf of the European Commission Directorate-General for Education and Culture entitled Outlook for the Development of Technologies and Markets for the European Audio-Visual Sector up to 2010 (June 2002), provides a detailed nation-by-nation analysis of trends in European viewing habits. 2. The Univideo Executive Summary (June 2003), elaborated by SIMMACO Management Consulting, provides data regarding purchasing trends in both the Italian and the European DVD sector as a whole. Data clearly show that the DVD player and disc market is the fastest growing, and consequently the most commercially significant, in the European home entertainment sector (www.univideo.org/dossier/studi.asp). 3. A decision to exclude products for the big screen was quite deliberate and mainly due to the fact that, at least in Italy, while technically speaking dubbing processes are similar for both types of programmes, for many feature films more time is dedicated to the whole process and often special actors are involved (Benincà, 1999). Indeed some directors are extremely concerned with the translational process (Stanley Kubrick and Woody Allen) while other producers go as far as choosing well-known voices (Buena Vista). Although most films are not dubbed with such care and given the fact that more Italians watch TV than go to the movies anyway, the researchers deemed that other products were more typical of what Italians actually watched. Furthermore, it seemed that the stressful work involved in dubbing scores of episodes of everyday TV products would by default produce poorer material in qualitative terms, yet it is what most Italians watch most of the time. 4. The authors wish to thank Piero Conficoni, webmaster at the University of Bologna’s Department of Interdisciplinary Studies in Translation, Languages and Culture (SITLEC), for creating the interactive website and programming the database (www.sitlec.unibo.it/dubbingquality). Perception of Dubbing by Italian Audiences 113 5. Special thanks go to Giuseppe Nocella for supervising the experimental design and data analysis. 6. A pop-under is a window that appears when surfers exit sites they have just visited. Unlike a pop-up, which appears as soon as surfers visit a site, thus running the risk of irritating users by interfering with what they are doing, a pop-under can only be seen once surfers have abandoned the site they have just visited. A pop-under, in fact, remains hidden behind the site in use, and is thus less intrusive than a pop-up. 7. CNEL is the Consiglio Nazionale dell’Economia del Lavoro (National Council for Economy and Work) and EURISKO stands for European Research on Consumption, Communication and Social Transformation. 8. Interestingly, the authors noticed a shift from già standing alone in episodes from the US series Felicity included in the corpus and instances of eh già in new episodes broadcast in June 2004. Furthermore, in August 2004 a popular weekly crossword magazine, La settimana enigmistica, featured the following clue: a three-lettered synonym for ‘yes’. References Alfieri, G. (1994) ‘Lingua di consumo e lingua di riuso’. In L. Seriani and P. Trifone (eds) Storia della lingua italiana. Vol. II, Scritto e parlato (pp. 161–235). Torino: Einaudi. Antonini, R., Bucaria, C. and Senzani, A. (2003) ‘It’s a priest thing, you wouldn’t understand: Father Ted goes to Italy’. Antares VI: 26–30. Baccolini, R., Bollettieri Bosinelli, R.M. and Gavioli, L. (eds) (1994) Il doppiaggio. Trasposizioni linguistiche e culturali. Bologna: CLUEB. Benincà, S. (1999) ‘Il doppiaggio: uno studio quali-quantitativo’. Quaderni del Doppiaggio II (pp. 53–148). Finale Ligure: Comune di Finale Ligure. Bollettieri Bosinelli, R.M., Heiss, C., Soffritti, M. and Bernardini, S. (eds) (2000) La traduzione multimediale. Quale traduzione per quale testo? Bologna: CLUEB. D’Amico, M. (1996) ‘Dacci un taglio, bastardo! Il doppiaggio dei film in Italia’. In E. Di Fortunato and M. Paolinelli (eds) Barriere linguistiche e circolazione delle opere audiovisive: la questione doppiaggio (pp. 209–16). Roma: AIDAC. Fuentes Luque, A. (2001) ‘Estudio empírico sobre la recepción del humor audiovisual’. In L. Lorenzo García and A.M. Pereira Rodríguez (eds) Traducción subordinada (II). El subtitulado (pp. 85–110). Vigo: Universidade de Vigo. Gallozzi, G. (2004) ‘Il silenzio dei doppiatori’. L’Unità, 18th February. Gambier, Y. (ed.) (1998) Translating for the Media. Papers from the international conference ‘Languages and the Media’. Turku: University of Turku. Gambier, Y. and Gottlieb, H. (eds) (2001) (Multi)Media Translation. Amsterdam and Philadelphia: John Benjamins. Heiss, C. and Bollettieri Bosinelli, R.M. (eds) (1996) La traduzione multimediale per il cinema, la televisione e la scena. Bologna: CLUEB. Ivarsson, J. (1992) Subtitling for the Media. Stockholm: TransEdit. Karamitroglou, F. (2000) Towards a Methodology for the Investigation of Norms in Audiovisual Translation. Amsterdam: Rodopi. Pavesi, M. (1994) ‘Osservazioni sulla (socio)linguistica del doppiaggio’. In R. Baccolini, R.M. Bollettieri Bosinelli and L. Gavioli (eds) Il doppiaggio. Trasposizioni linguistiche e culturali (pp. 51–60). Bologna: CLUEB. 114 Rachele Antonini & Delia Chiaro Pavesi, M. (1996) ‘L’allocuzione nel doppiaggio dall’inglese all’italiano’. In C. Heiss and R.M. Bollettieri Bosinelli (eds) La traduzione multimediale per il cinema, la televisione e la scena (pp. 117–30). Bologna: CLUEB. Pavesi, M. and Malinverno, A.L. (2000) ‘Usi del turpiloquio nella traduzione filmica’. In C. Taylor (ed.) Tradurre il cinema (pp. 75–90). Trieste: University of Trieste. 9 Transfer Norms for Film Adaptations in the Spanish–German Context Susana Cañuelo Sarrión Introduction This chapter summarises some findings on the relationships between cinema, literature and translation in the Spanish-German context.1 Firstly, I will show how the combination of three transfer forms – film adaptation, literary translation and audiovisual translation – can be systematically disentangled by using Polysystem Theory as an epistemological framework (Even-Zohar, 1990). Secondly, I will present a corpus of Spanish film adaptations and analyse its reception in Germany by applying a working model, that I call systematisation, which takes into account several selection factors. Finally, I will examine the extent to which the data allow for the detection of transfer norms. The systematisation model The systematisation model I propose in these pages has been designed to illustrate and explain all possible combinations of the three forms of transfer that take place when a literary work is filmed and both the book and the resulting film are translated into another language. Figure 9.1 illustrates the pattern of relationships. Figure 9.1 contains six elements: three literary works and three films. They are located in six systems (three literary systems and three cinematographic systems), which should be defined in each case depending on the works and cultures involved. This diagram not only takes into account the possible relations established between a given pair of languages, but also considers the possibility of using an intermediate 115 116 Susana Cañuelo Sarrión S1 LT1 SX LTX SY FX SK LTY SZ FZ SQ FY S: System LT: Literary Text F: Film Figure 9.1 or pivot language to carry out the audiovisual or literary translation. The transfer processes are represented by arrows. The broken arrows represent the literary translation processes, the solid arrows represent the audiovisual translation processes, and the dotted arrows represent the adaptation processes from literary work to screen. There are also mixed arrows (dash-and-dot), which indicate that the intersemiotic transfer has been combined with a linguistic one, that is, the film adaptation has been accompanied by a linguistic translation. The process always begins in the literary system in which the film adaptation has been done (LT1). The next steps vary from case to case and language to language; therefore, instead of using numbers, the other systems are marked with variables (x, y, k). From this diagram we can derive five possible basic combinations, from which other secondary alternatives can be also derived: Combination I: literary translation before film adaptation. Combination II: film adaptation before literary translation. Combination III: film adaptation starting from a translation. Combination IV: intermediate literary translation. Combination V: intermediate audiovisual translation. The diagrams corresponding to these five combinations are included as Figures 9.9–9.13 in the Appendix to this chapter. Some examples are Transfer Norms for Film in Spanish–German Context 117 discussed in the section below entitled Combinations where I present the most common combinations, followed by the Spanish film adaptations that have made it to Germany. These examples help to illustrate how each combination should be interpreted. Spanish production of film adaptations The corpus of film adaptations consists of all those Spanish feature films produced between 1975 and 2000 that are based upon Spanish literary works. The term ‘film’ refers to only those fictional full-length films produced for the cinema. I do not include in this concept short films, documentaries or TV-movies. ‘Spanish’ should be understood as relating to Spain, not to all Spanish-speaking countries. The corpus includes 311 film adaptations. Figure 9.2 shows the relationship between Spanish general film production and the production of film adaptations. Whereas general film production has experienced many upheavals, the production of film adaptations has remained comparatively stable. On average, literary adaptations represent 22% of the total film production in Spain, but only 15% if we consider exclusively the adaptations of Spanish literary works. One initial obstacle encountered in this study is the absence of any central institution from where to gather data on the transfer of Spanish films to Germany. There is no public institution, company, or corporation (neither in Spain nor in Germany) that systematically collects data on the reception of Spanish films in Germany. Several 160 140 120 100 80 60 40 20 19 7 19 5 7 19 6 7 19 7 7 19 8 7 19 9 8 19 0 8 19 1 8 19 2 8 19 3 8 19 4 8 19 5 8 19 6 8 19 7 8 19 8 8 19 9 9 19 0 9 19 1 9 19 2 9 19 3 9 19 4 9 19 5 9 19 6 9 19 7 9 19 8 99 0 Total of Spanish Films Spanish Film Adaptations from Spanish Literary Works Figure 9.2 118 Susana Cañuelo Sarrión Festivals 25% Cinema 21% DVD 1% Video 17% TV 36% Figure 9.3 14 12 12 10 9 8 8 9 8 7 6 6 6 6 6 5 4 3 3 2 2 2 1 0 0 1 2 3 2 1 0 19 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 1995 1996 1997 1998 2099 00 0 2 Figure 9.4 sources, including the Spanish ICAA (Instituto de Cinematografía y de las Artes Visuales) and the German SPIO (Spitzenorganisation der Filmwirtschaft), have given me information, but none of the lists provided is exhaustive. However, I have gathered sufficient data on the presence of Spanish films in Germany, mainly thanks to the Lexikon des internationalen Films (2000), which provides interesting information and helps to trace provisional transfer norms. According to the Lexikon des internationalen Films and its online database (www.filmevona-z.de), 410 Spanish feature films reached the German film industry during the period 1975–2000. However, this figure only refers to works distributed in commercial cinemas, television, video, and DVD. Other films were also shown at film festivals and Spanish weeks, where approximately 150 Spanish films were screened. Since 61 of the films were distributed in Germany in commercial ways and at festivals, the total number of Spanish films that reached the Transfer Norms for Film in Spanish–German Context 119 German market between 1975 and 2000 can be set at 499. Figure 9.3 shows the different ways in which they were distributed. The film adaptations of this corpus represent a small percentage – 15% – of the 499 Spanish films seen in Germany. They number 77 and their distribution by year is shown in Figure 9.4. There does not seem to be an obvious pattern of regularity in the annual reception of Spanish adaptations in Germany, and perhaps the most interesting factor is the constant ups and downs suffered in the number of imports. In contrast to the distribution of general films in Germany, most Spanish literary film adaptations did not enter the country by TV but through festivals, where 43% of the productions were seen. 34% of them were shown on television, 13% in cinemas and the remaining 10% were sold on VHS and DVD, as can be seen in Figure 9.5. Cinema 13% Festivals 43% DVD 1% Video 9% TV 34% Figure 9.5 Audiovisual translation in Germany Germany is traditionally a dubbing country, which means that films distributed through commercial channels are mostly dubbed. After the Second World War, cinema was used by the Allies to denazify the occupied territory and numerous American, French, British and Soviet films were imported. This activity led to the creation of four large dubbing centres in the country: France dubbed its films in Munich, Great Britain in Hamburg, the United States in West Berlin and the Soviet Union in East Berlin. Nowadays, the German dubbing industry is firmly established and is considered one of the best in the world. The most important dubbing studios continue to be located in Munich, Hamburg and Berlin. Films exhibited at festivals, on the other hand, are mostly subtitled. Sometimes in German but in most cases in English, since subtitles in 120 Susana Cañuelo Sarrión this language facilitate the distribution of the film all over the world and is frequently a requirement of the submission instructions of many festivals. Combinations Out of the 311 film adaptations of the original corpus, only 77 (25%) have been distributed in Germany. Of those 77, there are 39 whose corresponding literary work has not been translated into German and cannot therefore be assigned to any combination of the models discussed here. I should point out that the systematisation only considers combinations in which the three transfer processes are present, that is, film adaptation, audiovisual translation and literary translation. So, there are only 38 films which can be taken into consideration and classified as follows: Combination I 13 Combination II Combination III Total 22 3 38 A total of 13 film adaptations have followed Combination I, where the translation of the literary work is prior to the film adaptation. A literary text in Spanish is translated into German, once or more times. Later, the work is turned into a film in the Spanish cinematographic system. Finally, the film is either dubbed or subtitled into German and distributed in Germany. New translations of the literary work can also be done after the release of the film adaptation. Figure 9.6 shows the combination corresponding to the film La casa de Bernarda Alba (Mario Camus, 1987). Other examples are: Pascual Duarte (Ricardo Franco, 1976), Bodas de sangre (Carlos Saura, 1981), La colmena (Mario Camus, 1982) and El perro del hortelano (Pilar Miró, 1996). In this group, we find, among others, four adaptations of works originally written by Federico García Lorca, two by Camilo José Cela, and one by Ramón Del Valle-Inclán, Miguel de Cervantes, Félix Lope de Vega and Vicente Blasco Ibáñez. One of the main considerations is that they are some of the most celebrated and well known Spanish authors, and the works that have been adapted to the screen were written in past centuries or in the first half of the twentieth century. Therefore, it is not surprising that literary translations existed prior to the release of the film adaptations and that, in most cases, new literary translations have been published after the distribution of the films. Transfer Norms for Film in Spanish–German Context S1 S2 Bernarda Albas Haus E. Beck, 1950, 1968, 1978 F.R.Fries, 1987 1 La casa de Bernarda Alba Federico García Lorca, 1936 4 Bernarda Albas Haus E. Beck, 1992 H.M. Enzensberger, 1999 2 S3 S4 3 La casa de Bernarda Alba Mario Camus, 1987 Bernarda Albas Haus DubG, 1991 (TV: West3) Figure 9.6 S1 “El Sur”, in El Sur Adelaida García Morales, 1983 S4 3 Der Süden, A. Sorg-Schumacher & I. Bergmaier, 1989, 1992 (2nd ed.) 1 S2 El Sur, Víctor Erice, 1983 Figure 9.7 2 El Sur (Der Süden), SubG, 1984 (Spanish Cinema Week in Dortmund) DubG, 1985 (Cinema) S3 121 122 Susana Cañuelo Sarrión Twenty-two film adaptations have followed Combination II. First, a Spanish literary text is adapted to the screen in the Spanish cinematographic system. Subsequently, the feature film is subtitled or dubbed into German, and a translation of the literary text is later done in German. In this case, no translation of the literary work existed prior to the film adaptation. Examples include: Los santos inocentes (Mario Camus, 1984), Tiempo de silencio (Vicente Aranda, 1986), Las edades de Lulú (José Juan Bigas Luna, 1990), ¡Ay, Carmela! (Carlos Saura, 1990), El maestro de esgrima (Pedro Olea, 1992), and Historias del Kronen (Montxo Armendáriz, 1995). In most cases, Combination II encompasses film adaptations made shortly after the publication of the literary work in Spain, which explains why there are no prior literary translations in German. Figure 9.7 shows the combination corresponding to El sur (Víctor Erice, 1983). Only three of the translated movies followed Combination III: I Picari (Los alegres pícaros, Mario Monicelli, 1988), Uncovered (La tabla de Flandes, Jim McBride, 1994) and The Ninth Gate (La novena puerta, Roman Polanski, 1999). These are all Spanish co-productions based on Spanish literary works, but filmed in Italian (the first one) and English (the other two). It is difficult to ascertain which was the text that the screenwriter used to work on the adaptation. However, considering their nationalities, it is most likely that they used the translations of the literary works rather than the originals in Spanish. As for the audiovisual translation, it is also likely that all the productions have been S1 El Club Dumas, Arturo Pérez Reverte, 1993 S3 2 1 Der Club Dumas C. Schmitt, 1995, 1997 The Dumas Club, S. Soto, 1996, 1998 5 S2 Die neun Pforten C. Schmitt, 2000 3 La novena puerta DubS/SubS, 1999 The Ninth Gate Roman Polanski, 1999 S1: Spanish literary system S2: German literary system S3: British literary system S4: Spanish cinematographic system S5: German cinematographic system Figure 9.8 S4 4 Die neun Pforten, DubG, 1999 (Cine) DubG, 2000 (Video) DubG/SubG, 2000 (DVD) DubS: Dubbed into Spanish SubS: Subtitled into Spanish DubG: Dubbed into German SubG: Subtitled into German S5 Transfer Norms for Film in Spanish–German Context 123 translated from the Italian and English versions of the film. Figure 9.8 shows the combination corresponding to The Ninth Gate. It is interesting to notice that the title of the second literary translation, made after the release of the film, has been changed in order to match the title of the film, although the translator remains the same. Literary genres As may be expected, there is a clear predominance of the narrative genre – both in the original Spanish corpus (77%) and in the corpus of films translated into German (78%) – since commercial films are heavily reliant on narrative. This suggests that this genre is the easiest both to be adapted to the cinema and possibly to be translated into other languages. By contrast, of all the Spanish adaptations distributed in Germany only one is based on a work of poetry (A un dios desconocido, Jaime Chávarri, 1977). Film genres As for the film genres, the majority of adaptations made in Spain are dramas (53%), followed by comedies (27%), thrillers (12%), and adventure/action films (4%). A very similar ranking represents the adaptations that have made it to Germany: 58% are dramas, 16% comedies, 18% thrillers, and 5% action films. Worth noticing is the slight decrease in the number of comedies transferred into German, since normally humour does not travel well abroad. If we consider the total of Spanish films shown in Germany, the number of comedies is also lower, whereas thrillers fare considerably better. Literary authors In the case of the original authors of the literary works, there seems to be no evidence whatsoever of a direct relation between the number of adapted works into Spanish films and the number of film adaptations that have been translated into German. For example, six of Fernando Vizcaíno Casas’s literary works have been adapted for the screen in Spain, but none of the films has been ever distributed in Germany. And Miguel Delibes, who is the Spanish writer with most adaptations in the original corpus (9), has only seen one of the films distributed in Germany. It seems plausible that films based on literary authors who are well known in Germany are also better represented in German 124 Susana Cañuelo Sarrión cinematography. Indeed, this is the case with writers like Federico García Lorca, Alberto Vázquez Figueroa and Arturo Pérez Reverte, whose works are very popular among German readers. Thus, of the five Spanish film adaptations made from works by García Lorca, four have been released in Germany. In the case of Vázquez Figueroa, out of six films based on his novels, four have made it to Germany and Pérez Reverte has had five works adapted to the screen, of which four have been released in Germany. Film directors When we look at the different film directors, there is no clear relation between the number of adaptations made in Spain and the number of adaptations that have been exported to Germany. For instance, Francisco Betriú and Vicente Escrivá have both directed five adaptations each, but none of them has reached the German market. Similarly, Rafael Gil and Pedro Lazaga have done eight and six adaptations respectively, and none of them has been seen in Germany. Vicente Aranda is the director with the lion’s share in the corpus, but only four of his ten adaptations have been translated into German. Mario Camus, with eight adaptations in the Spanish corpus, is only represented with four among the adaptations released in Germany. Imanol Uribe, Gerardo Herrero, Rafael Romero Marchent and Pilar Miró, on the contrary, are the directors who have a higher proportional representation both in the Spanish corpus and among the adaptations exported to Germany (5–3, 4–3, 5–3 and 4–3 respectively). Although Carlos Saura is not the director with the highest number of adaptations in the Spanish corpus (only three), all of his works have been shown in Germany. That is hardly surprisingly, since Saura is one of the most famous Spanish directors in Germany, mainly thanks to his musical trilogy with the dancer Antonio Gades (Bodas de sangre, 1981; Carmen, 1983; El amor brujo, 1986). Saura is the director with most films (23) distributed in Germany in the period 1975–2000. He is followed by another of the most international Spanish directors, Pedro Almodóvar, of whom twelve productions have made it to Germany in the same period. Conclusion In the transfer of Spanish film adaptations from Spain to Germany there are some recurrent patterns that invite us to think about the existence of transfer norms; norms that come to the surface when we Transfer Norms for Film in Spanish–German Context 125 consider the three transfer processes discussed above: film adaptation, literary translation and audiovisual translation. Some norms are operative in the combinations followed by the film adaptations on their way to the German market. Of the five potential patterns outlined, most films followed the first two combinations (I and II), whereas combinations IV and V were not present at all. These norms can be called ‘combinatory norms’ and tell us about the steps followed in the process of transfer from one culture to another and about the languages involved. These norms do not seem to fit under the category of ‘preliminary norms’, defined by Toury (1995: 56–61) as those that operate before the translational process of transfer itself takes place. If we agree that the selection of films does not happen by chance, these norms reflect rather the existence and nature of a concrete transfer strategy, that is, the selection of elements of the Source System and the systemic mechanisms that govern this selection. Some regularities have been detected in the case of the preferred literary genres to be adapted to the screen (narrative, with 78%) and the film genres more likely to be translated (drama, with 58%). As regards the selection of directors and literary authors, there is no evidence of a direct relation between the number of adapted works in Spanish and the number of transferred adaptations into German, but it is noteworthy that some of them are more popular than others: Federico García Lorca, Alberto Vázquez Figueroa, Arturo Pérez Reverte, among the writers; and Carlos Saura, Imanol Uribe, Pilar Miró, among the filmmakers. However, further investigation into the mechanisms of export and import in the Spanish-German film context is needed in order to establish if a transfer policy really exists at this level. Regarding the reception of the Spanish films in Germany, no evidence of recurrent patterns has been found in the yearly distribution of adaptations, but there seems to be some regularity in the distribution channels (43% festivals, 34% TV) and the audiovisual translation modes implemented (dubbing for cinema, video and TV; subtitling for festivals). Another relevant fact is that Spanish film adaptations represent a small percentage (15%) of the total of Spanish films shown in Germany between 1975 and 2000. Nevertheless, in-depth research into the particularities of distribution and exhibition is needed in order to define the position and function of Spanish film adaptations in the German system. I have not dealt with operational norms, which govern all the linguistic decisions made during the individual processes of the transfer (Toury, 1995: 56–61). A comprehensive analysis of the three processes – that is film adaptation, literary translation and audiovisual translation – 126 Susana Cañuelo Sarrión involved in the transfer of the 38 Spanish works to Germany is needed. Such an approach will allow the researcher to discover some of the operational norms followed and to establish how the ‘combinatory norms’ may have influenced the different decisions taken during each of the individual transfers. Note 1. This is fully discussed in my PhD dissertation supervised by Dr Luis Pegenaute at the Universitat Pompeu Fabra in Barcelona. Between 2002 and 2004 the research was carried out at the University of Leipzig thanks to a scholarship from the foundations La Caixa and DAAD (Deutscher Akademischer Austausch Dients). I am grateful to them for the financial support and to the Institute for Applied Linguistic and Translation Studies in Leipzig, especially to Prof. Dr Wotjak for his assistance. Thanks also to Dr Patrick Zabalbeascoa and Dr Aline Remael for their suggestions. References Even-Zohar, I. (1990) Polysystem Studies. Special Issue of Poetics Today 11(1). www. tau.ac.il/~itamarez/works/books/ez-pss1990.pdf Filmstatistisches Taschenbuch. Wiesbaden: SPIO, 1976–2000. Filmstatistisches Jahrbuch. (2001) Baden-Baden: Nomos. Lexikon des internationalen Films. (2000) Munich: Net World Vision. Online database: www.filmevona-z.de Toury, G. (1995) Descriptive Translation Studies and Beyond. Amsterdam and Philadelphia: John Benjamins. Transfer Norms for Film in Spanish–German Context Appendix S1 S2 1 LT1 LT2 2 S3 S4 3 F1 F2 S: System LT: Literary Text F: Film Figure 9.9 Combination I: literary translation before film adaptation S1 S3/4 2/3 LT1 LT2 1 S3/4 S2 F1 2/3 F2 S: System LT: Literary Text F: Film Figure 9.10 Combination II: film adaptation before literary translation 127 128 Susana Cañuelo Sarrión S1 LT1 1 S2 LT2 SK LT3 2 SY F1 F2/3 SQ S3 F2/3 S: System LT: Literary Text F: Film Figure 9.11 Combination III: film adaptation starting from a translation S1 LT1 S2 F2/3 LT3 LT2 SY S3 SZ F2/3 S4 F1 S: System LT: Literary Text F: Film Figure 9.12 Combination IV: intermediate literary translation Transfer Norms for Film in Spanish–German Context S1 SX LT2 LT1 SY F1 SK LT3 SZ F2 SQ F3 S: System LT: Literary Text F: Film Figure 9.13 Combination V: intermediate audiovisual translation 129 10 Voice-over in Audiovisual Translation Pilar Orero Introduction Nowadays audiovisual translation (AVT) is a thriving field within Translation Studies. This is, however, a recent development. Although research in the audiovisual field dates back to 1932 (Franco and Orero, 2005), it remained in the realm of Film or Media Studies and it was only in the 1980s that it started to be studied from a translation perspective, within the discipline of Translation Studies. This transition from Film Studies to Translation Studies may account for the blurred terminology in use, the research guidelines and the somewhat unbalanced interest shown in the many modes within AVT. While subtitling and dubbing have been attracting interest in both research and teaching at university level, other techniques such as voice-over have been left aside or not clearly understood, as pointed out by Grigaravič ītė and Gottlieb (1999), Franco (2000), and Gambier (2003). Some scholars have analysed this unbalanced situation in the amount of attention paid to the many communication forms within AVT. Gambier and Suomela-Salmi (1994: 243) suggested the following possible explanation: Up till now, research [in AVT] has mainly been concerned with the subtitling and dubbing of fictive stories/fiction films. In the light of the huge variety of audio-visual communication, this may seem somewhat surprising; in fact, however, it reflects the prevailing orientation in translation theory, which is still highly dominated by literary translation. 130 Voice-over in Audiovisual Translation 131 One could argue that most research carried out in AVT has concentrated on dubbing and subtitling because these are the modes used to translate fictive stories and fiction films, objects of study which tend to be favoured by academics. However, it seems to have been forgotten in many studies that in some countries like Poland and the Baltic States voice-over is used as the translation mode for films. The traditional stance in most academic studies has been to associate voice-over with the translation of documentaries covering topics like nature and travel (Luyken et al., 1991). The general trend has been to consider it as a technique suitable for the translation of non-fictional genres, an approach that Franco (2000: 3) regrets: Translating reality must inevitably be a straightforward, nonproblematic activity. What such a belief implies is that translated foreign material within non-fictional output (e.g. interviews in news and documentaries and sometimes commentaries as well) constitutes uninteresting data for the purposes of research. Traditionally claimed as objective, deprived of the artifices of literary language or cinematic invention, factual programmes would and could not represent any real challenge to the translator or stimulus to the researcher; in sum, the translation of ‘real life’ would constitute a boring field of study. Still, voice-over is normally included when scholars want to provide a taxonomy of the many – and not universally used – audiovisual translation modes. In some cases, voice-over has been classified within the technique of dubbing by authors such as Luyken et al. (1991: 80), Baker and Braň o (1998: 75), Shuttleworth and Cowie (1997: 45), and Franco (2001: 290). No doubt, simplification and a lack of understanding of the media and its process have made voice-over to be seen in the same light as dubbing, which is certainly a different mode, subject to different translation and production processes. As a result, it comes as no surprise that reference works on AVT have not considered voice-over as a discrete entry (Luyken et al., 1991; Dries, 1995; Baker and Braň o, 1998; Shuttleworth and Cowie, 1997; SánchezEscalonilla, 2003). The unresolved terminology employed in the wider discipline of AVT (Orero, 2004: vii) also applies to the field of voice-over. Many definitions of the term have described voice-over in a misleading or inaccurate form. Thus, it has been referred to as a category of revoicing, lip synchronisation dubbing, narration and free commentary (Luyken 132 Pilar Orero et al., 1991: 71; Baker and Braňo, 1998: 75; O’Connell, 2003: 66); as a type of dubbing, either ‘non-synchronized dubbing’ (Dries, 1995: 9), or its opposite ‘doublage synchrone’ (Kaufmann, 1995: 438). It has also been described as ‘dubbing-with-voice-over’ (Baranitch, 1995: 309), as a type of interpreting (Pönniö, 1995: 303; Gambier, 1996: 8) and finally as ‘half-dubbing’ (Hendrickx, 1984). The process of voice-over has also been described as ‘the easiest and most faithful of the audiovisual translation modes’ (Luyken et al., 1991: 80; Díaz Cintas, 1997: 112). This definition, however, has not helped towards further understanding of the technique and bears little resemblance to the real process of translation. A possible reason for this reputed easiness and faithfulness of voice-over is its alleged disregard for synchronisation between source and target texts, pointed out originally by Luyken et al. (1991: 80) and later by Lambert and Delabastita (1996: 41), Franco (1998: 236), and Grigaravič ītė and Gottlieb (1999). To date, attempts to define voice-over seem to have focused on its reception. That is, voice-over is viewed as the final product we hear when watching a programme where a voice in a different language than that of the original programme is heard on top of the original soundtrack. This new voice is normally recorded some seconds into the beginning of the original utterance – and sometimes finishes before the actual person on screen – allowing the viewer to hear part of the original, although this practice is not universal. These technical dimensions mentioned by numerous translation scholars when defining voice-over may have a common starting point that can be traced back to Luyken et al. (1991: 80) who in their seminal book Overcoming Language Barriers in Television described voiceover not as a complex and highly specialised translation technique but as: The faithful translation of original speech. Approximately synchronous delivery. It is normally used only in the context of a monologue such as an interview response or series of responses from a single interviewee. The original sound is either reduced entirely or to a low level of audibility. A common practice is to allow the original sound to be heard for several seconds at the onset of the speech and to have it subsequently reduced so that the translated speech takes over. This contributes to the sense of authenticity in the translation and prevents a degree of mistrust from developing. Alternatively, if the translation is recorded as Voice-over in Audiovisual Translation 133 part of the original production, it may follow the original speech exactly. While studying the transfer of commercial videos, Mailhac (1998: 222) points to the possible source of inconsistency that surrounds the term voice-over: It should be pointed out that the term ‘voice-over’ is used with the meaning it normally has in English where it refers to the voice of an unseen commentator heard during a film, a television programme or a video. Therefore, it does not correspond to what is called ‘un voice over’ in French, since this refers to situations in which a voice giving a translation is heard simultaneously on top of the original voice (see Pönniö, 1995). The French equivalent of ‘voice-over’ would be commentaire. I have heard the term ‘over-voicing’ to describe the superimposition of a second voice in the context of an interview in a commercial video; ‘half dubbing’ also seems possible to refer to this type of superimposition when applied to a feature film dialogue. And Franco (2000: 32) concludes by stating that: The confusion seems to have arisen from the fact that terms adopted by Audiovisual Translation Studies, such as ‘voice-over’ and others common to factual translation (e.g. ‘commentary’), have all been arbitrarily borrowed from its predecessor Film Studies, whose concepts do not imply any translating activity. A new academic and research approach to voice-over and to AVT in general should take into consideration Film and Media Studies without forgetting traditional research methodology from the field of Translation Studies. It should focus on the translation process as much as on the reception of the audiovisual product, and should also determine the terminology which at present is – to say the least – muddled. In what now follows, I offer a detailed analysis of the two different types of voice-over translation according to the way it is carried out professionally. Voice-over in TV and radio It is true that voice-over is used both in TV and radio, as for example in the BBC World Service. In its news or current affairs programmes, what 134 Pilar Orero is said by someone whose mother tongue is not English is translated and voiced-over into English, in an attempt to convey the feeling of authenticity of the contents (Luyken et al., 1991: 80), the voice of the speaker (Pönniö, 1995: 304), and the accent or regional variation (Fawcett, 1996: 76).1 As Franco (2001: 290) explains: ... the type of delivery we hear in voice-over translation is an important strategic way of reassuring viewers that what they are being told in their own language is what is being said in the original language, although it is known that what they will be listening to is in fact only a representation of the original discourse. In Spain at least, voice-over is a more common mode of audiovisual translation than subtitling, and the leading translation mode when people speaking other languages appear live on TV broadcasts. The programmes range from news and sports to gossip, and even reality shows, and are broadcast on all Spanish TV stations, local as well as national, in Spanish or in any of the other three official languages (Basque, Catalan and Galician); on state-owned stations TV3, 24/3; City TV in Catalonia; TV1 or TV2 in Spain: and on privately owned national channels such as Tele 5 and Antena 3. When voice-over is used for the translation of an audiovisual programme, two voices are usually heard. We hear one voice in the background (the original speech) and the voice of the translation. When there is a speech or sound bite of President Bush or Tony Blair, the translation is usually heard through voice-over. In some exceptional cases three languages can be heard, as is the case when we hear Osama Bin Laden speaking, whilst being translated into English and then into Spanish or Catalan. Cases in which three languages can be heard simultaneously are known as pivot translation (Grigaravičītė and Gottlieb, 1999: 46), a field of study that also merits further research. The compilation of further data on the number and type of programmes as well as on the times of the broadcasts when voice-over production and post-production are used on TV is another much needed research project. Material of this nature will help us to contextualise and to put into perspective voice-over translation versus the two more popular forms of subtitling and dubbing. It would also provide information that could confirm, or reject, the popular association of voice-over and documentaries. A project of compiling data could prove to be challenging, given the large number of TV channels Voice-over in Audiovisual Translation 135 now on offer in Catalonia and Spain. The city of Barcelona alone has over ten TV stations which broadcast in Catalan plus all the TV stations – digital, satellite and terrestrial – which broadcast to mainland Spain. Still a methodological approach to the compilation of reliable data would be needed for any rigorous analysis with some serious conclusions to be undertaken. Translating for voice-over production Much translation for voice-over is carried out in the post-production phase, described by Luyken et al. (1991: 80) as ‘narration or re-voiced narration’, which so far is the mode most widely studied in academic works dedicated to AVT. However, there is an important market for TV and radio programmes with translation undertaken as part of the production process. This second type I refer to as translation for production. A post-production translation is the translation of a programme which is a finished product by the time it reaches the translator; it usually comes with a complete dialogue list, e.g. the BBC’s The Human Body (2001). In the case of translation for production, however, the translator has to work with rough, unedited material which will undergo several processes before being broadcast. The material to be translated can be made of interviews or already existing post-production programmes. When the team of journalists working on a programme are planning its contents and format, they may decide to use excerpts from some existing foreign programmes – regardless of the format or genre. It can be a documentary, sports personality interviews, sports events (World Swimming Championship, Formula 1 Racing, European Champions League football matches), or awards ceremonies (Oscars, MTV), all of which can be either live or a repeat. In order to develop these programmes the team of journalists, all working within the production department, will either buy a number of minutes of an already existing and edited programme or interview, or produce their own interview. If the first approach is chosen, these excerpts will be incorporated in the main body of the programme, either simply to provide images to illustrate the narration, or to provide authentic comments delivered by the actual speakers. For instance, in order to make a programme on the war in Iraq, the journalists may want to use declarations or interviews given in English by Tony Blair or George W. Bush, bought from a news agency; or they may prefer to use 136 Pilar Orero footage of a documentary in English such as Lost Treasures from Iraq,2 which they will buy from an international TV agency for which they only have to provide the topic and the company will offer the images. Alternatively, they may want to have an exclusive interview with the UN weapons chief inspector, Hans Blix, or with Bush’s private weapon inspector David Kay, who resigned on 23rd January 2004 after his report announcing that no weapons of mass destruction had been found in Iraq.3 The translator will then be sent the bulk of material that has been bought (some excerpts of Bush and Blair’s comments to translate, a 30-minute documentary on the Museum of Baghdad, and an in-house produced 35-minute interview with Hans Blix or David Kay). On some occasions the translator may be asked to go to the TV station to carry out the translation in situ. If the physical presence of the translator is not required at the television station, the material will be sent to the translator without a dialogue list. While this can be expected from an unedited home-produced interview, there is no apparent reason why a commercially broadcast programme will only come in audiovisual format, with no transcript. By contrast, the absence of a document with the written dialogue for post-production translation tends to occur only when the producing team buys an excerpt from a broadcast programme from an agency instead of purchasing it directly from the TV producer. Whatever the length and duration of the excerpt, it invariably reaches the translator only in audiovisual format. In Spain, the inclusion of translators in a TV or radio production team may be due to the tight production schedule of TV programmes or to the journalists’ lack of foreign language competence; this is still a major obstacle, perhaps likely to change in the near future, although present university education programmes in Spain have not been designed to address and rectify this particular problem. It should also be pointed out that journalists working from Spanish for British radio and TV stations usually encounter the same problem. Here we have seen some of the characteristic features of production in contrast to post-production voice-over. The material is unedited, it is always urgent,4 and it rarely comes with the written transcript; hence it constitutes a challenge to the listening and comprehension skills of the translator (Orero, 2005). It may be added that the translation will be edited in order to be synchronous in content with the images which it will accompany. A final point concerns the translator who never knows what will eventually be shown in the edited broadcast programme. Voice-over in Audiovisual Translation 137 Conclusion As we have seen in the previous section, the translation for voice-over production is characterised by several constraints: ● ● ● Time: translations have to be done within a very tight schedule. Translating from the screen: a written text to support the audiovisual material is not usually available to the translator. Working with unedited, rough material which will be edited after being translated, not by the translator but by another different professional. In this chapter, I have tried to raise the academic visibility of voice-over in general, and voice-over for production of TV and radio programmes in particular, seeing it as an important translation mode in need of much more scholarly attention, if only to match the popularity it enjoys on TV. Notes 1. Fawcett (1996) explains how, in recent years, TV stations in the UK have incremented the use of voice-over for documentaries replacing the use of standard English for the accent ‘appropriate’ for the programme. 2. www-oi.uchicago.edu/OI/IRAQ/iraq.html 3. ‘US chief Iraq arms expert quits. The head of the team searching for weapons of mass destruction in Iraq, David Kay, has resigned. Mr Kay said he did not believe Iraq possessed large stockpiles of chemical or biological weapons’ (http://news.bbc.co.uk/1/hi/world/americas/3424831.stm). 4. While some research on translation under time pressure has been undertaken in the field of literary translation (Jensen and Jakobsen, 1998; Jensen, 1999) the scope of AVT research remains largely unexplored. References Baker, M. and Braň o, H. (1998) ‘Dubbing’. In M. Baker (ed.) Routledge Encyclopedia of Translation Studies (pp. 74–6). London and New York: Routledge. Baranitch, V. (1995) ‘General situation with electronic media in Belarus’. Translatio, Nouvelles de la FIT-FIT Newsletter XIV(3–4): 308–11. Díaz Cintas, J. (1997) El subtitulado en tanto que modalidad de traducción fílmica dentro del marco teórico de los Estudios sobre Traducción. (Misterioso asesinato en Manhattan, Woody Allen, 1993). Valencia: Universidad de Valencia. PhD Thesis. Dries, J. (1995) Dubbing and Subtitling: Guidelines for Production and Distribution. Düsseldorf: European Institute for the Media. Fawcett, P. (1996) ‘Translating film’. In G.T. Harries (ed.) On Translating French Literature and Film (pp. 65–88). Amsterdam: Rodopi. 138 Pilar Orero Franco, E. (1998) ‘Documentary film translation: A specific practice?’. In A. Chesterman, N. Gallardo San Salvador and Y. Gambier (eds) Translation in Context (pp. 233–42). Amsterdam and Philadelphia: John Benjamins. Franco, E. (2000) Revoicing the Alien in Documentaries. Cultural Agency, Norms and the Translation of Audiovisual Reality. Leuven: KUL. PhD Thesis. Franco, E. (2001) ‘Voiced-over television documentaries. Terminological and conceptual issues for their research’. Target 13(2): 289–304. Franco, J. and Orero, P. (2005) ‘Research on audiovisual translation: some objective conclusions, or the birth of an academic field’. In J.D. Sanderson (ed.) Research on Translation for Subtitling in Spain and Italy (pp. 79–92). Alicante: Universidad de Alicante. Gambier, Y. and Suomela-Salmi, E. (1994) ‘Subtitling: a type of transfer’. In F. Eguíluz, R. Merino, V. Olsen, E. Pajares and J.M. Santamaría (eds) Transvases culturales: Literatura, cine, traducción (pp. 243–52). Vitoria: Universidad del País Vasco. Gambier, Y. (1996) Les transfers linguistiques dans les médias audiovisuels. Villeneuve d’Ascq (Nord): Presses Universitaires du Septentrion. Gambier, Y. (2003) ‘Introduction: screen transadaptation: Perception and reception’. The Translator 9(2): 171–89. Grigaravič ītė, I. and Gotttlieb, H. (1999) ‘Danish voices, Lithuanian voice-over. The mechanics of non-synchronous translation’. Perspectives: Studies in Translatology 7(1): 41–80. Hendrickx, P. (1984) ‘Partial dubbing’. Meta 29(2): 217–18. Jensen, A. and Jakobsen, A. (1998) ‘Translating under time pressure: an empirical investigation of problem-solving activity and translation strategies by nonprofessional and professional translators’. In A. Chesterman, N. Gallardo San Salvador and Y. Gambier (eds) (2000) Translation in Context (pp. 105–16). Amsterdam and Philadelphia: John Benjamins. Jensen, A. (1999) ‘Time pressure in translation’. In G. Hansen (ed.) Probing the Process in Translation: Methods and Results. Copenhagen Studies in Language 24 (pp. 103–19). Copenhagen: Samfundslitteratur. Kaufmann, F. (1995) ‘Formation à la traduction et à l’interprétation pour les médias audiovisuels’. Translatio, Nouvelles de la FIT-FIT Newsletter XIV(3–4): 431–42. Lambert, J. and Delabastita, D. (1996) ‘La traduction de textes audiovisuels: modes et enjeux culturels’. In Y. Gambier (ed.) Les Transferts linguistiques dans les médias audiovisuels (pp. 33–58). Villeneuve d’Ascq (Nord): Presses Universitaires du Septentrion. Luyken, G.M., Herbst, T., Langham-Brown, J., Reid, H. and Spinhof, H. (1991) Overcoming Language Barriers in Television: Dubbing and Subtitling for the European Audience. Manchester: European Institute for the Media. Mailhac, J.-P. (1998) ‘Optimising the linguistic transfer in the case of commercial videos’. In Y. Gambier (ed.) Translating for the Media (pp. 207–23). Turku: University of Turku. O’Connell, E.M.T. (2003) Minority Language Dubbing for Children. Screen Translation from German to Irish. Series XL. Vol. 81. Frankfurt: Peter Lang. Orero, P. (2004) ‘Audiovisual translation: a new dynamic umbrella’. In P. Orero (ed.) Topics in Audiovisual Translation (pp. vii–xiii). Amsterdam and Philadelphia: John Benjamins. Voice-over in Audiovisual Translation 139 Orero, P. (2005) ‘La traducción de entrevistas para voice-over.’ In P. Zabalbeascoa Terran, L. Santamaria Guinot and F. Chaume Varela (eds). La traducción audiovisual: Investigación, enseñanza y profesión (pp. 213–22). Granada: Comares. Pönniö, K. (1995) ‘Voice over, narration et commentaire’. Translatio, Nouvelles de la FIT-FIT Newsletter XIV(3–4): 303–7. Sánchez-Escalonilla, A. (ed.) (2003) Diccionario de creación cinematográfica. Barcelona: Ariel. Shuttleworth, M. and Cowie, M. (1997) Dictionary of Translation Studies. Manchester: St. Jerome. 11 Broadcasting Interpreting: A Comparison between Japan and the UK Tomoyuki Shibahara Introduction Although broadcasting interpreters play a significant role in the news programmes today, little is known about their role and their profession (Mitsufuji, 2002). An analysis of the interpretation style of broadcasting interpreters could serve as a useful reference for many other interpreters in terms of improving their delivery. Therefore, I believe it is important to introduce and compare the work of broadcasting interpreters in Japan and the UK. Based on my personal experience, I take a look at the role of these professionals working for the BBC (British Broadcasting Corporation) and NHK (Japan Broadcasting Corporation). In this comparison of broadcasting interpreting styles in both countries, I will attempt to explain the reasons behind some of the differences. At the same time, I will try to find out if there is a ‘common interpreting policy’ to be found between the two broadcasting organisations. Definition and history of ‘broadcasting interpreter’ in Japan In some countries and languages, the word for translator also encompasses that of the interpreter. However, in Japan, there is a clear-cut distinction between interpreter and translator, despite the fact that in general conversation the word translator is used synonymously with interpreter. The main difference between these two professionals is that the translator is the person who translates written material, and the interpreter, spoken dialogue, and although my official job title at the 140 Broadcasting Interpreting in Japan and the UK 141 BBC Japanese Unit was ‘broadcasting translator’, I prefer ‘broadcasting interpreter’. Japanese people, in general, tend to believe that interpreting is somewhat more difficult than translating. This may be due to the fact that for a long time the main purpose of foreign language education in Japan has been to absorb new information and technology by reading foreign books. As a result, foreign language education in Japan tends to put more stress on reading and writing than on listening and speaking. The Japanese feel that it is much easier to be a translator because they perform tasks which are similar to what they did in school. As they feel that it is more difficult to listen and to speak in English, it is then considered very hard to be an interpreter. When interpreter is combined with the word broadcasting, for many Japanese people it takes on star status. In actual fact, the term ‘broadcasting interpreter’ refers to a professional who performs mainly interpreting for broadcasting, although the job sometimes includes translating as well. It is very common for ‘ordinary’ interpreters to work only part-time as broadcasting interpreters, because job opportunities are very limited and therefore, few people can make a living by broadcasting interpreting alone. To the best of my knowledge, there are no full-time, in-house positions in Japan for broadcasting interpreters and they are employed on a freelance basis. The only exception is the BBC Japanese Unit, although interpreters then have to work in London. Broadcasting interpreters are usually able to watch beforehand the material that they are going to interpret. They can either prepare the Japanese translation or, if there is not enough time to do so, make rough notes that they will use while on air. Live coverage such as press conferences are interpreted simultaneously. In Japan, broadcasting interpreters made their first appearance on public television in 1969, when Apollo landed on the moon. The event was covered live on TV, and the Japanese audience listened to the simultaneous interpretation. Since top-notch interpreters were assigned to do this job, the performance was very impressive. It resulted in the general perception that broadcasting interpreters are somewhat superior to ‘ordinary’ interpreters. During the 1980s, it became a trend among Japanese TV stations to make use of sound-multiplex broadcast systems to revoice foreign news in Japanese. This trend reached a climax when in 1991 the Gulf War broke out. During this period, TV stations were forced to increase the number of broadcasting interpreters. It had both positive and negative effects. On the one hand, it enabled many young and promising interpreters to acquire 142 Tomoyuki Shibahara experience in broadcasting interpreting. On the other hand, and unlike the case of the Apollo Project, many Japanese viewers are well informed about war issues and jargon. As a result, broadcasting interpreting came to be criticised as inaccurate, with many viewers stating that it was very hard to make out what the interpreters were saying. Broadcasting interpreters at the BBC Japanese Unit In the case of the BBC Japanese Unit, interpreters are employed fulltime and work in-house. To be more precise, they work as contract employees of the BBC’s contract company. The length of each contract is one year, renewed annually. It offers financial stability for the interpreters and enables them to have a fixed schedule, which means that they do not have to work very late at night or very early in the morning unless they agree to do so. The fact that interpreters work five days a week for a prolonged period of time is beneficial to them, enabling them to accumulate and enhance necessary knowledge and skills. This is a great advantage in terms of keeping the quality of broadcast high. However, this employment style can be a double-edged sword. Firstly, even if the workload increases, there is no change in salary. For example, when NATO bombed Kosovo in 1999, the interpreters covered the daily NATO press conferences live every day for more than two months. The quantity of simultaneous interpretation at least quadrupled during that period. But since press conferences took place within regular working hours, there was no extra remuneration. In addition, the pay scale is based on seniority, and there was no merit pay. The work of the BBC Japanese Unit consists of providing Japanese translation for BBC World TV, which transmits mainly news and documentary programmes. The broadcast in Japan started in 1994, and it was introduced as a successor to BBC World Service radio’s Japanese language service, which ceased its operations in 1991. There are three major tasks that are assigned to the broadcasting interpreters: 1. Interpreting the news programmes. 2. Translating documentary programmes and checking the translation. 3. Voicing-over of documentary programmes. I would like to focus on the first task since it is the one most similar to the activity carried out by broadcasting interpreters working for NHK, the biggest broadcasting organisation in Japan. Broadcasting Interpreting in Japan and the UK 143 News programmes consist of two parts, the lead and the piece. The lead is the introduction which is then followed by the piece or the correspondent’s report. The piece is also called ‘TX’. Transcripts for the lead are usually filed in the computer system beforehand. It is the responsibility of the interpreter who is in charge of interpreting the anchor on that day to prepare the Japanese translation. Since the interpreter reads aloud the translation at the very same time as the programme goes on air, the translation has to be as long as the original and fit within the same time frame as the English source text. If the interpretation is too long it will spill over into the TX, and will obstruct the interpretation of this section. This can have a very negative effect since the lead and the piece are interpreted by different interpreters. There are almost no transcripts available for interpreters who are in charge of the piece but correspondents’ reports are usually filed in the computer system in video format; interpreters can download the footage in order to prepare the Japanese translation. As in the case of the lead, when read aloud the translation should fit within the time frame of the original. The translation should also occur in concert with what is happening on the TV screen so that there is no conflict between the interpretation and the picture. If there is not enough time to prepare the Japanese translation, interpretation takes place simultaneously while on air. Simultaneous interpretation is also called for in the case of live coverage, when for example the anchor interviews a guest in the studio, or the correspondent files a live report. Interpreting policy and quality control The quality of the Japanese is greatly emphasised by the editor, stressing that the standards should be comparable to NHK news. Under the slogan of ‘less is more’, interpreters are asked to edit the information so that the interpretation is not only accurate but also easy for the viewers to understand. Delivery is considered important and interpreters are offered elocution training in Japanese. Interpreters are asked to comply with the ‘less is more’ policy even when they are interpreting simultaneously; a rapid and somewhat garrulous delivery, which is rather common in simultaneous interpretation, is not welcomed. On BBC World TV, the interpreter’s name does not appear on the screen, making it difficult for the viewer to identify the interpreter when there is any misinterpretation. Although editors are in charge of checking the output of the interpreters, they tend to focus on purely technical matters such as voice level, paper rustle, etc. rather than on 144 Tomoyuki Shibahara the linguistic merits of the translation. Besides, given the time pressure, it is almost impossible for the editors to check the entire output. As for checking word usage, this is basically left to the interpreters to cross-check among themselves. Under this system, the translation of new words tends to cause debate among the interpreters. In order to solve this problem, there have been attempts at forming a Translation Standardisation Committee in charge of making the use of certain terms consistent. However, it is very difficult to create a consensus among the interpreters and, so far, the attempts have not resulted in any committee being formed. Nonetheless, a Standard Translation Database System is now in place, where interpreters take turns to define the translation of new words and add them to the database. In any case, editors have the final say in deciding which Japanese term should be used. Broadcasting interpreters at NHK Broadcasting interpreters for NHK are provided by NHK Joho Network Inc., an affiliate company of NHK. The Bilingual Center, one of the divisions of NHK Joho Network Inc., is in charge of selecting and arranging interpreters and translators. More than one thousand interpreters and translators in some 60 languages are registered at the Bilingual Center. Since NHK interpreters work on a freelance basis, they have considerable flexibility. They can decide if they do not want to work very late or very early hours. When there is a high volume of work, interpreters are called in more often, which means an increase in their income, as for example during the reporting on the War in Iraq in spring 2003. However, working on a freelance basis means that there is no income security, and it can be very difficult for interpreters to rely entirely on the income from NHK to make a living. Besides, it is rather difficult for young interpreters to work for NHK in the first place because it is a very competitive environment and most of the positions have already been taken by experienced interpreters. Had it not been for the emergency situation presented by the war in Iraq, many interpreters might not have had a chance to work for the NHK. NHK broadcasting interpreters work for two channels: BS-1, a satellite news and sports channel; as well as General TV, a terrestrial channel. In both cases, broadcasting interpreters work for the news programmes only. NHK has many overseas documentary programmes but these are translated separately by broadcasting translators. Voice-over is also done separately by voice actors and actresses. Although BS-1 has a larger and more constant demand for interpreters, because it broadcasts many Broadcasting Interpreting in Japan and the UK 145 foreign news programmes, I would like to limit my explanation below to the case of General TV. The tasks of interpreters working for General TV are: (1) Simultaneous interpretation: this is applied when something urgent happens and there is no time to prepare a translation beforehand. In the case of the War in Iraq, there were three 8-hour shifts a day (00:00–08:00; 08:00–16:00; and 16:00–24:00), and two interpreters were always standing by. When the initial air strike took place, NHK ran the live coverage of ABC News, which was interpreted simultaneously. Interpreters were also asked to monitor other news channels such as BBC, CNN, and Fox. If editors decided to run any of these channels, interpreters were asked to interpret simultaneously. In addition to simultaneous interpreting, interpreters are asked to provide whispering interpreting (chuchotage) for NHK editors. For this task, interpreters covered mainly the news conferences of the White House or the US Central Command. Editors listened to the chuchotage in order to decide which part of the news conference should be cut out and used – usually with subtitles – in the news bulletin following. In this kind of work interpreters are then asked to pin-point the footage with the video technician. Having found where the footage begins and ends, interpreters are asked to interpret the contents. In most cases, the interpretation is used to make subtitles in Japanese and, when there is a shortage of manpower, interpreters are also asked to help with the subtitling. (2) Monitoring of foreign news channels in case editors decide to use short footages from these. (3) Reading the Japanese translation of short footages: sometimes interpreters are asked to voice-over the Japanese translation in the actual telecast. Professional voice actors and actresses are rarely used for this task unless it is something very important and there is enough time to do so. (4) Providing translation to make subtitles for short footages. The last two tasks are assigned to interpreters who have not reached the level of simultaneous interpreting or who have opted out of simultaneous interpreting. They follow similar working shifts as simultaneous interpreters and two to three interpreters are assigned to each shift. The difference in the functions between simultaneous interpreters and other interpreters is clearer in NHK than in the BBC Japanese Unit. Despite helping with the above two tasks, simultaneous interpreters 146 Tomoyuki Shibahara and other interpreters basically work separately, dealing with different types of tasks. Interpreting policy and quality control The person responsible for the selection of interpreters considers accurate interpretation very important together with a high degree of fluency in simultaneous interpretation. As in the case of the BBC Japanese Unit, interpreters are expected to edit and to organise the information. It is also important to avoid words which are not allowed to be used on air, such as terms which may be considered discriminatory. Both on BS-1 and General TV, the interpreter’s name appears on screen. The programme is shown in the NHK news room where editors can watch it and, in the event there is a problem with the interpreting, they can quickly point them out to interpreters. As for the standardisation of certain translation terms, there is at least one set of guidelines for the General TV team. BS-1 channel provides a bulletin board that interpreters can refer to for the translation of new words and on some occasions they hold study sessions, as happened before the War in Iraq broke out. Comparison between BBC Japanese Unit and NHK (General) TV Interpreters are employed in-house, full-time at the BBC Japanese Unit, while interpreters employed by NHK work on a freelance basis. This may be because interpreters at BBC World TV have to cover an entire news bulletin, whereas interpreters at NHK have to interpret only parts of the news bulletin. Since job security is not guaranteed at NHK, interpreting fees are higher than those paid by the BBC Japanese Unit. Working five days a week, interpreters in NHK can earn roughly twice as much as interpreters in the BBC Japanese Unit working full-time. The work of the BBC Japanese Unit requires broadcasting interpreters to be all-rounders, while their counterparts at NHK are required to specialise in a certain area. This may be because of the different nature of the contents of the programmes broadcast by the channels for which they work. BBC World TV tends to follow a fixed pattern with news and documentary programmes broadcast every thirty minutes. Interpreters have to cover both types of programmes, although documentaries are translated and voiced-over beforehand. On BS-1 and NHK General TV news programmes are not always run every hour, and they can be Broadcasting Interpreting in Japan and the UK 147 followed by sports or variety programmes, which do not necessarily require broadcasting interpreters. Difference in the interpretation policy and quality control The BBC Japanese Unit likes to stress the ‘broadcasting’ dimension of this profession. In other words, it encourages interpreters to edit the information in order to make the target text clearer to the viewers. Although this is also done by interpreters at NHK, more emphasis seems to be put on the accuracy of the interpreting. This may be due to the fact that NHK is the biggest authority among TV stations in Japan and considers it one of its duties to set high standards. However, this does not mean that BBC World TV sacrifices accuracy, as it has to compete with other TV channels such as CNN and NHK, and their interpreting has to be more than ‘just accurate’. The greatest difference in the quality control is whether interpreters are under constant monitoring. It can be said that NHK has a very strict system in this respect, while the BBC Japanese Unit has a more democratic, if not flexible, system. Some interpreters at the BBC Japanese Unit felt that the name of the interpreter should appear on the screen; the reason for this being that it could provide an incentive for interpreters to take more responsibility for their work and, consequently, boost the overall standard of interpreting. Having personally worked in both organisations, I have found that adding the name on the screen does indeed have a great effect on the interpreters. However, this proposal was rejected at the BBC Japanese Unit because of financial considerations. Conclusion Interpreting is a way of communicating and the effectiveness of the communication may be measured by judging how well the message is conveyed. No matter how faithful the interpretation is to the original, it means very little to the viewers if the final output is very hard to understand. In the case of broadcasting interpreting, it may be assumed that viewers are not knowledgeable about all the subjects discussed. Therefore, it is not sufficient to ‘convert’ the Source Language into the Target Language and hope that the viewer will understand the message behind the facts. Interpreters have to understand the message in the Source Language and, if necessary, edit the information to make it easier for the 148 Tomoyuki Shibahara viewers to understand. It may be said that any type of interpreting implies editing information in one way or another. However, in broadcasting interpreting it is essential that interpreters actively edit the information to ensure that communication is as effective as possible. This editing leads, of course, to yet another question as to how free or faithful an interpreting act should be. References BS Hōsō Ts̄yaku Gur̄pu (1998) Hō sō ts̄yaku no sekai (The World of Broadcast Interpretation). Tokyo: ALC. Mitsufuji, K. (2002) ‘Interpreting as audio-visual translation (AVT)’. Interpretation Studies 2: 87–98. Part III Accessibility to the Media This page intentionally left blank 12 Interlingual Subtitling for the Deaf and Hard-of-Hearing Josélia Neves Introduction Considerable mileage has been covered since the early days, in the late 1950s, when audiovisual translation (AVT) began to be addressed as a subject in its own right. Concerns have moved from the dubbing/ subtitling debate to more specific domains such as the analysis of issues pertaining to discourse analysis, technical constraints and audience design. Comprehensive lists of audiovisual translation typologies (Luyken et al., 1991; Gambier, 1996; Díaz Cintas, 2003) have reflected the gradual change in scope and are making space for newer forms of language transfer within the audiovisual context. Indeed, the lines once drawn between language transfer types in the media are becoming less and less visible. Principles and techniques are merging to give way to specific offers, directed to specific audiences. This implies that the very concept of ‘mass’ media is changing; technology is now allowing masses to be broken down into smaller groups and products are tailor-made to the expectations and the needs of defined sub-groups. AVT will inevitably need to follow the general trend in the audiovisual market and, rather than aiming to cater for a general audience, audiovisual translation now finds itself focusing on the needs of smaller distinct audiences in order to respond to them in a more adequate manner. This concern with making translations accessible to the intended receptor is not new. As a matter of fact, in his seminal work ‘Principles of correspondence’, Nida (1964) calls our attention to the fact that translations need to be directed towards the different types of audiences. This scholar takes up the topic later (1991: 20) to specify that translators need to look at ‘the circumstances in which translations are to be used’. Ever since, academics and translators have gone to great lengths to find 151 152 Josélia Neves theories and practices that reflect such concerns for their audiences. Kovačič (1995: 376) questions the meaning of ‘reception’ in subtitling, to see it as a multilayered construct in which factors such as ‘the sociocultural issue of non-TV context influencing the process of receiving subtitles’, ‘the attitudinal issue of viewers’ preference for subtitling over dubbing or vice versa’, ‘the perceptual issue of subtitle decoding (reading and viewing) strategies’, and ‘the psychological or cognitive issue of the impact of cognitive environment on understanding subtitles’ are of paramount importance in audience design. Gambier (2003: 185) adds that ‘these four aspects (socio-cultural, attitudinal, perceptual and psychological/cognitive) could be used to inform a model for research on subtitle reception’. These very same aspects have proved to be of utmost importance in the still ongoing study of SDH in Portugal; and it is with these in mind that I address a situation that is changing just as writing and reading are changing. In fact, technological and policy changes are offering new challenges to all those working within AVT and it would be interesting to see practices improving, by taking advantage of the new products for the benefit of audiences who should not be seen as minorities but as one of the many parts of a fragmented reality. It is in this context of fragmentation and plurality that a different type of subtitling solution, interlingual SDH, has emerged. A new concept to most people, interlingual SDH is gradually gaining visibility, particularly in the DVD market where, further to intralingual subtitles for the hearing-impaired, a track is provided with interlingual subtitles for the hearing-impaired in a language other than that spoken in the original soundtrack. The analysis of 200 randomly chosen DVDs, made available in video rental shops in Portugal, showed that close to 38% contained intralingual SDH in English; 9% contained a track with interlingual SDH (15 titles had interlingual SDH in German; and 3 in Italian); and none of such titles contained SDH into Portuguese. Indeed, these numbers show that not many DVDs offer interlingual SDH, and those that do offer it, seem to do so by applying the conventions of traditional subtitling on DVD, which, I believe, do not cater for the needs of people with hearing impairment. However unusual and scarce the offer, this ‘new product’, in its hybrid make-up, lends itself to reflection and allows one to address problems that are shared with the more common types of subtitling: both interlingual subtitling for hearers and intralingual SDH. In its complexity, interlingual SDH might be regarded as a perfect example of transadaptation (Gambier, 2003). On the one hand, as happens with traditional interlingual subtitling, one is faced with an Interlingual Subtitling for Deaf and Hard-of-Hearing 153 instance of translation between two different languages. On the other hand, as happens with intralingual subtitling, further to taking speech into writing, there is a need for the adaptation of multiple aural messages (speech, sound effects and music) so as to produce a visual (most often verbal) substitute for the information that cannot be picked up by people with hearing impairment. In my view, what most strongly determines the nature of this kind of subtitling is the need to ‘adapt’ a product to a target audience or addressee, which, according to Nord (2000: 196), is: ... not a real person but a concept, an abstraction gained from the sum total of our communicative experience, that is, from the vast number of characteristics of receivers we have observed in previous communicative occurrences that bear source analogy with the one we are confronted with in a particular situation. In other words, if we are to consider subtitling as ‘translational action’ (Vermeer, 1989: 221), serving a functional end, its skopos needs to be perfectly understood by all those involved in the commission. Quite often, the commissioners of SDH, and the subtitlers themselves, are not fully aware of the particular needs of their ‘clients’ for not much is given to them in terms of audience design or reception analysis. In fact, only by knowing the distinctive features of the target audience will people be reasonably aware of the possible effects of their work on the receptor. Only then can anyone aim at the utopian situation where the ‘new viewer’s experience of the programme will differ as little as possible from that of the original audience’ (Luyken et al., 1991: 29). No distinction appears to be made between being deaf and being hardof-hearing in terms of intralingual subtitling. Both conditions go hand-in-hand as if they were complementary or even interchangeable. Guidelines for intralingual subtitling assume that their subtitling solutions cater for the needs of all alike and, in so doing, I would suggest that they are catering for the needs of neither. According to Nord (2000: 195): ... the idea of the addressee the author has in mind, is a very important (if not the most important) criterion guiding the writer’s stylistic and linguistic decisions. If a text is to be functional for a certain person or group of persons, it has to be tailored to their needs and expectations. An ‘elastic’ text intended to fit all receivers and all sorts of purposes is bound to be equally unfit for any of them, and a specific purpose is best achieved by a text specifically designed for this occasion. 154 Josélia Neves If subtitlers are to produce functionally adequate texts for these particular people, it seems fundamental that, however difficult it may be, one clarifies what is meant by ‘deaf’ and by ‘hard-of-hearing’, to which I add another term: ‘Deaf’. Deaf vs. hard-of-hearing (HoH) Medically and clinically speaking, one is considered to be ‘deaf’ whenever hearing loss is so severe that one is unable to process linguistic information through hearing alone.1 On the other hand, one is considered to be ‘hard-of-hearing’ when there is a hearing loss, whether permanent or fluctuating, which adversely affects an individual’s ability to detect and decipher some sounds. In other words, despite the hearing loss, there is still some residual hearing. Rodda and Grove (1987: 43) draw the line between audiological and social understanding of these terms: Defining hearing loss is a fairly simple matter of audiological assessment, although the interpretation of the simple pure-tone audiogram is more difficult. Defining deafness is exceedingly complex; it is as much, if not more, a sociological phenomenon as an audiological definition. This particular reference to the sociological dimension of deafness is of utmost importance when it comes to knowing our addressees better. In fact, it leads us to a different concept, the concept of ‘Deaf’ which refers to the cultural heritage and community of deaf individuals, that is the Deaf culture or community. In this context, the term applies to those whose primary receptive channel of communication is visual. It is also in this context that we refer to the Deaf as a minority group, with a language of their own (national sign language) and with an identity that differs from that of hearers and that of the hard-of-hearing. Further to this, we are dealing with a group of people whose first language is not that of hearers in their country. The (oral) national language is, even for those Deaf people who are Bi-Bi (bilingual and bicultural), a second language with all that this entails. With all this in mind, I will try to reformulate the traditionally canonical ‘for the deaf and hard-of-hearing’ by readdressing hearing and deafness in terms of the two ends of a scale that spans between two distinct poles: being a ‘hearer’ and being ‘Deaf’. Here too, I prefer to address deafness in sociological rather than audiological terms. To be ‘hard-of-hearing’ means to be somewhere in the ‘grey zone’, midway Interlingual Subtitling for Deaf and Hard-of-Hearing 155 between two distinct realities. Usually, people who are hard-of-hearing identify themselves with the hearing community. They have acquired the condition through age or disease but they mainly partake of the social order of the community in which they were raised. Their mother tongue is the spoken language of their national group. They have a notion of the sound systems which inhabit their environment. Even in cases of severe hearing loss, there is usually some residual hearing and, above all, sound is still stored in a fairly accessible memory bank, ready to be retrieved through effective stimuli. Being Deaf, on the other hand, means belonging to a world in which there is no sound, not necessarily meaning that there is silence. Emmanuelle Laborit, a French Deaf actress and author of an autobiography, The Cry of the Gull (1998),2 describes silence as ‘the absence of communication’, adding that ‘I’ve never lived in complete silence: I have my own noises that are inexplicable to hearing people. I have my imagination and it has its noises that are inexplicable to hearing people’ (ibid.: 10). Further to this, Deaf people communicate among themselves using their natural language: sign language. Laborit (ibid.: 88) clarifies: That’s right. As far as I am concerned, sign language is my voice and my eyes are my ears. Frankly, I don’t feel deprived of anything. It’s society that makes me handicapped, that makes me dependent of hearing people, that makes it impossible to contact a doctor directly, that makes me need to have conversations translated, that makes me have to ask for help to make a phone call or for captioning on TV (there are so few captioned programs). With more Minitels and captioning, I or rather we, the deaf, could have better access to culture. There wouldn’t be a handicap, a deadlock, a border between us. Up until fairly recently, Deaf children were brought up to ‘oralise’. This meant that they were taught to pronounce words and to make use of lip reading to understand speech. Communicating through sign language was not widely accepted and Deaf people were forced to use the national oral language, regardless of the fact that their hearing and speech apparatuses were not tuned to such a task. The ways in which Deaf children are educated determine their development and their perception of the world. Their proficiency in the use of language will be of paramount importance in their ability to decode messages. Needless to say, Deaf people’s reading competence will necessarily be different from that of hearers, and that of the hard-of-hearing as well. And subtitles are all about reading. 156 Josélia Neves Reading film ... in search of relevance Watching a film is all about decoding information that is conveyed through multiple channels (speech, sound, and image). Reading film is a complex process. It is problematic to say that full access to audiovisual texts is ever attained, even in the case of people with no impairment. In accepting the dialectic interaction between the producer and the receiver in the construction of meaning, and in the knowing that the audiovisual text is a perceptive whole that does not equal the sum of its parts, it is as obvious that decoding polysemiotic texts is a demanding task for all. Speech is usually decoded through cognitive processing, and sound effects and visual signs are often impressionistic and concurrent. Image, sound and speech are interrelated, quite often in a redundant manner. Such redundancy makes access to the message easier and it is only when there is no direct access to one of the elements, (either for some sensory impairment or for lack of knowledge in the Source Language) that subtitles gain importance. Interlingual subtitles usually bridge the gap between the SL and the language of the target audience; intralingual subtitles replace more than speech, they are expected to account for paralinguistic information and for acoustic elements. If they are to be fully integrated with the visual and auditory channels, subtitles must also guarantee a certain degree of redundancy in relation to concurrent information, fitting in with the whole in an integrated manner, guaranteeing ease rather than adding to the load of decoding effort. Díaz Cintas (2003: 195) sheds light on the issue by reminding us that: Even for those with adequate command of the foreign language, every audiovisual product brings with it a range of additional obstacles to comprehension: dialectal and sociolectal variation, lack of access to explanatory feedback, external and environmental sound level, overlapping speech, etc., making translation of the product crucial for the majority of users. This is so much more pertinent in the case of people with hearing impairment. To a greater degree than in the case of people who do not understand the SL, to the Deaf and HoH subtitles are essential, rather than redundant. They are the visual face of sound. For the HoH they are a stimulus and a memory exercise; for the Deaf, they are the only means to gain access to aural information. However redundant, sound Interlingual Subtitling for Deaf and Hard-of-Hearing 157 and image tell different stories. Watching images alone will not allow us to grip the whole, just as listening without viewing will never allow for a full understanding of the whole audiovisual construct. When subtitling for these specific audiences – the Deaf and the HoH –, it is up to the subtitler to turn into written words both the dialogue that is heard and the sound effects that are only apprehended. In other words, subtitlers need to be familiar with filmic codes so that they may capture the different effects in order to transfer sensations and emotions into words. Deciding what is relevant is a difficult task, for relevance is determined by the needs of the addressee. Ideally, as mentioned before, subtitlers should aim at producing equivalent effects on their audience as those produced in the target audience of the original. But, as Gutt (1991: 384) puts it: ... this raises the question of what aspects of the original the receptors would find relevant. Using his knowledge of the audience, the translator has to make assumptions about its cognitive environment and about the potential relevance that any aspects of the interpretation would have in that cognitive environment. Knowing the addressee’s cognitive environment is relatively easy for subtitlers who translate into their mother tongue and for an audience who is believed to belong to their social group. But when it comes to subtitling for the Deaf, hearing subtitlers rarely have true knowledge of the cognitive environment of their target audience. This may be due to the lack of specific training in the area, or even due to the fact that subtitlers are not aware that their ‘translation action’ is specifically directed towards addressees that do not share the subtitler’s language or culture. According to Gutt (ibid.: 386) and his approach to translation and relevance: ... whatever decision the translator reaches is based on his intuitions or beliefs about what is relevant to his audience. The translator does not have direct access to the cognitive environment of his audience, he does not actually know what it is like – all he can have is some assumptions or beliefs about it. And, of course, as we have just seen, these assumptions may be wrong. Thus our account of translation does not predict that the principle of relevance makes all translation efforts successful any more than it predicts that ostensive communication in general is successful. In fact, it predicts that failure of communication is likely to arise where the translator’s assumptions about 158 Josélia Neves the cognitive environment of the receptor language audience are inaccurate. This alone may account for many of the problems often found with subtitles for the deaf and HoH, which I think should be rightfully named ‘subtitling for the Deaf and HoH’. What is relevant? How do these two addressees (for they are indeed two different communities) perceive the world? How do they read film? And how do they read words (subtitles)? How much information do they need for meaning to be fully understood? When are words too much/too many? When are they not enough? Only when subtitlers find answers to these questions will they be better equipped to determine their addressees’ profile. Only when that is done will they be able to decide on issues such as how much and which information is to be presented in the text; how this information is to be structured; and which linguistic and stylistic devices are to be used to present the selected information, so that the translation may yield ‘the intended interpretation without putting the audience to unnecessary processing effort’ (Gutt, 1991: 377). Further, and keeping the audiences’ needs in mind, I would posit that subtitlers should question what supplement begs to be added for their benefit. In fact, when asked to comment upon the open subtitles offered on Portuguese television programmes, the criticism that was most often made by Deaf respondents was that subtitles were ‘difficult’. Deeper probing led to the conclusion that the respondents were referring to the great effort they put into decoding (any) written text and particularly into the reading of subtitles. Not many people referred to the need for supplementary information on sound effects. Most of the respondents had little notion that such information could be added at all, as at the time of the survey in early 2003, there was no significant offer of intralingual subtitles for the Deaf in Portugal. They demanded nothing, because they expected nothing, they did not know they could get anything different. But they voiced their frustration about not following subtitles: ‘subtitles come and go too rapidly’; ‘the vocabulary is too difficult’; ‘the sentences don’t make sense; they are too long and convoluted’, were some of their comments. These reactions to common, open subtitles call for further reflection on the different needs of the Deaf audience. One might question what could be done to attain the minimax effect (Kovačič, 1994: 246) to improve reception or to guarantee the viewer’s comfort. The answer is not always straightforward. But further discussion on how the Deaf Interlingual Subtitling for Deaf and Hard-of-Hearing 159 read might shed some light on the issue. Laborit (1998: 120) speaks of her reading experience in the first person: I was one of the rare students in the Morvan program who read a lot. As a general rule, deaf people don’t read very much. It’s hard for them. They mix up the principles of oral and written expression. They consider written French a language for hearing people. In my opinion, though, reading is more or less image-based. It’s visual. If we can extrapolate, what seems important here is the fact that Deaf people in general do not enjoy reading very much, and whereas the Deaf have this ‘visual’ reference base for the reading process, hearers complement it with an auditory reference system. Further to this, many have not developed the tools that allow them to take a step forward from simple word-processing to processing of a higher order such as inferencing and predicting, (often done subconsciously by the hearer), and planning, monitoring, self-questioning and summarising (metacognitive techniques that are specific to highly skilled readers). Normative studies on the reading abilities of Deaf people (Di Francesca, 1972; Conrad, 1979; Savage et al., 1981; Quigley and Paul, 1984) substantiate this idea that Deaf people attain very poor standards in reading. The results of the well known study conducted by the Office of Demographic Studies at Gallaudet College, USA, in 1971, indicated that ‘although the reading skills of deaf students increase steadily from 6–20 years, they peak at a reading level equivalent to Grade 4 in the United States school system (approximate chronological age 9 years)’ (Rodda and Grove, 1987: 165). It is widely accepted that subtitles are constrained by time and space to the point of making language hostage to parameters that will dictate options when writing subtitles. Such parameters vary in nature. Some pertain to the programme itself (genre, global rhythm, type of action); others to the nature of subtitles (spatial/temporal features; position of subtitles on the screen; pauses between subtitles; gap between dialogue and subtitles; density of information); and others to textual and paratextual features (semantic/syntactic coherence; register; role of punctuation). Further to these constraints many more will derive from factors such as the medium (cinema, VHS/DVD, TV); country of transmission (AVT policies and practices); producers’, distributors’ and clients’ demands; or, in the case of television broadcasting, the time of transmission and the expected audience. The problems that hearers find in reading subtitles will obviously be all the more acute for the Deaf viewer. Luyken et al. (1991) suggest that 160 Josélia Neves average adult reading speeds hover around 150 to 180 words per minute. This number varies according to the manner in which the text is presented, to the quantity and complexity of the information, and to the action on the screen at any given moment (De Linde, 1995: 10). The 6-second rule has been widely accepted as rule of thumb for ‘readable’ subtitles. On being questioned about the adequacy of such a ‘norm’ for the specific needs of the Deaf, and with reference to studies on how the Deaf read subtitles, d’Ydewalle states that ‘the 6-second rule should be replaced by a 9-second rule as they are typically slow readers’ (personal email, 17 February 2003). Reducing the amount of information is not, in my opinion, the solution to this need for extra processing time. Reduction is often achieved through the omission of information or by sacrificing interpersonal meaning. Redundancy is a feature of all natural languages and serves to make messages better understood. Such redundancy – phonetic, lexical, collocational or grammatical in nature – serves mainly to make up for possible interference, or noise. In subtitles, losing redundancy for the sake of economy is common practice, often resulting in a greater processing effort on behalf of the reader/viewer. In fact, subtitling goes against the grain of most translation practices in that rather than extending, subtitles usually seek to condense as much information as possible in as little space as possible. This often results in reduction strategies that can make decoding rather taxing, particularly for those who are reading subtitles in their second language. In the case of the Deaf reader/viewer, redundancy is of utmost importance, for such elements will make reading less demanding. This obviously adds tension to the difficult equilibrium between economy and expenditure. Economy cannot be had if it is at the expense of meaning. The first priority, when subtitling for this particular audience, needs to be the conveyance of meaning as fully as possible in as ‘readable’ a manner as possible, and only then, in as condensed a form as possible. Quite often, extra reading time might need to be given to allow for the reading of longer subtitles; however, if drawn out subtitles mean the use of simpler structures or better known vocabulary, it may well be worth sacrificing synchronisation and having subtitles appearing a little earlier or staying on a little longer, thus adding to reader comfort. Cohesive devices are also easily sacrificed in the search for economy. When cohesion is given by the image, subtitles may jeopardise cohesive devices; the problem arises when cohesion needs to be reinforced through language for lack of visual redundancy. What is conveyed Interlingual Subtitling for Deaf and Hard-of-Hearing 161 through subtitles is never the message but a possible version of the message and special care needs to be taken so that nothing can contradict that particular version. It is easy for the image to add, confirm or even contradict the verbal message conveyed. For Deaf audiences, who depend on visual cues to assist language decoding, special care needs to be taken so that subtitles may be truly complementary and so be seen as part of the whole, never as an obtrusive, or even cumbersome, add-on. Language has its own means of guaranteeing internal and external redundancy. Speech naturally involves linguistic, paralinguistic and non-linguistic signs. Paralinguistic signs cannot be interpreted except in relation to the language they are accompanying. On the other hand, non-linguistic signs are interpretable and can be produced without the co-existence of language. Non-linguistic signs or natural signs such as facial expressions, postural and proxemic signs, gestures, and even some linguistic features ‘are likely to be the most cross-culturally interpretable’ (Cruse, 2000: 8) but also the source of many misunderstandings in cross-cultural communication. Such kinetic elements are no greater a problem to the Deaf than they are to hearers. However, paralinguistic signs are more often hidden from the Deaf person, for they are only sensed in the tone or colour of voice in each speech act. There are times when such paralinguistic signs actually alter the meaning of words; and more often than not, punctuation cannot translate the full reach of such signs. Whenever such signs have informative value, there is a need for explicitation, one of the ‘universals of translation’ according to authors like Toury (1980: 60). In subtitling for the Deaf, explicitation is a fundamental process to compensate for the aural elements that go missing. In the case of paralinguistic information, there might even be a need to spell out what can only be perceived in the way words are spoken. Sound effects (such as ‘phone rings’) are often straightforward to describe, but expressing what comes with the tone of voice (irony, sadness, irritation) can be difficult. In feature films and series, paralinguistic features are most frequently found in moments when the story is being pushed forward by emotional interplay or when characters reveal something in themselves that goes against their words. This could mean that adding extra information might alter the intended pace or cut down on the tension. Finding adequate solutions for the problem is a challenge for those working in the area. The use of smileys to convey such information has proved effective thanks to the economy of their pictorial nature. Deaf TV viewers in Portugal have reacted most 162 Josélia Neves favourably to the introduction of commonly used smileys in the subtitling of a primetime soap opera. 3 Even though the introduction of information about paralinguistic signs may be considered redundant for hearers, it is fundamental for the Deaf if they are to get a better perception of the expressiveness of the intersemiotic whole. A study conducted by Gielen and d’Ydewalle (1992: 257) concludes that ‘redundancy of information facilitates the processing of subtitles’. If Deaf viewers are to gain better access to audiovisual texts, such redundant components of speech will need to be made explicit, and then too, be redundant in terms of the subtitles that convey speech utterances. Such redundancy will be a fundamental element to boost the limited short-term working memory characteristic of Deaf individuals (Quigley and Paul, 1984). Short-term working memory is of crucial importance in the reading process for it allows the reader to understand complex sentences and linked ideas. Coherence and cohesion is only possible if the reader can keep different fractions of information active so that, once processed together, meaning might be achieved. This does not mean that Deaf people are cognitively impaired and unable to process information efficiently; what this means is that they make use of other strategies (strong visual memory) to process information. As Rodda and Grove (1987: 223) put it: Hearing impairment does not incapacitate their central comprehension processes. Provided deaf readers can grasp the semantic context of a message, they seem to be able to exploit the syntactical redundancy of natural language and to comprehend its contents with surprising degree of efficiency. Deaf viewers will benefit from subtitles that are syntactically and semantically structured in ways that will facilitate reading. Long complex sentences will obviously be more demanding on their short-term working memory. Short direct structures, with phrasal breaks (e.g. avoiding the separation onto different lines of the article from the noun), will ease comprehension and make the reading of subtitles far more profitable. This does not mean that Deaf people cannot cope with complex vocabulary. In fact, subtitles may be seen as a useful means to improve the reading skills of Deaf viewers, as well as an opportunity to increment both their active and passive vocabulary. However, difficult vocabulary should only be used to some specific purpose, and only when there is enough available time for the processing of meaning. In fact, this principle could apply to all sorts Interlingual Subtitling for Deaf and Hard-of-Hearing 163 of subtitling for, as Gutt (1991: 380) reminds us when referring to translation in general: ... rare lexical forms [ ... ] are stored in less accessible places in memory. Hence such unusual forms require more processing effort, and given that the communicator would have had available a perfectly ordinary alternative, [ ... ] the audience will rightly expect contextual effects, special pay-off, from the use of this more costly form. In the case of intralingual subtitling, where verbatim transcription of speech is frequently sought, pruning text is particularly difficult. In interlingual subtitling, functional shifts are less exposed and, therefore, it may be easier to adapt text to the needs of the Deaf addressee. In the case of intralingual subtitles, editing might be understood as not giving the Deaf all that is given to hearers. Writing every single word on the screen, transcribing every utterance, is not, in my view, serving the needs of this particular audience. Not having enough time to read subtitles and to process information; not understanding the meaning of certain words; or not being able to follow the flow of speech, cannot be understood as being given equal opportunities. Paraphrasing, deleting superfluous information, introducing explanatory notes, making explicit the implicit, may mean achieving functional relevance for the benefit of the target audience. Borrowing Reiss’ terminology (1971: 161), in SDH we need to make ‘intentional changes’ for our readers are definitely not those intended for the original text. In order to gain accessibility for those who cannot hear, we need to strive for ‘adequacy of the TL reverbalisation in accordance with the “foreign function” ’ (ibid.) that is being aimed at. Gutt (1991: 377) also sheds light on this issue: Thus if we ask in what respects the intended interpretation of the translation should resemble the original, the answer is: in respects that make it adequately relevant to the audience – that is, that offer adequate contextual effects; if we ask how the translation should be expressed, the answer is: it should be expressed in such a manner that it yields the intended interpretation without putting the audience to unnecessary processing effort. Making sound visible As mentioned before, audiovisual texts are multimodal in their making and sound plays an important role in their narrative force. Most of the 164 Josélia Neves time, hearers make very little effort to process sound in a film. Unlike speech, that requires cognitive effort to be decoded, sound effects and instrumental music convey meanings in a discreet way. According to Monaco (1981: 179), ‘we ‘read’ images by directing our attention; we do not read sound, at least not in the same conscious way. Sound is not only omnipresent but omnidirectional. Because it is so pervasive, we tend to discount it’. Furthermore, in contrast to speech and to paralinguistic features pertaining to oral expression, sound and music need no translation. A hearer will quite easily pick up the suggested meanings transmitted through sound effects that often underline images, sustaining them and guaranteeing continuity and connectivity. Processing music is somewhat more difficult. However, hearing viewers have grown to understand filmic conventions and have come to associate musical types with certain genres and with particular filmic effects. Quite often, film music has been taken from its intersemiotic context to grow in the ear of radio listeners, and to gain a life of its own. Yet, while associated to image, its meaning is strongly felt even if its existence is subtle and little more than a suggestion. In fact, ‘it makes no difference whether we are dealing with speech, music, or environmental sound: all three are at times variously parallel or counterpuntal, actual or commentative, synchronous or asynchronous’ (Monaco, 1981: 182) and it is in this interplay that filmic meaning grows beyond the images and the whole becomes artistically expressive. One might question the pertinence of describing music and sound effects to people who have never been able to hear. There is no question that to the HoH, a comment on sound effects or music may activate the aural memory and may even guide viewers to consciously direct their residual hearing capacities towards relevant content. Those who have lost their hearing later on in life will recall previous experiences and may find cultural, social and historical references in certain melodies or rhythms. But even those who were born deaf can, never having heard, make those very same connections creating their own references, for even though they cannot hear, they live in a world of sound, which they perceive through vibrations and they learn to appreciate through other sources (literature, television, etc.). Once again, a Deaf person’s testimony may clarify doubts (Laborit, 1998: 17): I was lucky to have music when I was a child. Some parents of deaf children think it’s pointless, so they deprive their children of music. [ ... ] I love it. I feel its vibrations. [ ... ] I feel with my feet, or my whole Interlingual Subtitling for Deaf and Hard-of-Hearing 165 body if I’m stretched out on the floor. And I can imagine the sounds. I’ve always imagined them. I perceive music through my body, with my bare feet on the floor, latching onto the vibrations. The piano, electric guitar, African drums, percussion instruments, all have colors and I sway along with them. That’s how I see it, in color. [ ... ] Music is a rainbow of vibrant colors. It’s language beyond words. It’s universal. The most beautiful form of art that exists. It’s capable of making the human body physically vibrate. I firmly believe that every effort should be made to convey sound and music visually. Present subtitling systems do not allow for much more than alphabetical characters in the form of subtitles. Many teletext systems do not allow for the use of symbols, such as , and a # is often used to indicate the presence of music. New strategies are already found in the subtitling of DVDs, where special care is given to identifying music through the use of symbols; yet the same cannot be said for television for, at present, technical conditions are still hindering progress in this domain. I believe that with the advent of digital technology, it will be possible to convey music visually, thus adding more information for those who cannot hear. Until that day comes, subtitlers will need to be sensitive to sound and music to be able to decode their inherent messages and to find adequate and expressive solutions to convey such sensations verbally. Even though it may be difficult to find words that fully convey the expressive force of sound and music, the translator (subtitler) should try to produce an analogous aesthetic effect (Nord, 1996: 83), as well as to reproduce the expressive content of the original in the case of thematic music. Conclusion To conclude, it seems appropriate to address the issue of SDH anew in order to bring to the fore commonly held and often erroneous notions about SDH: that SDH is intralingual subtitling; that the Deaf and HoH are one and the same group; and that these specially conceived subtitles are for hearing impaired viewers alone. In the first case, and until fairly recently, this might have been a fact. Indeed, and particularly on television, SDH has mainly been the intralingual transposition of speech from the oral to the written mode. However, interlingual SDH might now be found on DVD, even if as yet only in a small number of films and languages. This may lead us to believe that common interlingual subtitles are sufficient and are used by hearers and hearing-impaired alike. The latter assumption may be 166 Josélia Neves true, but the former can only be a misconception, for Deaf viewers will not only be reading subtitles in a second language but will be relating them to yet a third lingua-culture, that of the original audiovisual text. It would appear obvious that the future of interlingual subtitling for the deaf and HoH lies ahead for, as Díaz Cintas (2003: 200) puts it: ... failing to account for this type of subtitling would imply a tacit acceptance of the fallacy that the deaf and hard of hearing only watch programmes originally produced in their mother tongue, when there is no doubt that they also watch programmes originating in other languages and cultures. This in turn would mean that they are forced to use the same interlingual subtitles as hearing people, when those subtitles are, to all intents and purposes, inappropriate for their needs. This takes us to the second notion that, regardless of the languages that might be at stake, SDH serves the needs of Deaf and hard-of-hearing alike. As mentioned above, it is misleading to consider the needs of each of these groups similar in nature. It goes without saying that most of the time neither of the two gets its due in what is presently offered. It would be excellent to be able to arrive at a compromise that would not be too taxing to either of the two. However, in an ideal world, these two audiences should be given different subtitles. Whilst that ideal situation is not yet with us, it seems only natural to me that one should try to address those at the far end of the spectrum, the Deaf, in the knowledge that the hard-of-hearing will be the ones left to choose between two not very adequate solutions. If they choose to follow SDH they will obviously need to make light of the extra care that has been put into the making of such subtitles without succumbing to the feeling of being patronised. In fact, such subtitles were made with a different audience in mind and should, therefore, be read as such. As for the last notion, that SDH can only be of help to Deaf and HoH viewers, this could not be any falser. Indeed, the extra information about sound in intralingual SDH and the care with reading time and reading ease, might be felt as quite unnecessary for the majority of people, but it might be of use to those who, for some reason, do not master the target language. In this group I would include poor readers; nonnative speakers trying to learn the language of the subtitles; and a vast number of people with cognitive impairment. At this point in time, only DVD can offer differentiated solutions in up to 32 available tracks, but with television going digital and Interlingual Subtitling for Deaf and Hard-of-Hearing 167 interactive, audiences will be further fragmented and will be looking for a translation solution that best suits their specific needs. This will mean that subtitles will be viewed on call, and open subtitling, as we now know it, will no longer make sense in its undifferentiated existence. The option of a translation solution will obviously require that multiple solutions (dubbing, interlingual subtitles, intralingual and interlingual SDH, adapted subtitling) be produced for each audiovisual programme. The question of financial feasibility and cost effectiveness is clearly pertinent. Language transfer of any kind costs money and it is difficult to foresee whether providers of audiovisual materials will be willing to pay for various versions of the same programme to suit all needs and preferences. In short, the special care that goes into making SDH easily read would most certainly be welcome in all sorts of subtitling solutions and particularly helpful to many viewers who, not being hearing-impaired, are less proficient readers. The extra information about sound, music and paralinguistic features, quite unnecessary features for hearing viewers, might be seen to raise aesthetic awareness and to those who are competent audiovisual consumers provide an additional information channel. I truly believe that, rather than SDH, subtitles for the Deaf (SD) are essential for this particular group, helpful to the HoH, and not necessarily disruptive or annoying to many hearers, provided that they are made with sensitivity, sensibility and coherence. Notes 1. See Hands & Voices (www.handsandvoices.org/resource_guide/19_definitions.html) for detailed definitions. 2. The original Le Cri de la mouette was published in 1994 and has since been translated into a great number of languages. It is widely accepted by the Deaf community as being one of the most illustrative accounts of the difficulties the Deaf person finds in trying to be part of the hearing society. 3. In the knowledge that the Deaf use smileys in their SMS messages, a set of eight smileys were used in Portuguese subtitles to depict emotions and tone of voice. After being tested on cards with a restricted group of Deaf people, they were introduced in the subtitling of the Brazilian soap Mulheres Apaixonadas, aired daily in Portugal by SIC, a commercial channel which introduced SDH, via teletext, for the first time on 6 September 2004. :( :) :-/ :-S :-& sad happy angry surprised confused (drunk, dizzy, ...) 168 Josélia Neves ;-) ;-( :-O :-º irony /second meanings acting out to be sad or angry screaming speaking softly References Conrad, R. (1979) The Deaf School Child. London: Harper & Row. Cruse, A. (2000) Meaning in Language. An Introduction to Semantics and Pragmatics. Oxford: Oxford University Press. De Linde, Z. (1995) ‘ “Read my lips” Subtitling principles, practices and problems’. Perspectives: Studies in Translatology 3(1): 9–20. Díaz Cintas, J. (2003) ‘Audiovisual Translation in the third millennium’. In G. Anderman and M. Rogers (eds) Translation Today. Trends and Perspectives (pp.192–204). Clevedon: Multilingual Matters. Di Francesca, S. (1972) Academic Achievement Test Results of a National Testing Program for Hearing Impaired Students. Series D, 9. Washington, DC: Office of Demographic Studies, Gallaudet College. Gambier, Y. (1996) ‘Introduction: la traduction audiovisuelle un genre nouveau?’. In Y. Gambier (ed.) Les transferts linguistiques dans les médias audiovisuels (pp. 7–12). Villeneuve d’Ascq (Nord): Presses Universitaires du Spetentrion. Gambier, Y. (2003) ‘Screen transadaptation: an overview’. Génesis – Revista de Tradução e Interpretação do ISAI 3: 25–34. Gielen, I. and d’Ydewalle, G. (1992) ‘How do we watch subtitled television programmes?’. In A. Demetrion, A. Efklides, E. Gonida, and M. Vakali (eds) Psychological Research in Greece: Vol. 1 Development, Learning and Instruction (pp. 247–59). Thessalonica: Arestotelian University Press. Gutt, E.A. (1991) ‘Translation as interlingual interpretive use’. In L. Venuti (ed.) (2000) The Translation Studies Reader (pp. 376–96). London and New York: Routledge. Kovač ič, I. (1994) ‘Relevance as a factor in subtitling reductions’. In C. Dollerup and A. Lindegaard (eds) Teaching Translation and Interpreting 2 (pp. 245–51). Amsterdam and Philadelphia: John Benjamins. Kovač ič, I. (1995) ‘Reception of subtitles. The non-existent ideal viewer’. Translatio, Nouvelles de la FIT-FIT Newsletter XIV(3–4): 376–83. Laborit, E. (1993) Le Cri de la mouette. Paris: Robert Laffont. Laborit, E. (1998) The Cry of the Gull. Washington: Gallaudet University Press. Luyken, G.M., Herbst, T., Langham-Brown, J., Reid, H. and Spinhof, H. (1991) Overcoming Language Barriers in Television: Dubbing and Subtitling for the European Audience. Manchester: European Institute for the Media. Monaco, J. (1981) How to Read a Film. The Art, Technology, Language, History, and Theory of Film and Media. New York and Oxford: Oxford University Press. Nida, E. (1964) ‘Principles of correspondence’. In L. Venuti (ed.) (2000) The Translation Studies Reader (pp. 126–40). London and New York: Routledge. Nida, E. (1991) ‘Theories of translation’. TTR – Traduction, Terminologie, Rédaction. Studies in the Text and its Transformations 4 (1): 19–32. Interlingual Subtitling for Deaf and Hard-of-Hearing 169 Nord, C. (1996) ‘Text type and translation method, an objective approach to translation criticism: Review of Katharina Reiss. Möglichkeiten an Grenzen der Übersetzungskritik’. The Translator 2(1): 81–8. Nord, C. (2000) ‘What do we know about the target-text receiver?’. In A. Beeby, D. Ensinger and M. Presas (eds) Investigating Translation (pp. 195–212). Amsterdam and Philadelphia: John Benjamins. Quigley, S. and Paul, P.V. (1984) Language and Deafness. California: College-Hill Press. Reiss, K. (1971) ‘Type, kind and individuality of text. Decision making in translation’. In L. Venuti (ed.) (2000) The Translation Studies Reader (pp. 160–71). London and New York: Routledge. Rodda, M. and Grove, C. (1987) Language, Cognition and Deafness. London: Lawrence Erlbaum Associates. Savage, R.D., Evans, L. and Savage, J.F. (1981) Psychology and Communication in Deaf Children. Toronto: Grune & Stratton. Toury, G. (1980) In Search of a Theory of Translation. Tel Aviv: The Porter Institute for Poetics and Semiotics. Vermeer, H.J. (1989) ‘Skopos and commission in translational action’. In L. Venuti (ed.) (2000) The Translation Studies Reader (pp. 221–32). London and New York: Routledge. 13 Audio Description in the Theatre and the Visual Arts: Images into Words Andrew Holland Introduction The UK’s Royal National Institute of the Blind (RNIB) has defined audio description (AD) as ‘an enabling service for blind and partially sighted audiences [ ... ] describing clearly, vividly and succinctly what is happening on screen or theatre stage in the silent intervals between programme commentary or dialogue – in order to convey the principal visual elements of a production’.1 Essentially, it is an attempt to make accessible a work of theatre – or any other audiovisual production – for an audience who are either blind or have partial sight by giving in a verbal form some of the information which a sighted person can easily access. In broad terms, a description in the theatre consists of two main elements. The first is a description of the set and costumes – the design or style of the production. This information is usually given before a performance, and may be pre-recorded. The second element is a description of the action which takes place during the play. In theatre, this has to be given live in order to accommodate changes in pace which are an integral part of live performance. Audience members are present in the auditorium and wear a headset (which can use infrared or radio technology) to listen to the description. This experience is very similar to the way in which you might listen to simultaneous interpreting. The crucial skill for the describer here is to time the description so that it does not interfere with the words being spoken by the actors from the stage. For films and recorded TV programmes, of course, the description of action is timed and recorded to run alongside the soundtrack. 170 Audio Description in Theatre and Visual Arts 171 The process of audio describing – some examples The simple definition of description as ‘an enabling service’ which concentrates on the ‘principal visual elements’ of a play, film, or work of art, does not get us very far in understanding how to describe something. It leaves a number of questions unanswered. In theatre we might ask the question: What makes a character? Is describing what they wear and what they look like enough? In the visual arts the same question pertains: By describing the physical characteristics of a work of art, are we really describing the art? It was Braque (1952) who noted: ‘One can explain everything about a painting except the bit that matters’.2 Before going into these questions, I want to consider three examples of description, all of which have something in common. The first is from Act 1, Scene V of Twelfth Night by William Shakespeare, and occurs when Olivia quizzes her steward Malvolio about a young man at her gate: ‘What kind of man is he?’ Olivia asks. ‘Why, of mankind’, comes Malvolio’s reply. ‘What manner of man?’ ‘Of very ill manner; he’ll speak with you, will you or no.’ ‘Of what personage and years is he?’ She enquires. ‘Not yet old enough for a man, nor young enough for a boy; as a squash is before ‘tis a peascod, or a codling when ‘tis almost an apple. ‘Tis with him in standing water between boy and man. He is very well favoured, and he speaks very shrewishly. One would think his mother’s milk were scarce out of him.’ The other two examples are from the book The Man Who Mistook His Wife For A Hat, by Oliver Sacks.3 When examining a patient who appeared to have problems with his sight, Dr Sacks handed him an object to identify. The patient described it as (1985:13): ‘A continuous surface [ ... ] infolded on itself. It appears to have [ ... ] five outpouchings, if this is the word [ ... ] A container of some sort ? [ ... ] It could be a change-purse for coins of five sizes’. A second object was described like this: ‘About six inches in length [ ... ] A convoluted red form with a linear green attachment [ ... ] It lacks the simple symmetry of the Platonic solids, although it may have a higher symmetry of its own’. What do these three extracts have in common? They certainly give us details about what or who they are describing, but do they clarify or confuse? 172 Andrew Holland In the first example, Malvolio is being deliberately obtuse, and his answer tells us more about him than it tells us of the ‘young man at the gate’. We could perhaps make a stab at his age, but his height? Build? Hair colour? Skin colour? Looks? Clothes? Demeanour? Social standing? None of these details are divulged. In the other two examples the descriptions are certainly detailed and closely observed, but do they communicate anything at all to a third party? How much easier would it have been for us to understand, if we knew that the ‘continuous surface, infolded on itself, with its five outpouchings’ was, in fact, a glove. Sacks (1985: 9) tells us about his patient that he could only see parts, not the whole picture: ‘His eyes would dart from one thing to another, picking up tiny features, individual features [ ... ] A striking brightness, a colour, a shape would arrest his attention and elicit comment – but in no case did he get the scene-as-a-whole’. In contrast, an audio description aims to provide the details that allow the picture as a whole to be formed. Working for the theatre Vocaleyes (www.vocaleyes.co.uk) was set up in 1998, following two research projects I undertook for the Arts Council of England. It was constituted as a charity to provide professional audio description services to theatres and to increase the provision and quality of access to visually impaired people. At the time, although a number of theatres provided audio description, this was mostly done on a volunteer basis, and the describers were resident in theatres. Initially Vocaleyes served the needs of touring theatre. The rationale behind this was that with the volunteer describers being resident in theatres, a production which toured to, say, ten of these theatres would be described by ten different sets of describers. At each theatre the describers would have to write the description from new and some would have very little time to prepare. Some descriptions would be good and some not so good. Some describers would have had training, others would be self-taught. The language used would be different; the actions described would be different too; the ability of the describer to time their script and to present it in sympathy with the mood of the play would vary from place to place. No other part of the artistic output would be subject to these regional variations. The result from the point of view of access is that a blind or partially sighted person attending each of the ten theatres on this tour would experience ten totally different productions, whereas their sighted companion would experience the same production ten Audio Description in Theatre and Visual Arts 173 times. Access, then, is not neutral. It changes the entire nature of the artistic experience. Vocaleyes aimed to create a high quality professional service by placing the describers with the touring production rather than with the receiving theatre. Having the describers move around the country with the production meant that the description could remain consistent from theatre to theatre, and even develop as the describers became more and more familiar with the production, and as the production itself developed over the length of a tour. Placing the describer nearer the creative process, it also allowed describers to become familiar with certain directors and the specific challenges presented by their work, making them more able to describe – in Dr Sacks’ phrase – ‘the picture as a whole’. Impartiality When I trained to be a describer we were constantly reminded to be ‘impartial’ and ‘objective’. Our job was to ‘say what we see’. The more description I undertake, however, the more impossible I feel that position to be. We were warned away from talking to creative people: ‘If you talk to the director or the actors’, we were told, ‘they’ll tell you what they want you to see’. While there is a certain amount of logic in this assertion, there is a counter claim that this information – what they ‘want you to see’, their artistic intention if you like – is a valuable thing to know about. To my mind, a describer should wish to know as much as possible about the art they are describing, a position which appears more obvious in other fields which have nothing to do with theatre or the visual arts, such as quite literal fields, for example cricket. There are limited parallels between audio description and cricket commentary. In some ways they are the complete opposite: in cricket, not much happens and the commentators have an awful lot of time to describe it – so they talk about the weather and the buses and the birds – whereas in theatre, masses happens and the describer has no time at all. In cricket, the action is spontaneous whereas in theatre (most theatre at least) the action is rehearsed and repeated. But a good cricket commentary relies on the commentator not only knowing the rules – what’s going on – but also what the two teams are trying to achieve by their various tactical manoeuvres. A commentary given by someone who knew nothing of the game would make no sense. The listener would not know what was being attempted, or what achieved. The whole narrative of the 174 Andrew Holland game would be lost and instead it would be reduced to isolated, abstract actions with bat and ball – in the same way that the glove was reduced to isolated, abstract physical details. Now, to some people cricket is just a whole load of isolated, abstract actions with bat and ball, but not to followers of the game, and it is these people who the commentary is there to serve. The same is true in theatre. The people who come to the theatre have an interest in it or they would not be there. They will not all have worked in the theatre, in the same way that cricket followers have not all played cricket. But unless the person who describes the action of a play has a good understanding of theatre as a whole the narrative of the action will be reduced to a string of meaningless details. So describers should not be afraid to engage with the creative team on a production. They should find out as much as they can. Once in a position of knowledge, it is then up to the describer to choose whether to use this information or not. Decision-making in audio description I would like to offer an example from a production of Henrik Ibsen’s Ghosts (see for example, Michael Meyer, 1980). The set was the garden room of Mrs Alving’s country house. It was a large, square room, with walls about thirty feet high. The walls were panelled with wood and painted a pale, grey-blue. In the wall to the left, a tall door led through to a dining room. In the wall to the right, a slightly smaller door led to a hallway. Most of the rear wall of the room was cut away, and the room extended back to form a glass conservatory, beyond which was a view of bleak, steep mountains. Placed around the walls were a number of simple, white-painted wooden chairs. In the centre was a white, circular pedestal table. I talked to the director about the set and he said to me: ‘It’s a room with no history’. While I could understand what he meant, I felt that the phrase was too loaded to use directly. It didn’t actually describe anything that was there. I only knew what he meant by the phrase ‘It’s a room with no history’ because I could see the room. But the phrase did bring out something about the atmosphere the room gave off; how ordered it was. Something that was an important part of the design. There was nothing in the room specific to its owner, Mrs Alving. No pictures or photographs or ornaments on the mantlepiece. No knick-knacks. No clutter. So while the phrase ‘This is a room with no history’, would have a limited meaning to a visually impaired person Audio Description in Theatre and Visual Arts 175 on its own – and indeed could be interpreted a number of ways which might be misleading – a description could express the concept through the physical details of the room. A description could structure the details in such a way as to emphasise the ordered sterility, to bring out how the distant bleak mountains and the grey, flat sky made the room feel isolated, cut off. This feeling, which is being reinforced to a sighted audience all the time they are watching the play, needed to be present in the earlier description so that the set had an artistic force, and was not just a geographical placing of walls and furniture. In short, so that it was a ‘set’ and not just a room. A decision which describers also have to make is where to strike the balance between the literal truth and imaginative truth of something. A production of Philip Pullman’s The Firework-Maker’s Daughter, for instance, used a host of imaginative illusions in its telling of the story; a physical style of theatre for which the production company Told by an Idiot is well known. As the heroine Leeyla lights the fuse of a firework, a woman dressed in red, with red gloves, their fingers ending in red metallic tassels, comes on to the stage and moves along the fuse, her tasselled fingers waggling to become sparks. The two levels – one literal and one imaginative – coexist in the same moment, and there is comic potential in the fact that the spark is, in fact, a human being. In another scene, Leeyla is on a journey. She wears a simple Chinesestyle costume with a conical straw hat. As she walks, an actor takes the hat and holds it at a distance so that it resembles the volcano to which Leeyla is travelling. The actor pulls a party popper underneath so that streamers fly up through a hole in the hat, like the volcano erupting. As sighted viewers we are able to appreciate this moment both in terms of the narrative and how it is technically achieved. We are admiring the picture that is created as well as the artistry involved in its creation. Any description of this moment therefore has to do justice to the physicality of the performances as well as supporting the narrative. It needs to reflect the technical or literal truth as well as the imaginative truth. Too far in one direction and we miss out on the performance style. Too far in the other direction and we wonder why someone has stolen Leeyla’s hat. But where to strike the balance will vary from play to play. Let us take a look at two quick examples from a stage production of Peter Pan. Peter and the Lost Boys find themselves at the Mermaid’s lagoon, a shimmering bay of green-blue sea which laps against crystal cliffs. In the centre of the bay is a large pock-marked rock which at low tide resembles a human skull. During the play, characters dive extravagantly off the 176 Andrew Holland rock into the sea, swimming around, with their heads bobbing in the waves as they tread water. At least this is the illusion. The reality is that the sea is made from sections of green-blue cloth which stretch out from the rock in the centre of the bay to the cliffs at the back. Each section of cloth is edged with elastic, and is constantly moving to recreate the lapping of the sea. When the characters dive, they are actually jumping down through the gap between two billowing cloth sections on to a crash mat beneath, then poking their heads up between the cloth and the crystal cliffs, as though treading water. The question is, which version of this description should we give? The first suggests the water is real, which it is not. The second is prosaic and undermines the illusion. We do not want crash mats in this beautiful lagoon. Again, a sighted audience is able to sustain both worlds simultaneously, so in the end we went for something of both. The introductory notes described both the way the sea worked technically and the impression that was created. During the play itself, we supported the illusion of it being sea, while slipping in the odd word which reminded people of the way the illusion was achieved. The other big question in Peter Pan is how to describe the flying. No matter how clever the mechanism involved, an audience is always aware that it is not ‘real’. But to describe this great moment in literal terms – ‘Peter, Wendy and the boys get hooked up on to wires and are then hoisted into the air, kicking their legs and waving their arms about as though flying’ – would be to seriously short change our audience. So again, we need to find a way to do both, dispensing with the technical stuff before the story begins, so that when the time comes and Peter sprinkles his magic dust, the children do what those characters do – and fly. The live nature of theatre I mentioned earlier that unlike cricket, where the action is spontaneous, in most theatre the actions are rehearsed and repeated night after night. But this is not always the case; a fact which again emphasises the advantage of working closely with a company and knowing how they work. Some directors, for instance, do not block their actors’ moves. This means that from one performance to the next, the actors will – externally – do different things. If you only have the opportunity to see the production once or twice, description can be tricky as describers are constantly being surprised. They have timed the description of a certain action to fit within a certain pause. But in the next performance, not Audio Description in Theatre and Visual Arts 177 only is the action different but the length of the pause is too. This is an unusual way of working and may seem confusing to some. But to the director, it emphasises the live, unfixed nature of theatre which distinguishes it from other art forms. You do not want to reheat the same meal every night but to use the same ingredients to cook it afresh. Understanding something of this approach is important if you want to describe more than just the actions. It means that, as a describer, you can begin to look at the production differently, focusing on the internal motivations of the characters rather than their outward behaviour. An example of what happens when you do not adequately reflect motivations occurred during one description I heard. In David Halliwell’s play Little Malcolm and His Struggle against the Eunuchs (1967), art student Malcolm holds court in his dingy flat, discussing art and revolution with his college mates. But despite the bitterness of the weather, he does not light the metered gas fire, telling his friends that he has no money. He sends them off on an errand and as soon as they leave, Malcolm takes money from his pocket, crouches by the fire and lights it, warming his hands. While he is doing this a new character, Anne, comes in. The staging of this production highlighted Malcolm’s selfish duplicity. As he crouched by the fire with his back to the door, the door slowly opened, making it appear to us that he might get caught. The tension was then dispelled when another character – one we had not met before – entered. But Malcolm himself was clearly rattled by the possibility that it could have been his friends returning. That he could have been caught. The description of this moment, though, remained entirely literal, giving us a sequence of external actions with no underlying message: ‘The friends went out. Malcolm lit the fire. Anne came in’ – and all the subtleties of the carefully constructed moment were missed. Not only did the description fail to articulate the drama, it also failed to give us Malcolm’s reaction, and therefore to extend our knowledge of his character. The narrative was reduced to a series of actions, but description needs to be more than this. Training audio describers But if we cannot ever be wholly objective in our work, we can try to be non-judgemental. When training theatre describers I use a number of exercises where trainees describe photographs or illustrations. This is to consider a number of issues which affect description generally – how information is given and received – and to separate them off from those issues which relate specifically to audio description in the theatre. 178 Andrew Holland A couple of years ago I was training describers in a regional theatre and we were looking at some Victorian Spot illustrations, the kinds of sketch, sometimes humorous, which were used to illustrate books or pamphlets. In one particular drawing, a man walked along a street. The trainee’s description went along these lines: ‘He is wearing an overcoat which is not of the first quality. He carries an umbrella, for effect. And a briefcase which probably contains little more than his sandwiches’. The coat, umbrella and briefcase were all there, certainly, but I asked the trainee how she knew, for instance, about the sandwiches in the briefcase? ‘Because I know men like this!’ she said. In this example the describer is judging the character rather than describing him, and coming up with a whole lot of detail the picture itself did not contain. There was no evidence that the overcoat was ‘not of the first quality’ – no fraying of the collar, no marks of wear, no sagging hem. And the two other details could not possibly be obtained by looking at this image alone. If the next frame, so to speak, had shown the man in the rain, would his umbrella have been up? And there was certainly no way, using sight alone, that we could know the contents of the man’s briefcase. Of course, were the man a character in a play, we would be given a whole lot more information about him. He might, during the course of the action, open his briefcase and reveal ‘little more than sandwiches’. But this might be intended to surprise an audience. In this case it is the action which reveals something, not the man’s physical appearance. So a balance has to be struck which recognises that although we cannot be wholly impartial, we do have a duty to be non-judgemental. But is this balance the same in every work? Christopher, the narrator in Mark Haddon’s novel The Curious Incident of the Dog in the Night, makes many very astute observations about the difficulties of describing people, especially where emotions are concerned (2003: 19): I find people confusing. This is for two main reasons. The first main reason is that people do a lot of talking without using any words. Siobhan says that if you raise one eyebrow it can mean lots of different things. It can mean ‘I want to do sex with you’ and it can also mean ‘I think that what you just said was very stupid’. Siobhan also says that if you close your mouth and breathe out loudly through your nose it can mean that you are relaxed, or that you are bored, or that you are angry – and it all depends on how much air comes out of your nose and how fast and what shape your mouth is Audio Description in Theatre and Visual Arts 179 when you do it and how you are sitting and what you said just before and hundreds of other things which are too complicated to work out in a few seconds. This seems to me to get to the heart of the problem. Using our sight we are exposed to hundreds of physical hints at any one time. We are processing this information to work out attitudes, emotions and relationships. This processing is constant and, for the most part, we are unaware of it. In the theatre we will be interpreting and re-interpreting a character all the time she is on stage; all the time she is talking; all the time she is listening; all the time she seems disengaged with the central action. Sometimes even by her absence. A describer has two problems. The first is that the description cannot be continuous. To fit around the dialogue of a play, a describer may have only a couple of seconds to sum up a particular emotion. If the description concentrates purely on the physical details – the raised eyebrow, for instance – this may not tell a visually impaired member of the audience what the entire facial expression suggests to a sighted audience member. The second problem is that the information is verbalised and can only be given in a linear way whereas sight, although focusing on certain details, can also hold other details in the background. Verbalising something gives it a prominence in that one thing is mentioned and not another thing. This means that it is the describer who is choosing what the audience should focus on. During an intense dialogue between Henry and Stella, Henry walks over to the window. Well, how does he walk? Is the move casual or is it more of a stomp? Is he displaying the same impetuosity we have seen in him before? Or is this behaviour something new to us? This arrogant strut, is this a side of his character we have not seen before? And is the fact that Henry is going towards something important? Is it more important that he is moving away from Stella? Or that he is turning away so that she cannot see his face? Or perhaps our focus should be on Stella. Is her reaction to what Henry does more significant? For each seemingly simple moment there will be many possible descriptions and, as I said before, there may be only one or two seconds in which to describe it. The mere fact that as describers we have to choose to emphasise one thing rather than another, means that – whether we like it or not – we are making an artistic decision. We are contributing to how a piece of art is experienced by a member of the audience. The more we know about the piece of work, the more likely we are to make the right choice of what to describe. 180 Andrew Holland Audio description in the visual arts The issue of ‘interpretation’ becomes more prominent when describing the visual arts, if for no other reason than the fact that in theatre a visually impaired member of the audience is integrating the information given by the describer with any residual sight they have and also with what is coming from the stage aurally. A description will often be only minimal because the voice of the actor is doing so much. With a piece of visual art, however, less is likely to come from the work itself. This means that the description will take on a greater role in interpreting as well as describing the material. In 2002, Vocaleyes worked with the RNIB and English Heritage on the Talking Images project. We undertook three case studies to examine the use of audio description within the visual arts, working with galleries who were interested in opening up their collections for blind and visually impaired visitors. We wanted to examine the language of description, and cover a range of work from historical to contemporary and from figurative to abstract. An example from one of the case studies will serve to illustrate something of the process we went through. This involves the language of description, specifically how ‘interpretative’ a description should be. For the first case study we worked on an exhibition of drawings and late reliefs by the abstract artist Ben Nicholson, curated by Kettle’s Yard in Cambridge. Working with a group of visually impaired users, we tried out a number of different descriptions of the same work. This first version attempted to concentrate on the forms of the work and remain as much as possible ‘un-interpretative’: 1968 Ramparts (oil on carved board) A rectangular background, some 19 inches high and 21 inches wide – that is about 48 by 53 centimetres – is painted a smooth earthy brown. Standing proud of it is a slightly smaller rectangle – this one divided up into a number of smaller, overlapping shapes. At top and bottom are areas of white. Between them a line of three differently sized rectangles. The one to the left is brown like the background. The central one is a darker brown, and the third, a lighter, orangey brown. The line created by these three rectangles starts off – to the left – as horizontal and almost central. But a little way across, the line shifts so that the two rectangles centre and right slope downwards. To the Audio Description in Theatre and Visual Arts 181 far right is a tall rectangle – painted the same brown as the background. Two other forms seem to float above the relief. Their colour is similar to the two white sections. Both are similar in shape – a trapezium – with parallel sides, horizontal tops, but with a bottom edge which slopes down towards the right. One is positioned within the top white section and to the right. Its slanting edge runs along the top edge of the slanting brown line. Carved within it is a circle – the inner edge painted white. The other trapezium sits next to it – just left of centre – and a little lower. In this, another circle has been inscribed rather than cut. What does this rather literal description tell us? It was greeted by the focus group with a stunned silence. Then, after a little while one of the group said: ‘Well, you’re on a hiding to nothing there’. And they were right. There is a possibility that after listening carefully to this description a number of times we could perhaps make a stab at reconstructing what it looks like: which shape fits where; what overlaps what; something of its composition. But what is the point of that? As a sighted person I do not go through that process. I do not look at a work in order to memorise what goes where. My encounter with the work is on a completely different level. I am looking for signs of intent, of humour. I am responding to certain elements of composition or colour. I am asking questions about why I respond in the way I do. As a sighted person I have some kind of emotional relationship with the piece. I experience the arrangement of physical details – it is not enough just to know they are there. The second description that we tried out with the focus group allowed for a level of interpretation. Colours, for instance, were given tactile qualities which reflected the physical process the artist used to make the piece. So the areas at top and bottom became areas of ‘frosty, silvery white – scratched and rubbed in places to create an uneven surface, like snow drifting across dirty ice’. Another area had been ‘roughly scraped so that the grubby brown of the hardboard shows through like a stain’. With the three differently sized rectangles in the centre, colour was described in terms of how it is perceived by the viewer – the visual tricks it plays. So, the central, darker rectangle with its blacker sheen ‘makes it seem to sink back away from us into the relief, although in reality it stands proud of the one to the left’. The line of rectangles to which this belongs ‘progresses in a line across the work from left to right’. Their shift downwards ‘as though a 182 Andrew Holland geographical fault has sheered this layer and pushed it bodily down. Now sloping, these two rectangles seem in danger of sliding out of the composition – squeezed out from between the frosted white sections at top and bottom’. The language creates a narrative within the piece, and makes reference outside the world of the work to our own world to try to capture some of the work’s dynamic. The sense we get as sighted viewers is of the shapes somehow acting on each other. Although they are static, the angles, different depths and contrasting colours of the various geometric shapes create a sense of slow, powerful movement. There are internal forces which although dynamic, achieve a kind of balance or equilibrium. As a sighted viewer I experience both the feeling of movement and the feeling of stasis. So to describe the piece simply in terms of the relative position of each individual shape is to reduce it to pointlessness. It could be argued that the references to snow, ice and geographical faults are a personal judgement which cannot be justified. But the focus group got much more from this description than the earlier, uninterpretive version. They felt they had had some kind of encounter with the work. They felt that it was worth listening to, whereas the first version, although accurate, gave them nothing at all. The interdependence of the senses Back in 2004 I attended a talk given by experimental psychologist Dr Charles Spence in which he described a series of experiments which showed how interdependent our senses are. People’s experience of touch, for instance, is affected by sight and sound. In one experiment, subjects were asked to touch a number of pieces of sandpaper which they could not see. Wearing headphones, they could hear the sound their finger made as it rubbed the sandpaper. The experiment showed that the subject’s experience of the sand-paper was directly influenced by the sound-frequency that was played to them as they touched the paper. ‘Rougher’, they said, and ‘smoother’, even though the sand-paper stayed the same. The fact that the senses are interdependent is something that artists have used in their work for centuries, and much figurative art goes to great lengths to transcend the visual. Berenson (1896: 5) writing about Giotto stated how: ‘Psychology has ascertained that sight alone gives us no accurate sense of the third dimension’ and that ‘every time our eyes recognise reality, we are, as a Audio Description in Theatre and Visual Arts 183 matter of fact, giving tactile values to retinal impressions’. He argued further that: Painting is an art which aims at giving an abiding impression of artistic reality with only two dimensions. The painter must, therefore, do consciously what we all do unconsciously – construct his third dimension. And he can accomplish his task only as we accomplish ours, by giving tactile values to retinal impressions. His first business, therefore, is to rouse the tactile sense, for I must have the illusion of being able to touch a figure ... before I shall take it for granted as real. Cloth has weight, weave and texture. The simple glove that we talked about earlier may not have ‘five outpouchings’ but it might, perhaps, be fine, intricate, old, coarse, supple, warm, frayed, discarded. Clothes may be rich, delicate, finely woven, roughly stitched, worn formally or with a casual disregard. Rooms have three dimensions; urns are cracked; a dog balances precariously; bodies have mass and density; are fragile, infirm; water has luminosity; bodies sweat; woodland reeks of decay. A description has to work just as hard to ‘give tactile values’ to the ‘retinal impressions’ described, to appeal to more than the visual. Conclusion This brings me back to why I feel that the definition given at the beginning of this chapter, that is, of audio description as ‘an enabling service’, is inadequate: because what it picks up on is the language of discrimination and disability rather than the language of artistic endeavour and achievement. The danger is that we regard audio description as we would a screen reader – a simple access tool, a kind of functional software, a window to meaning. With a screen reader there is a direct equivalence between the text which is written and the text which is voiced. There are few choices or decisions for the software to make. The relationship between them is clear and unambiguous. Some things may be a little hard to understand because of a strange intonation, but we can work out what the meaning is. But audio description is not like this. It can never be transparent. By its very nature it will change the experience someone has of the art. I also wonder if the term ‘enabling’ would be acceptable in any other area of the arts? When translating a play, for instance, are we simply ‘enabling’ someone with another language to know what the play is about – its storyline, its characters, its themes? Or do we want our 184 Andrew Holland audience to experience the play directly, with all its complexities? With no direct equivalence between one language and another, a translator has to make a series of decisions about which one of several possible meanings or nuances to go with. This decision will inevitably have a knock-on effect on the rest of the play, and the translator may even be accused of skewing the text in some way. But the option not to make choices, to translate literally and include, perhaps, alternatives within the text, while it might enable people to know about the text, would create an un-watchable piece of theatre. The audience would be constantly reminded that they were watching a translation of ‘something else’. They would come out of the theatre more aware of the problems faced by the translator than of what the playwright wanted to say. In the same way, audio description is a way of translating artistic material from one medium to another. There is no direct equivalence between a moment on stage and the words chosen to describe it. The exhortation to be ‘impartial’ doesn’t recognise this fact – and too often has the effect of focussing the audience’s attention on the enabling tool – forcing them to remain distant from the actual piece. For me, and for many visually impaired people, that is not enough. Description should aim to get to the heart of a work of art and to recreate an experience of that work by bringing it to life. It should not be content with telling someone the physical details of something they cannot see. When you leave the art gallery, you want to come away discussing the art, not the description. So in order to be non-intrusive a description has to make decisions and not pretend it is not there. I have tried to argue that in order to write a meaningful description, an audio describer has to do more than ‘say what he sees’, that this phrase is a nonsense which attempts erroneously to divorce ‘seeing’ from ‘understanding’, and that in this process of understanding the other senses are involved alongside sight. The quotation given earlier by Oliver Sacks seems to me to get to the heart of what a good description should be. Of his patient he said (1985: 9): ‘He failed to see the whole, seeing only details, which he spotted like blips on a radar screen. He never entered into a relation with the picture as a whole’. Good description must allow the viewer to enter into a relation with the object, person or painting being described ‘as a whole’. This means integrating the description so that it becomes part of the artistic experience, rather than keeping that experience at arms’ length. And the third example, the one the patient described as a ‘convoluted red form with a linear green attachment’? Dr Sacks suggested the patient Audio Description in Theatre and Visual Arts 185 smell it (1985:12–13): ‘Beautiful’, the patient said – ‘An early rose. What a heavenly smell!’ Notes 1. Quoted from the introduction to A National Standard for Theatre Audio Description – AUDEST report, 1998. 2. ‘Il n’est en art qu’une chose qui vaille. Celle que l’on ne peut expliquer’. Trans. Ben Nicholson. 3. The article of the same name was originally published in the London Review of Books, 1981. References Berenson, B. (1896) The Florentine Painters of the Renaissance. New York and London: G.P. Puttnam’s Sons. Braque, G. (1952) Cahiers 1917–1952, Paris. Translated by Ben Nicholson [Quoted by Peter Khoroche in his introduction to Ben Nicholson: Drawings and Painted Reliefs (2002). London: Lund Humphries, Ashgate Publishing]. Haddon, M. (2003) The Curious Incident of the Dog in the Night. London: Vintage. Halliwell, D. (1967) Little Malcolm and His Struggle Against the Eunuchs: A Play. London: Faber. Meyer, M. (Trans.) (1980) Ibsen Plays: Ghosts, The Wild Duck, The Master Builder. London: Methuen. Pullman, P. (1995) The Firework-Maker’s Daughter. London: Corgi Yealing. Sacks, O. (1985) The Man Who Mistook His Wife For A Hat. London: Gerald Duckworth. Shakespeare, W. (1623) Twelfth Night. In S. Wells and G. Taylor (eds) (1986) The Oxford Shakespeare. The Complete Works. Oxford: Oxford University Press. 14 Usability and Website Localisation1 Mario De Bortoli and Jesús Maroto Ortiz-Sotomayor Introduction Established multinationals or small companies expanding into overseas markets need a well-coordinated international sales and marketing effort in order to succeed. A website can serve as a company’s premier marketing tool, a facilitator of direct sales, and a technical support mechanism; but it can also be used for purposes of public, customer, investor or employee relations. When users are able to interact successfully with a website, positive impressions and attitudes about both the site and the associated organisation are created. Hence, having a web presence, like advertising, should make possible the boosting of a corporate image. This effect is crucial because the website may be related to branding, especially if it is a vehicle for sales. To this end, websites are often customised, or localised, for foreign markets, taking into account local language issues, business and social standards, and aesthetic preferences. Localising a website is a complicated but necessary task. The idea of making different versions of a site for other cultures demonstrates a willingness to show the consumers that the organisation is willing to accommodate their needs. According to research carried out by Hayward and Tong (2001), users perceive a company more favourably (that is, as more trustworthy, more likeable) when they see a version of the company’s website in their mother tongue, regardless of the user’s proficiency in the English language. Written text plays a crucial role on the web, as most websites – particularly corporate sites – are content-based. Too many companies have found themselves in trouble by entrusting their translation to someone in the company who has travelled the world and is ‘fluent’ in six languages, or to people who happen to be bilingual but have no 186 Usability and Website Localisation 187 localisation background whatsoever. We have all had the experience of laughing at copy poorly translated into English. However, failing to gain market share because of linguistic issues is not a laughing matter. In any case, successful localisation involves a lot more than simply translating content. The manner in which people carry out tasks can differ from culture to culture. For example, approaches such as the ‘shopping trolley’ or ‘shopping cart’ metaphor found on many websites may not transfer accurately to some cultures, which may dramatically limit the usability of an e-commerce site, thus reducing potential revenue in those countries. Having made large investments, companies should not allow their websites to be downgraded in their international versions. Fortunately, there is a growing acknowledgement amongst international businesses that each of their foreign markets is best served by its own culturally specific website. Furthermore, there is an important financial case behind this awareness since non-English speaking Internet users alone represent over 64% of the world online population (Global Reach).2 The fact is that although many international businesses have had their English websites translated into the languages of their main foreign markets, some of these sites have not performed as successfully as their home versions. For a site to be well received and successful – today more than ever – it should address those intangible aspects that make a group of people a community, rather than deal exclusively with customising the obvious, superficial items such as measurement units or currency. Failure of localisation Quite apart from the linguistic dimension, localising the content of a website is not easy; from a technical point of view, localising web content poses some daunting challenges. Websites come in many shapes and forms, from a few pages of HTML created in basic text editors to vast scripted or database-driven sites. Internal company sites, intranets, are also becoming more popular for the private dissemination of information in a structured manner. Timeliness and up-to-the-minute content are rapidly becoming the key discriminators of a company’s website, and as the web is a global phenomenon, the speed at which this content is localised is also becoming an issue for many companies. In addition to the linguistic and technical issues, website localisation also faces cultural challenges. According to Hofstede (1991), culture dictates how people from a specific location view and react to images and messages in relation to their own patterns of acting, feeling and 188 M. De Bortoli & J.M. Ortiz-Sotomayor thinking, often ingrained in them by late childhood. Any differences in these patterns are displayed in the choice of symbols, rituals and values of any given cultural community. A culture influences the perceptions, thoughts and actions of all its members, and it is this shared influence that defines them as a group. These various sets of ideas and expectations that culture provides are all brought to bear when interacting with technology. Dialogue between humans and computers is constrained not only by the design laws of the computer, but also by the user’s understanding of the world and its norms. If the design of a computer system does not match the user’s understanding of the task in hand then the interaction between the two will be sub-optimal. Products designed in one culture for use in another often fall into this category. This is generally the result of the two common errors in the localisation process discussed below. Designers do not necessarily know about other cultures The first stumbling block encountered by most localisation projects is the limits of human intuition about other people. However much we believe we know about a group or an individual, as human beings we tend to be extremely poor at anticipating their requirements. When considering the needs of users from another country, even well intentioned designers may be unaware of their own bias and ignorance about the people of that culture. Often they are unable to filter out interface features, which may handicap users from other cultures (Fernandes, 1995). Some designers attempt to circumvent these problems by enlisting a friend or colleague who has lived in the target culture and can speak the language. Unless these individuals have spent a considerable amount of their time in that country working on markedly similar projects, it is unlikely that their insights will be any more accurate than those of anyone else. In reality, successful localisation begins with the recognition that we do not necessarily know the requirements of other cultures when a project is launched. What is really needed is a systematic approach to gathering information about the users of a particular product or service in other countries. Cosmetic changes are not enough The second most common mistake is to pay attention only to superficial differences between cultures in the belief that this represents an Usability and Website Localisation 189 adequate attempt at localisation. In order to improve the quality of what is essentially guesswork, designers tend to use guidelines, helping them address the features which vary superficially across cultures. These guidelines cover such aspects as basic differences in currency and calendar. Guidelines often serve to give designers and management the mistaken impression that the website has been localised and effectively ‘fireproofed’ for cross-cultural usability problems. In fact, culture actually influences interaction with computers at levels significantly deeper and less observable than the use of local calendars or currencies. Human Computer Interaction (HCI) research shows that successful interaction depends on more than just using the correct language. For authors like Bourges-Waldegg and Scrivener (1998) and French and Smith (2000), interaction is also dependent upon the culturally embedded meaning of objects such as icons and metaphors – for example, the desktop or the shopping cart. Whilst the USA and the UK share a common language, a website which, for instance, has recourse to the metaphor of the white pages (the US phone directory) to help users find individuals’ contact details may not be appropriate for use in the UK, where yellow pages are the standard. This is despite the fact that to all appearances the site may not appear to be in need of localisation. What is needed, therefore, is a new definition of effective localisation and its scope. Extending the scope of localisation through the inclusion of HCI expertise As mentioned above, website localisation efforts have traditionally been concerned mainly with translation and character encoding issues. However, it is our belief that this alone is not sufficient to meet the technology needs of users from other cultures. Indeed, there is a considerable amount of evidence detailing the difficulties and failures experienced by users of culturally inappropriate systems. The extent of human diversity is such that the mere translation of an interface from one language to another is not always sufficient to meet the needs of another culture. HCI approaches model systems from the user’s perspective and therefore are well placed to inform localisation requirements for a site if employed early enough in the design process. During the 1990s cross-cultural HCI research expanded from issuing guidelines and importing models from the social sciences (Hall, 1976; Hofstede, 1991) 190 M. De Bortoli & J.M. Ortiz-Sotomayor to developing its own frameworks (Bourges-Waldegg and Scrivener, 1998; French and Smith, 2000). Papers with a global dimension regularly feature in all major HCI conference programmes. However, despite this explosion of research interest, the number of designers using HCI support for cross-cultural interactive systems remains low. Additionally, a worrying number of misinterpreted theories have been imported piecemeal from other fields such as social science, linguistics and cognitive psychology. This is not uncommon in interface design where imported theories are often adopted and cited by designers unaware of the research background. These theories then gain credibility within the design community at the expense of other findings (Green et al., 1996). There is a clear need for culturally sensitive technology, a need which is currently not being met. What is obvious is that designers should now be expected to add localisation to their skill-set. The interdisciplinary nature of localisation means that it does not lend itself to ‘one-size-fitsall’ solutions, which can be learnt and applied in identical fashion to all projects. The background of each applicable theory and the subtleties of local culture and language must be understood fully. Whether the product in question is software, a website or a mobile phone, preparing the user interface for use in an international context calls for expertise from a variety of fields. Designers are required for their creative abilities. Equally, properly qualified linguists and translators are necessary not just to translate content, but also to ensure that the essential meaning of each message is communicated adequately. However, successful interaction cannot be boiled down to a simple matter of aesthetic preferences and translation. This ignores the behaviour on the part of the user: people all over the world have different, culturally rooted responses to stimuli and act accordingly. For example, Chinese consumers seem to prefer shopping at online stores that offer the possibility of bargaining in spite of the price they actually end up paying sometimes being greater than that at other stores (Liang and Doong, 2000). HCI professionals utilise a range of approaches, from cognitive models to usability testing and user-participatory methods, which help determine the requirements of different cultures. These can be applied to make recommendations on the subsequent process of localisation. If properly implemented at an early enough stage by professionals with the correct knowledge of localisation issues, these recommendations can significantly improve the end result, increase return on investment and improve the relationship of a brand with its target group of consumers. Usability and Website Localisation 191 The market In many ways, the choice of one brand over another is increasingly becoming a political choice by consumers, who express a whole range of values when buying a product. Therefore, for the product to be successful it must represent and address the consumer’s concerns and beliefs. And this is where a simple literal translation can never be enough. Each market has to be scrutinised and culture-specific solutions found. Consumers only respect a brand that respects them. There are many opportunities for brands that are willing to listen to their markets and are prepared to go the extra mile in the localisation process. Brands have to understand that markets are there to be seduced, not patronised, and the only way to seduce them is to know how they think and feel, and act accordingly. From a marketing perspective, brands need to realise that being international means that each section of their target market is as important as the home one. German or Polish website users believe that they are entitled to as much attention from a brand as their American or British counterparts. After all, they may be paying just as much money for the product. At the end of the day it is for the seller to make an effort in a transaction, not the buyer. Generally speaking, the recipe for international success is to convince people that the product is good and that it was produced with the consumer’s needs in mind. Each linguistic community will need to have this explained from a slightly different angle, determined by priorities of the local culture. Problems in this process occur when English-speaking, and especially American, businesses tend to confuse increasing Internet use worldwide with increasing Americanisation of other cultures. There seems to be the misperception that if people in another country have access to the web, they will already have been exposed to enough Western influences to be able to use sites of Anglo-American origin. In fact, despite its history, there is nothing inherently American or even Western about Internet use, as evidenced by the latest estimates for Internet users by language growth stating Chinese has grown by 622% and Arabic by 2062% between 2000 and 2008 (Internet World Stats: online).3 All of the above means that an international campaign, or a multilingual website, cannot be researched and developed in English, and then sent for translation, an attitude that is viewed as arrogant by many. As stated earlier, superficial approaches to localisation do not adequately meet the needs of users in different countries, and do not create the positive awareness that companies crave for their brand. In order for localised sites to perform effectively, differences in culture must be 192 M. De Bortoli & J.M. Ortiz-Sotomayor reflected in the design of each one. To achieve this, the localisation process has to be integrated at a much earlier stage of the planning or creative process than is now done. Every aspect should be discussed and studied before its development, and only then implemented. This represents a revolution in localisation as we know it at present – but a necessity for most companies at a time of increasing dependency on foreign markets. In fact, a better-integrated localisation approach might mean significant savings on post-production adaptations via economies of scale and the pooling of assets and resources via new technologies, as well as greater branding consistency and, as a result, greater return on investment. The reduced cost of localisation achieved in this way might, for example, also allow entry into previously unviable markets. Conclusion Products such as websites expose companies to global markets, but few of them pay adequate attention to the vast audiences outside their own borders. Among those that do, most are only prepared to pay the absolute minimum for translated versions of their main site. Whilst superficially localised, these sites do not usually fulfil their intended function as they may remain culturally unsuitable. Localisation must start from an in-depth knowledge of the local culture and requirements, then address those requirements within the framework of existing local structures. Cultural differences affect interaction at levels significantly deeper than language. Addressing these differences requires the early attention of professionals with expertise in a variety of fields, including user-centred design, HCI, psychology and linguistics. Only a synthesis of this expertise can guarantee the reaping of the enormous benefits that globalisation can offer to truly international businesses. Notes 1. This chapter would not have been possible without the assistance and cooperation of Robert Gilham, HCI consultant at Amberlight Partners. 2. An online marketing multilingual website at http://global-reach.biz 3. Information available on www.internetworldstats.com/stats/.htm References Bourges-Waldegg, P. and Scrivener, S.A.R. (1998) ‘Meaning, the central issue in cross-cultural HCI design’. Interacting with Computers 9(3): 287–309. Usability and Website Localisation 193 Fernandes, T. (1995) Global Interface Design: A Guide to Designing International User Interfaces. Boston, MA: AP Professional. French, T. and Smith, A. (2000) ‘Semiotically enhanced web interfaces for shared meanings: Can semiotics help us meet the challenge of cross-cultural HCI design?’ IWIPS 2000. Baltimore, USA, 23–38. Green, T.R.G., Davies, S.P and Gilmore, D.J. (1996) ‘Delivering cognitive psychology to HCI: the problems of common language and of knowledge transfer’. Interacting with Computers 8(1): 89–111. Hall, E.T. (1976) Beyond Culture. Garden City, NY: Anchor Press. Hayward, W.G. and Tong, K.K. (2001) ‘The influence of language proficiency on web site usage with bilingual users’. www.psy.cuhk.edu.hk/~usability/ research/HaywardTong.pdf Hofstede, G. (1991) Cultures and Organizations: Software of the Mind. Maidenhead: McGraw-Hill. Liang, T. and Doong, H. (2000) ‘Effect of bargaining in electronic commerce’. International Journal of Electronic Commerce 4(3): 23–43. This page intentionally left blank Part IV Education and Training This page intentionally left blank 15 Teaching Screen Translation: The Role of Pragmatics in Subtitling Erik Skuggevik Introduction It is an inescapable fact that subtitling, as a form of interlingual translation, must simplify information, if not wherever possible, then wherever necessary. This need to simplify arises because subtitles do not replace the original language of the film, they coexist with it, as well as with the other audio and visual channels of the film, and even – should the subtitles be too complex – compete with them. Thus, having to match reading speed to speaking speed invariably leads to choices about what to prioritise in subtitles, and this forms an equal part of the whole translation process. The subtitlers, in many ways, have nowhere to hide. They present their translated rendition of whatever is spoken at the precise moment when it is said, and any viewer with a grasp of the original language is able to make an instant comparison. This is what terrifies any novice, and even some experienced subtitlers; a fear that can manifest itself in too rigid an adherence to the original syntax and to the denotative values of the spoken words, often resulting in – amongst other things – over-long subtitles and literally-translated metaphors. The confidence to translate subtitles with a degree of economy might improve with experience and sensitivity, but as an alternative to the intuitive skills acquired in this process, we can also develop or demonstrate such awareness through subtitling pragmatics. In this chapter we shall be looking at the speech act analyses of Grice (1975) and Jakobson (1960), and consider a number of aspects that, to varying degrees, may be said to form part of the emerging field of subtitling theory. 197 198 Erik Skuggevik The five levels of competence in subtitling In order for us to delineate the field into which we are about to delve, we may briefly outline five levels of subtitling competence. Grice’s cooperative principle and its maxims can be associated with the third, while Jakobson’s speech act theory can be aligned to the fourth. 1. The first level is technical competence, that is the ability to deal with the sheer practical demands of the job as it appears to most working subtitlers: use of software, line breaks, positioning on the screen, time and space restrictions, use of italics, etc. This is an area where mistakes are easily quantifiable, as there are clearly identifiable rights and wrongs. 2. The second level concerns the linguistic skills of the students; the translators’ expertise and sensitivity to their own and to the Source Language (SL). These are language skills that students will have to draw on in any translation work. 3. The third level refers to the translators’ understanding of social and cultural (non-linguistic) aspects and the awareness of their relative values. Any number of dictionaries cannot be a substitute for the hands-on experience of living and breathing the way of life of another culture, its language use and the hierarchies of social values. 4. The fourth level is possibly the most elusive analytically but also the most universal: comprehension of the psychological or emotional dimension inherent in the action that accompanies the spoken words. 5. There is, in my opinion, a fifth level which is the competence which allows taking all these previous areas into consideration in a holistic exercise of determining strategies based upon the limitations and possibilities on offer, in order to formulate any given subtitle. I shall return to some additional considerations about this fifth level of competence at the end of this chapter. Within the discipline of pragmatics, many of these dimensions are considered as part of the importance that context, readership, cultural expectations and implicatures play in communication. However, as we shall see, subtitling places rather specific demands on the type of analysis needed since its main concern is the transfer of the communicative act itself, not in its entirety (like literary translation), but partially, as the visual and audio channels are running concurrently and are complementary. Role of Pragmatics in Subtitling 199 Subtitling and implicatures One problem area characterising subtitling results from violations of the maxims of the cooperative principle (Grice, 1975), or differently put, the problems of transferring the operative differences of these maxims between language-cultures without having recourse to the social or cultural landscape in which they exist. We should note here that the terms conventional (or standard) implicature versus conversational implicature may be most easily recognised through reference to denotative and connotative values respectively. There is a close parallel distinction between stated meaning (as both denotative value/standard implicature) and implied meaning (as both connotative value/conversational implicature), and although there are instances where the two sets of distinctions have their different uses, the overlap is sufficient to explain why both sets of terms are used at times interchangeably. For our purposes, the term ‘standard’ (Levinson, 1983) is used in preference to Grice’s ‘conventional implicature’, to make it visually (and phonetically) easier to distinguish from conversational implicature. In order for us to investigate the usefulness of Grice’s formulations of the cooperative principle and to what extent it applies to subtitling, it might be convenient to summarise the initial list of maxims provided by Grice (quoted in Katan, 1999: 211): ● ● ● ● The maxim of Quantity: give as much information as needed. The maxim of Quality: speak truthfully. The maxim of Relation: say things that are relevant. The maxim of Manner: say things clearly and briefly. According to Grice (1975), these maxims will inform the choices made by both speakers and listeners when deciding how to communicate or interpret information: as standard implicature when no maxim has been violated, or as conversational implicature when one or more of them have been flouted. Looking around, particularly at social interaction in the Western world which we may identify as problematic (from our own experience or from observation), it becomes evident that there are a few other possible maxims in operation: ● ● ● Be polite. Do not embarrass. Do not dominate. 200 ● ● Erik Skuggevik Do not be too subservient. Do not exaggerate your status. The fact that there are maxims other than the ones suggested by Grice is not surprising. He freely admits that his list is not exhaustive, and hints that there may be additions. That there may be hierarchies involved in the importance of the maxims is also something Grice (1975: 46) concedes himself when he states that: ‘It is obvious that the observance of these maxims is a matter of less urgency than the observance of others; a man that has expressed himself with undue prolixity would, in general, be open to milder comments than a man who has said something he believes to be false’. In other words, it is worse to lie than to talk too much. But this may not be the order of priorities in every culture, as clearly pointed out by Baker (1992: 233): In some cultures “Be polite” indeed seems to override all other maxims. Loveday (1982: 364) explains that “No” almost constitutes a term of abuse in Japanese and equivocation, exiting or even lying is preferred to its use.’ If this is true, it would suggest that the maxims of Quality and Manner are easily overridden by considerations of politeness in some cultures. Grice’s maxims, and their hierarchical potential, must then be seen as culture-specific, and although they provide a method of making explicit the motives and inferences of speakers, they do not assist translation and are in general better suited to describe communicative failures than prescriptive strategies. Indeed, it is worth registering that the relationship between the cooperative principle and its maxims is analogous to the relationship between the Saussurean langue and parole, where Grice’s definitions are firmly rooted in the description of the parole of maxims and implicatures. In the field of commercial film translation, the market is dominated by US (and to some extent British) films, and for many European cultures the Anglo-Saxon hierarchical organisation of Gricean maxims and their own have large areas of overlap, some more than others. Below, however, we shall have a quick look at a particular difference in nuance between Norwegian and English maxims. Norwegian and English variation The gap between the operative maxims in American or British societies and Norwegian society is generally very narrow. However, one Role of Pragmatics in Subtitling 201 difference lies in the maxim of being institutionally (or minimally) polite. When subtitling a large number of occurrences of statements ending in the words ‘please’ or ‘thank you’ into Norwegian (Swedish and Danish would be much the same) the tendency is often to omit these words, and this is not just because it is space-saving: to include the Norwegian semantic units vennligst, takk or er du snill (denotative/ connotative equivalents of ‘please’ or ‘thank you’) might often seem excessively polite, and could therefore activate the connotative, Norwegian interpretation (conversational implicature) of irony or sarcasm. A scene at an English restaurant where Mr A says, ‘Could I have the menu, please?’ presents us with a standard implicature in English, but a Norwegian Mr B who says, Kan jeg få menyen, takk/er du snill? runs the risk of invoking the conversational implicature ‘I need a menu and I am getting impatient’. The omission of ‘please’ therefore seems logical, and rarely presents a problem in Norwegian subtitles. Indeed this difference in linguistic code is summed up in what many Norwegians describe as a British preoccupation with ‘please’ and ‘thank you’. Occasionally, however, the violation of this (institutional) politeness maxim becomes foregrounded, as when the withholding of a ‘thanks,’ which would normally have been omitted in a Norwegian subtitle, gives rise to a conversational implicature in English. The following scene takes place in a restaurant, where the waiter attending a table finds the guest quite uncooperative: ENGLISH NORWEGIAN - Would you like pepper? - Vil du ha litt pepper? - No. - Nei. - I was only asking. - Jeg bare spurte. - And I said no. - Og jeg sa nei. To a British audience it is not difficult to see that the reason for the waiter’s remark (and he gets our sympathies) was prompted by a feeling of ‘a void’ in the normal communicative pattern. It is virtually obligatory to say ‘No, thanks’ when declining an offer in English, whereas in Norwegian, judged by the subtitles at least, the feeling created is that the waiter is over-sensitive and that the guest has a valid point. One might say ‘No, thanks’ in Norwegian as well, but then it is an added 202 Erik Skuggevik degree of politeness. In Britain it is obligatory. While the English version violates the maxim of politeness, the Norwegian does not. Since Grice’s proposal offers no methods of transferring maxims between cultures, we shall return to this point after we have looked at Jakobson’s communication functions. The kind of difference in cooperative hierarchies discussed above could in theory be formulated between different languages and cultures as the closer two cultures are, the less divergence between the maxims. The further removed from each other cultures are, the more important it becomes to use strategies to act as a form of intermediary area in which comparisons of relative value might be conducted with greater ease. In order to provide a possible Rosetta Stone between the difference in relative value of culturally defined maxims, we can turn to an analysis of communication, essentially more psychological than cultural, based on the assumption that there must at least be a high value attached to the maxim of Relation (be relevant), as indeed communication seems pointless if there did not exist a maxim that addressed the question ‘why are we talking at all?’. Here Jakobson’s (1960) speech act analysis, originally a linguistic model, may provide us with an approach making explicit the dynamics; this is important in cases of violation of the maxims of Grice’s cooperative principle, that is when culturally codified conventions are challenged, to the point when the subtitler’s skill will be put to the test. Communication in subtitling Jakobson (1960) proposes that there are six necessary components of any speech act. I will extend these to encompass all forms of communication as the functional principles involved are exactly the same. Jakobson’s components are obligatory in the sense that if any one of them is removed no communication can take place. They are illustrated in Figure 15.1. Sender Figure 15.1 Code Contact Context Message Jakobson’s six components Receiver Role of Pragmatics in Subtitling 203 We shall briefly list some examples of these six categories: ● ● ● ● ● ● Sender (or addresser): Me, the Prime Minister, a person from the sixteenth century, etc. Code: Spanish, mathematical indexes, legalese, cockney, morse code, body language, etc. Contact: Person to person, a newspaper, Internet, telephone, runes on a stone, TV, etc. Context: Attending a lecture, sitting on the underground, walking the dog, etc. Message: (The lexical/semiotic content) ‘Meet me at 11pm’, ‘There will be tax rises’, etc. Receiver (or addressee): You, the President, the phone operator, a person in the twenty-third century. As an example of a situation involving these components we could pick anything that may be viewed as communication, but the one that you, in your present role as reader, are experiencing will serve as an illustration. You (the receiver) are reading this paper, the factual content of which is the message, written by me (the sender). The contact that makes this possible is the book you have in front of you. The context is the situation you are in right now, sitting at home, at college, on the bus – maybe you have been asked to read the paper or found it by accident. The code employed is English, somewhat academic, specialised in the field of Translation Studies. Remove one of these ingredients and you will not be able to make any sense, contact or inferences from your experience (although the removal of context alone is an impossible proposition, as there can be no context-free experience). These formal components of the communicative act can be seen as universal. The recognition and identification of each one of them (albeit subliminally) lies at the core of any analysis of communication. For instance, in order to correctly interpret any violation of a Gricean maxim, it would be necessary to first – and this is an unspoken assumption in Grice’s proposal – eliminate the possibility that there may be a breakdown in communication, due to faults in the code, mode of contact, misunderstandings of context, or misidentification of receiver/sender. Only then can any non-adherence to accepted maxims be given an inference or implicature, and any message interpreted accordingly. Once the presence of these six components is established, it becomes clear that at different times any one of them may seem more central to the tenor of a communicative act. Translation theory scholars have 204 Erik Skuggevik Metacommunicative Code Phatic Emotive Sender Contact Referential Conative Receiver Context Performative Message Figure 15.2 Jakobson’s six functions of communication1 long recognised that particular linguistic features may be linked to a corresponding problem or situation – e.g. Bühler (1934: expressive, appellative, informative); Reiss (1971: informative, expressive, operative). Each of Jakobson’s constituents may be seen as capable of providing the focus for a type of communication, either because there is a semantic problem (code, contact), a particular objective (sender, receiver), an informative purpose (context) or a particular aesthetic imperative (message). As acknowledged by Jakobson (1960), communication displays certain features when it gravitates towards each of the six constituents, and he ascribes to them the functions in Figure 15.2. Jakobson’s functions could easily form the basis of a further classification of ‘text types’ leading to a more comprehensive version than Reiss’s model, (informative, expressive and operative – in Jakobson’s terms referential, emotive and conative), although Reiss (1971: 172) does not believe this is a fruitful approach. Nord (quoted in Munday, 2001: 74)2 added the phatic function to Reiss’s text types, leaving Jakobson’s performative (poetic) and metacommunicative (metalinguistic) functions to be incorporated into a potentially more complete system. However, Jakobson himself never suggested that these functions should define text types as such. On the contrary, the functions, as a dominant factor of any communication, are in constant flux and in any act of communication the focus could shift quickly from one to another, as different approaches may be employed in order to achieve a particular goal: cajole, tempt, flatter, threaten, reason... For subtitlers this is of course a most familiar occurrence. Let us briefly look at a summary of Jakobson’s (1960) definitions: 1. Emotive: When the communication is focused on the sender (addresser), expressing how s/he feels and thinks: ‘I am totally Role of Pragmatics in Subtitling 205 2. 3. 4. 5. 6. opposed to this plan!’, ‘Never mind your husband, what about me?’, ‘I love this place’, etc. Metacommunicative (metalinguistic in Jakobson’s original usage): Communication about communication (redefinitions, reformulating, explaining words) or when the code itself is in focus, see the example below from the film Meet the Parents. Phatic: When language is used just to establish or maintain contact as in ‘How are you doing?’, ‘Nice day today, isn’t it?’, etc. Referential: It takes place when the communication is appropriate to the context, e.g. when in a meeting, it is appropriate to discuss the agenda; when lecturing, it is appropriate to impart information. Referential communication is focused on something external, independent facts, descriptions, relevant information, etc. Performative (poetic in Jakobson’s original usage): Communication that indulges in the poetic, aesthetic or musical qualities of the code. It is important to remember that this does not necessarily imply ‘something beautiful’, just that communication works performatively. Most modern adverts do in fact perform visually and musically by associating aesthetic values with their product (Barthes, 1977). Conative: Not to be confused with ‘connotative’. Communication focused on the receiver’s (addressee’s) needs and situation: ‘Don’t worry, nothing bad will happen, you just take your time’, ‘Is that something you can live with?’, ‘I won’t forget you’, etc. Multiple choices As far as subtitling is concerned, it is of course imperative to recognise that it is not the words themselves which necessarily determine the communicative function of a statement, but the way they are given expression; their relationship to tone of voice, body language and situation. As an example, we shall look at a phrase subtitlers often encounter, ‘I’m OK’, heard in many films, from which several examples are given below. ‘I’m OK’ simply seen on a page will make most people interpret it (by a kind of ‘cultural collocation’) as a standard implicature: ‘I am fine, no need to do anything’. But the phrase is often uttered to express entirely different sentiments. These are indicated, most importantly, by context, mood and body language, something that becomes clear when attempting to subtitle a large number of versions of ‘I’m OK’ 206 Erik Skuggevik into Norwegian. In the following six examples, different connotations of ‘OK’ mean that they cannot be expressed by one and the same phrase: 1. Referential: factual response to context. Nate asks Lisa whether she wants another drink. She smiles, shakes her hand and replies: ‘I’m OK’ – Nei, takk [No, thanks]. 2. Conative: seeking to reassure. Claire puts her arm around Russell, who is visibly upset following an emotional ordeal. He is shivering, looks blankly straight ahead and says quietly: ‘I’m OK’ – Det går bra / Jeg klarer meg [It goes well / I’ll manage]. 3. Performative: the delivery of the message is rhythmically adapted to give sensual or poetic expression to the words. Asked how he is feeling, John, sitting in a hot Jacuzzi, sipping champagne, replies in an overly slow, breathy voice, pausing between each word, lingering for a long time on the letter ‘o’: ‘I ... am ... ooooo ... kay’ – Jeg ... har det ... deilig! [I have it lovely]. 4. Emotive: statement of feeling. Someone runs jubilantly out from a doctor’s surgery, having just received news that s/he does not have some suspected disease, shouting to a friend: ‘I’m OK!’ – Jeg er frisk! / Alt er bra! [I am well (healthy)! / All is good!]. 5. Phatic: confirming that the communication channel works. The chairman of a telephone meeting checks that all the connected participants are hooked up, and asks: ‘Can everyone hear me?’. Several of the participants answer: ‘I’m OK’ – Alt i orden [All is in order]. 6. Metacommuncative: terms used to illuminate other terms. A confused foreigner asks a friend: ‘What does “I am swell” mean?’ The friend replies: ‘It means “I’m OK” ’ – Det betyr: Jeg er ok [It means, I am OK]. Although the Norwegian subtitling choices above no doubt might have been expressed differently by another subtitler, the fact remains that there is no way of translating the words ‘I’m OK’ into Norwegian (or any other language) without considering the underlying connotative message. Code In the early stages of the film Meet the Parents we have a situation where the daughter brings her new boyfriend home to meet her parents. However, the reunion serves to isolate the boyfriend as he witnesses a Role of Pragmatics in Subtitling 207 stream of private language, childish nicknames and small ‘games’ between father and daughter: Hi, Daddy! Hi! - I missed you so much, Pamcake. - I missed you too, Popjack. ← Daughter runs up to dad. ← Father hugs. ← Daughter Oh, boy! Oh, boy! Oh, boy! ← Father, as he swirls daughter around in an embrace. Shortstack, shortstack, coming up! ← Both father and daughter as they ‘stack’ hands. Where’s my widdle girl? ← Mother, with apron appearing in doorway. Mommy! My mom! ← Daughter, goes to hug mother. Should we choose to analyse the scene above in terms of Gricean maxims (even if Grice does not deal with conventions or the cooperative principle in relation to group dynamics), there are a number of maxims in such interactions that are clearly relevant to Greg, the ‘forgotten’ guest. There certainly is a breach of the maxim of relation (relevance), because to Greg, the whole scene becomes an embarrassment as it does not involve him at all. The scene also demonstrates a breach of the maxim of politeness. But Grice’s maxims cannot tell us (for translation purposes) in what way the language in the scene communicates that sense of exclusion. Following Jakobson’s categories, what we see in this extract is the practice of a kind of specialised, somewhat regressive, childish language, practised by the family only, and in Jakobson’s terminology, this scene has at its centre its very own code. Greg, the alienated outsider, has no access to this private language. How do we translate words like ‘pamcake’, ‘popjack, ‘shortstack’, ‘shortstack coming up’, or ‘my widdle girl’? The words are confusing, certainly, if all you have is a dictionary. Although there are referential references in the dialogue, at the centre, most importantly, is the use of a private code, a code that constitutes a secret family bond. This is functionally what we need to hold on to. The translated subtitles should not ignore the alienating effect of this private code because Greg’s alienation and embarrassment is what the film revolves around from that moment on. Now imagine you are watching a film sequence where Mr Jones and Mrs Smith are talking about aeroplanes. But they are both pilots. 208 Erik Skuggevik Perhaps their conversation is, purely phatic, that is, not so much about aeroplanes but just about talking to kill time, or to just stay in touch. That can give the subtitler good guidance in what choices to make, to avoid language that accidentally makes it sound more technical than necessary, or at least make sure that the tone of the translation reflects the phatic function of the scene as a whole. Or imagine the following scene in a film: a coal miner exhausted after a terrible ordeal down a burning mine shaft exclaims: ‘I think I need a cup of tea’. On the face of it, this seems very clear. The statement appears referential in nature, has a clear outside physical component and is grammatically straightforward. Indeed, I cannot imagine too many different ways of translating this line into Norwegian. But imagine for a second that in your country drinking tea is something that not many people do and those who do belong to the social elite. So against this background a coal miner wanting a cup of tea would seem a bit strange. What do you do? In the UK, having a ‘cuppa’ can be associated with caring, hospitality, safety and normality. The expression ‘I think I need a cup of tea’ in this situation is not primarily referential in nature, it concerns itself with the situation: it is an expression of the speaker’s feelings. The function, according to Jakobson, is emotive, that is, it has to do with the feelings of the speaker and is actually expressing a need (albeit with well-known British understatement) for comfort and normality. If both good knowledge of the English institution of tea or the English language is lacking in the Target Language (TL), there may be a distinct advantage in translating this request for tea in a more general way. But how? Given that we identify the function of this statement as emotive, we should remain within that sphere, and we could maybe consider: ‘I need something soothing/relaxing to drink’. The meaning of ‘No’ We have discussed, above, the situation where a waiter might have felt slightly offended by the guest’s failure to say ‘no, thanks’ when asked whether he wanted pepper on his food. We recognised that there was a violation of the Gricean maxim of politeness which caused this offence in English, whereas in Norwegian the expectation is virtually absent. Admittedly, this is not a major problem in Norwegian subtitling, but all the same, it is interesting to analyse further. How should we view the importance of the flouting of the maxim of politeness in this situation? According to Jakobson’s functions, ‘thanks’ and ‘please’ are an institutionalised form of conative behaviour since politeness (as standard Role of Pragmatics in Subtitling 209 implicature) usually carries a conative function. Omission of this obligatory response makes the guest perform an emotive act; his answer has become self-centred. In Norwegian it is more culturally acceptable to give an emotive or referential answer to an offer with no conative element. When subtitling this dialogue, we should endeavour to reflect the guest’s rudeness, as was the waiter’s interpretation. How do you make the emotive ‘no’ more emotive or impolite without adding more words? One method is simply to add an exclamation mark (more commonly used in Norwegian). This would not interfere with the word economy, and could almost imperceptibly accentuate the unexpectedness of the guest’s response. This strategy could also work in English: - Would you like pepper? - No! - I was only asking. - And I said no. At the beginning of this discussion we mentioned the possibility of a fifth, holistic, subtitling level of competence and some further remarks on the subject might seem appropriate at this point. A number of related observations have been made by Ivarsson and Carroll (1998) and Gottlieb (2003), but we shall have a look at two which have received limited attention. Dynamic coexistence As language users we have a relationship to words which is not only technical or semantic, but also sensuous or, at least as far as the speaker is concerned, performative. In subtitling we need to give consideration to a certain mirroring of dynamism between what is said and how it reads. By ‘dynamic coexistence’ I refer to the relationship between the spoken and the written word; for a subtitle to sit comfortably with what we hear, we also need to get a reasonable semblance of the aural impact of the words. What we are referring to here is how we read a written text. As pointed out by De Linde and Kay (1999: 20), it is important that we take into account the inner voice that makes people, even when reading, pronounce the words: Studies in homophonic reading support the notion that inner speech is significant in the reading process. Most studies confirm that the speed and/or accuracy of silent reading is influenced by the sounds 210 Erik Skuggevik of words. Likewise, sentences containing a series of words that rhyme or alliterate (tongue-twisters) are more difficult to read both silently and aloud. We need to remember that subtitles are a ‘one-way street’ in the sense that, unlike words on a page, we cannot go back and check or re-read. A subtitler must therefore ensure the viewer gets the message the first time around by not overcrowding the subtitles with phrases or words that require concentrated reading. Most subtitlers recognise that the performative element inherent in swearing or slang must be given attention but so must the rhythm of idiomatic expressions or the poetic nature of amorous communication. The notion of dynamic coexistence might also rule out the possibility of translation on plain readability alone, as when letter combinations and words read with such difficulty that – never mind the correctness of the semantic and grammatical equivalence – the viewer simply cannot disentangle the subtitle before it has disappeared. A consideration of the dynamic coexistence of speech and subtitles should also ensure that we let short expressions stay short in translation, or give a semblance of verbosity in the subtitle if the speaker is verbose. Sometimes, by chance, it happens that an entire sentence of as many as ten words can be expressed accurately by a short phrase in translation, but much as subtitles strive for economy of expression, in such cases one might have to add words simply in order to mirror the time it takes to read a subtitle with the time it takes for it to be spoken in the original. The importance of body language Although originally presented in order to define poetics, Jakobson’s functions are far from being just linguistic. In a visual medium such as a film the viewer will be bombarded with clues as to what is going on. Communication is also body language and usually coincides functionally with the spoken word. Somebody stroking someone over the head will most often match that with words fulfilling the same function – in this case the conative, centred on the receiver. A person wildly waving their arms about in discussion will normally use words that reflect the emotive function; someone using their hands to delineate or draw shapes and spatial relationships will equally be inclined to using referential language; a handshake is the physical equivalent of establishing verbal contact, both given a phatic function, etc. Sometimes the subtitler may be struggling with a particularly obtuse or unusual phrase, Role of Pragmatics in Subtitling 211 sentence or sequence to translate, even if they have a general idea of what the meaning of it is. In such instances it may be a useful strategy to simply switch off the sound and view communication at the level of body language: who is relating to whom, in what manner, what are the power dynamics in the situation, etc. This exercise can often unlock – or at the very least confirm – some of the psychological relationships and messages in a particular scene. We do this in our own lives every day, only then we cannot switch the sound off! However, if body language does not match words, the effect is of course entirely different. It is quite impossible to imagine a person with a referential body language, using his hands in a manner of somebody making a business presentation, while at the same time declaring his undying love for his girlfriend. If somebody’s physical action is stroking another person’s hair (conative), but the concurrent verbal action is talking about the weather (phatic), or themselves (emotive), then there is a jarring of functions. As viewers, we may sense that a film sequence expresses something rather odd without knowing how or why. However, this way of analysing may give a clue about how to combine different elements in order to reflect what is odd. These kinds of relational dynamics may be structured as shown in Figure 15.3. This is obviously virtual reality and it is not suggested here that such tables should be drawn up; there are far too many variables in communication to do this. But it is possible to make explicit what functional relationship the spoken words have to a particular audiovisual sequence as an aid to solving a particular subtitling problem. In cases where there is a clash between body and verbal messages, the subtitler must of course only focus on the expression of the verbal element, as the interrelationship of physical and verbal actions are intentionally at odds, a priority that also holds in cases of poor-quality acting. Verbal function Physical function Phatic (talking about the weather) + conative (comforting) = Insincerity Emotive (talking about oneself) + conative (comforting) = Self-centredness Conative (declaring one’s admiration for someone) + referential (using hands factually like a politician) = Patronising Figure 15.3 Cumulative functions Cumulative effect 212 Erik Skuggevik Conclusion The process of breaking down the communicative functions of filmic dialogue, with a view to transferring them to subtitles, can be seen as a largely structuralist activity, and as such amounts to more than just pragmatic considerations. Indeed, subtitling choices are not simply modified by the associative functions we have discussed but are, in fact, defined by them. In subtitling, the balance between action, meaning and words needs to be accurate, which is successfully incorporated by the fifth competence mentioned in our discussion. When confronted with a subtitling task, we must ask ourselves: ‘What is most important here?’, and the various areas outlined as levels of competences need to be considered in reverse order. First, we need to understand the communicative function, and consider how to formulate this expression. Secondly, we must try to align this expression with the cultural associations of the situation, including any possible violations of Gricean maxims. Thirdly, we must try to get close to the meaning of the spoken words themselves. But in our search for the words, or the right cultural associations and potential implicatures, we must not forget the communicative function, since this is the most basic of any equivalence we might be trying to recreate.3 By looking at the different interpretations and translations of the phrase ‘I’m OK’ we have shown that creating equivalence at word level may not be possible. In all translation acts there will be a compromise, but what we cannot ignore when subtitling is the communicative function performed by a phrase or sentence. Notes 1. Jakobson uses the terms ‘metalinguistic’ and ‘poetic’, which in this discussion, have been replaced by ‘metacommunicative’ and ‘performative’ in order to move from a linguistic to a more general semiotic dimension. 2. No doubt Jakobson (1960) must have been widely read, including by Nord, but his typology does not seem to have had enough impact to make text type analysis incorporate all six functions. 3. The existence of equivalence at a more general level, for example between subtitle and filmic dialogue, is arguable, at least in a conventional way. It must be remembered that subtitles are not in themselves a complete product as they are symbiotically tied to the film. They are the result of an interlingual repositioning of selected elements of the original verbal content taken from, and reintroduced back into, the overall narrative of the film. It is also a basic premise that for equivalence to be a relevant concept, a replacement of the original needs to have taken place. Strictly speaking, subtitles live a life of coexistence, not replacement. Role of Pragmatics in Subtitling 213 References Baker, M. (1992) In Other Words. London: Routledge. Barthes, R. (1977) Image, Music, Text (Transl. Stephen Heath). London: Fontana Press. Bühler, K. (1934 [1967]) Sprachteorie: Die Darstellungsfunktion der Sprache. Stuttgart: Gustav Fisher. De Linde, Z. and Kay, N. (1999) The Semiotics of Subtitling. Manchester: St. Jerome. Gottlieb, H. (2003) Screen Translation: Six Studies in Subtitling, Dubbing and VoiceOver. Copenhagen: University of Copenhagen. Grice, H.P. (1975) ‘Logic and conversation’. In L. Cole and J.L. Morgan (eds) Syntax and Semantics (pp. 41–58). New York: Academic Press. Ivarsson, J. and Carroll, M. (1998) Subtitling. Simrishamn: TransEdit. Jakobson, R. (1960) ‘Closing statement: linguistics and poetics’. In R. DeGeorge and F. DeGeorge (eds) (1972) The Structuralists: From Marx to Levi-Strauss (pp. 85–122). New York: Anchor Books. Katan, D. (1999) Translating Cultures: An Introduction for Translators, Interpreters and Mediators. Manchester: St. Jerome. Levinson, S.C. (1983) Pragmatics. Cambridge: Cambridge University Press. Loveday, L.J. (1982) ‘Conflicting framing patterns: the sociosemiotics of one component in cross-cultural communication’. Text 2(4): 359–74. Munday, J. (2001) Introducing Translation Studies: Theories and Applications. London: Routledge. Reiss, K. (1971) ‘Decision making in translation’. In L. Venuti (ed.) (2000) The Translation Studies Reader (pp. 168–79). London: Routledge. 16 Pedagogical Tools for the Training of Subtitlers Christopher Taylor Introduction Several research projects exploiting the potential of the ‘multimodal transcription’ as devised originally by Thibault and Baldry (2000) have pointed to useful tools for the screen translator with particular reference to the subtitler. The methodology as adapted here consists in the drawing up of a table consisting of rows and columns (see Figure 16.1) of the Italian soap opera Un posto al sole. The first column indicates time in seconds while the second column contains the individual frames of a film which may be, as in the example in Figure 16.1, of one-second duration. The third column (visual image and kinesic action) describes in some detail what is visible on the screen and what action is taking place. This is effected through a series of codes (‘D’ stands for distance, ‘CP’ for camera position, ‘VS’ for visual salience, ‘VC’ for visual collocation and ‘VF’ for visual focus) and features such as perspective, lighting effects, primary and secondary visual salience, gaze vectors and many other elements are thus described. The fourth column describes the sounds that are to be heard in each frame, including musical accompaniments and dialogue. The final column is reserved for the appending of subtitles where required. The minute description contained in the middle columns allows for a thorough analysis of the multimodal text in all its manifestations, and sensitises the analyst to the complicated meaning-making resources contained therein. The purpose of this methodology, as far as subtitling students are concerned, is that it enables them to base their translation choices on the meaning already provided by the other semiotic 214 Pedagogical Tools for Training Subtitlers 8 CP: horizontal D: medium close VS: Renato gesticulating, women passively absorbing. VC: apartment furnishings. Gaze: Renato looks at both women, looking for agreement. Lighting natural. Renato moves far right, hands still forward and moving frenetically. Questo non è un desiderio... This isn’t just a wish! 9 (as above) VF: Renato looks down at Giulia (power pose). Hands folded down, palms up. Questa è una pazzia! This is madness! 10 (as above) Giulia opens palms in gesture of frustration. Renato, calmati! Renato, calm down! 215 Figure 16.1 modalities contained in the text (visual elements, music, colour, camera positioning, gestures). This process, though time-consuming from a practical point of view, has proved to be an extremely valid pedagogical instrument. Students at Trieste University have been encouraged to carry out this type of analysis over a whole range of film types (feature films, soap operas, cartoons, advertisements, television series, news broadcasts, documentaries, etc.) as material for their graduate dissertations. The results have shown how the various film types require different treatment when translating for subtitles. An obvious example is that of the nature documentary where the correspondence between the visual and the verbal is at its highest, as compared to an animated scene from a soap opera where the dialogue may refer to many events and people outside the immediate setting. 216 Christopher Taylor However, given the impracticality of the method for professional purposes with anything longer than a short television advertisement, a second stage in the research brought in the concept of phasal analysis, following the ideas of Gregory (2002) who used this term to describe his analyses of written texts. Thus, rather than individual frames, film texts can be divided or parsed into phases, sub-phases and ‘phaselets’ based on the identification of coherent and harmonious sets of semiotic modalities working together to create meaning in recognisable chunks, rather in the manner of written texts. Such texts can be analysed in terms of their phasal construction, which includes the transitions that separate the phases and that behave somewhat like conjunctions or discourse markers in the written mode; various devices from the clear cut from one shot/scene to another, to fading in or out, to a blurred screen heralding a dream sequence, etc. are employed. This further extension of the original tool provides the basis for a thoroughgoing analysis of any film text, and will provide what is hoped will be a useful addition to the pedagogy of film translation. This chapter reports on the progress being made in refining this tool within the environment of university courses in screen translation. It also closely adheres to the ideas of Kress and van Leeuwen (1996) relating to multimodal texts and visual grammar. Discourse, Design, Production, Distribution and Interpretation (DDPDI) Kress and van Leeuwen (1996) describe the creation of multimodal texts in terms of four, not necessarily chronological, stages – discourse, design, production and distribution. They also add a final stage completing the whole process, that of interpretation on the part of the receiver of the multimodal text. Firstly they use the term ‘discourse’ to refer to an abstract level of ‘text’ that transcends even the concept of genre. In a sense it hangs in the air from where it can be channelled into the language use associated with identifiable genres and genrelets through instantiation. For example, there exists a discourse of ‘love’ which may manifest itself in talk between lovers, in written love letters, on the problem pages of tabloid newspapers or in a Danielle Steele novel or, what is of particular interest for this chapter, in a film sequence. Design involves the use of all the semiotic resources available to the multimodal text creator (words, images, music, etc.) in instantiating Pedagogical Tools for Training Subtitlers 217 discourse. It involves, for example, choosing between the verbal and the visual, or deciding the weighting of various semiotic modalities. Such design decisions vary from individual to individual and from age to age. Consider, for instance, commercial advertisements which once were a purely verbal phenomenon, a written text appearing in a newspaper, whereas now they are the embodiment of the sophisticated multimodal text, at times using a vast array of meaning-making resources. Feature films can also exploit the whole gamut of semiotic modalities, especially combining them to create meanings that transcend the significance of the individual elements. Film design thus lies in the hands of scriptwriters, set designers, costume designers, make-up artists, lighting engineers, etc. Production can be described as the organisation of the expression, the putting into practice of the designed discourse. Film production is a complicated process; film is a complex semiotic event involving many participant ‘producers’. Actors, of course, produce the spoken discourse. Cameramen are responsible for the production of the visual discourse, musicians for the soundtrack and sound engineers for accompanying effects. It is also true that production is often shared. For example, while recognising that the camera crews create the visual text, the actors also contribute in some way by dint of gesture, body language, movement and the like. The star of the film may sing to a musical accompaniment. Many other ‘producers’ may be involved and the long list of acknowledgements that appear at the end of any film is testimony to this. But then, in the final analysis, it is the director who could be said to produce the whole multimodal event, or at least exercise control over that ‘production’. The distribution stage is that in which the multimodal product is presented to, potentially, millions of receivers. Music, for example, is distributed as CDs, on television, or on the stage. Film is distributed in cinemas, on television, as video cassettes and, increasingly, as DVDs and through online streaming. Home movies, holiday videos and wedding shoots are examples of home-made multimodal texts where perhaps a small group of people handle the whole process from discourse through distribution to interpretation, but in the case of film, the number of people involved may be very large and disparate and the stages may be handled totally separately. As all communication depends on interpretation, the result of discourse, design, production and distribution needs to be interpreted by the end user. The act of interpretation completes the process, and 218 Christopher Taylor not necessarily in a predictable way. As Kress and van Leeuwen (1996: 8) point out: ‘Which discourses interpreters or users bring to bear on a semiotic product or event has everything to do, in turn, with their place in the social and cultural world’. And this is interpretation at a first remove. The question of interpretation becomes much more important when the original multimodal text is a translated product. DDPDI and film text We shall now examine Kress and van Leeuwen’s stages as they relate specifically to film text, and subsequently as they relate to film subtitling. The discourse of film on one level can be seen as the discourse of the particular theme or subject matter, e.g. Mafia discourse in The Godfather, or cowboy discourse in John Wayne vehicles, and these include the nonverbal elements such as Armani suits in the one instance and guns and holsters in the second. At another level, film language is a law unto itself and differs from real language use. Starting from the premise that film scripts are ‘written to be spoken as if not written’ (Gregory and Carroll, 1978), any comparison with real, spontaneous spoken language should demonstrate similar patterns in both. However, experiments carried out at the University of Trieste tend instead to point to the difficulty of reproducing genuine oral discourse in film. Comparing the scripts of fifty modern films set in contemporary, ‘real’ environments with an equivalent-sized sample of spoken language from the Cobuild Bank of English corpus, it was discovered that, at least in terms of a number of chosen variables, the film scripts differed considerably from the corpus sample. The use of six discourse markers (‘now’, ‘well’, ‘right’, ‘yes’, ‘OK’ and ‘so’), all typical of authentic spoken language, is much more marked in the corpus sample than in the film scripts. Other experiments that tested for the frequency of tag questions, hedges and vague language (‘sort of’, ‘kind of’, etc.) in television series produced similar results. In terms of design, in the design of a film the discourse is channelled into something with possibly multiple aims. For example the discourse of ‘love’ in the film Notting Hill provides a vehicle for the classic boy–girl development of the plot, but is also designed to entertain and amuse. It creates feelings of empathy, antagonism or eroticism by using the semiotic modalities at its disposal. The shift from script to screen with the skilful use of multimodal resources such as music, light, camera angle, etc. can create a new semiotic phenomenon. At Pedagogical Tools for Training Subtitlers 219 this point it is sometimes hard to distinguish design from production. Just as actors can break out of the scripted strait-jacket and redesign an original script (transcriptions of film dialogue often show discrepancies in relation to the original script, as the actors begin to ‘feel’ the part and, as a result, render the dialogue more authentic), they can also redesign the ‘stage directions’ and produce some semiotic element that enhances, modifies, even ‘brands’ the original discourse. Consider Bogart and Bergman’s famous kiss in the film Casablanca, as compared to the mechanical design of a porno movie. Finally, distribution occurs after the director has completed the editing process and presented a finished product. The director’s role is crucial in the development of the semiotic potential that enables film to become a recognisable ‘language’ (Kress and van Leeuwen, 1996: 85), to be interpreted by an audience whose size can never be predicted. The extent and direction of distribution depends partly on the intrinsic worth bestowed by the first three stages in the filmmaking process and the subsequent popularity or critical acclaim of the product, but also depends on commercial and political factors which are instrumental in terms of audience reach, while audience interpretation, inasmuch as it manifests itself in positive or negative response, is equally instrumental in maintaining or improving distribution levels. Phasal analysis The focus of the study now shifts back to Gregory’s (2002) concept of phasal analysis. Language functions to realise social interaction; that is, it combines the three components of the context of situation, namely field, tenor and mode (Halliday, 1994), to create discourse. In practically any text, including multimodal texts, it is possible to identify discrete phases characterised by factors that bind that particular part of a text together as a single, homogeneous unit. The three components of the context of situation (field, tenor and mode) relate directly to the three (meta)functions of language delineated by Halliday (1994): ideational, interpersonal and textual functions. The ideational function of language is that of providing interlocutors with the means to talk about the world, in the general sense, and their experiences within the world. The interpersonal function enables relationships to be established, maintained and altered and for roles to emerge in communication acts. The textual function is what allows communicators to construct discourse and to present information 220 Christopher Taylor cohesively and coherently. In terms of phasal analysis, those continuous and discontinuous stretches of discourse which can be identified as phases share ideational, interpersonal and textual consistency and congruity. A clear example is provided by any television soap opera. A number of parallel plots or stories run simultaneously, but to maintain the interest of the viewers, they are presented clip by clip in a seemingly endless succession of short alternating scenes or scene sequences. The latter can be considered discontinuous phases, often divisible in turn into subphases and sub-subphases, which retain a number of features that are intrinsic, even unique, to those phases. Thibault and Baldry (2000: 320) were the first to apply Gregory’s phasal analysis to multimodal texts, explaining that phases in this sense are characterised by ‘a high level of metafunctional consistency [ ... ] among the selections from the various semiotic systems (music, action, dialogue, etc.)’. Gregory (2002) uses the term ‘communicating community context’ (CCC) as a macro version of the concept of the context of situation in which an act of communication takes place. It refers to the dimensions of time, geographical location, ideologico-political situation, etc. The CCC therefore partially reflects and partially constructs the functioning of language in conversation, but in the case of film, the CCC surrounds a group of actors being directed to play a scene in what we shall call an ‘artificially produced situation’ (APS). This APS must relate, like all other communicating community contexts, to the participants’ knowledge of the world and, naturally, to the language they are using. However, the actors’ knowledge of the world (time, place, tradition, culture, etc.) in the form of sets of schemata variously described as scripts (Minsky, 1975), frames (Schank and Abelson, 1977), mental models (Johnson-Laird, 1980) or scenarios (Sanford and Garrod, 1981), is harnessed to an imaginary character’s gnostology. This would provide some explanation as to why scriptwriters and actors find it difficult to recreate convincing dialogue but also raises another question. Could it be that the great actor’s gift (or acquired skill) is that of constructing and reconstructing schemata to the point of making an APS seem real? The phasal analysis of a (multimodal) text provides ‘a connection with conceptual and gnostological analysis – a basis for the investigation of how speakers construct reality’ (Gregory, 2002: 342). And this applies equally to the unreal situations that we find in film. Even in film situations, the participants interact coherently in terms of all the semiotic modalities in play, and viewers understand films (usually) because they make an effort to ‘refer language to some understandable social Pedagogical Tools for Training Subtitlers 221 interaction’ (Fries, 2002: 347). This understanding/interpretation is assisted here by phasal analysis which throws up identifiable patterns of use. Just as in written texts, within the continuous and discontinuous stretches of discourse, items re-occur and co-occur to create recognisable circuits and chains of meaning, similarly, in multimodal texts, nonverbal elements will re-occur and co-occur in the same way, and in an integrated relationship with the verbal elements, forming clusters of semiotic modalities. The ‘Bundaberg Beer’ text Figure 16.2 shows three frames from a short Australian advertisement for Bundaberg ginger beer. The pictures in the frames best represent the three macro-phases identified within this multimodal text. In T Visual frame VI + kinesic action Soundtrack 1 Man walking purposefully through outback. Music/no speaking 2 Action with bottle. ‘You can’t walk away from the truth’ ‘Naturally brewed from our own family owned company’ ‘Non puoi fuggire dalla verità’ ‘Prodotta in modo naturale dalla nostra azienda di famiglia’ Attention focussed on product. TRUE BREW GINGER BEER GINGER BEER QUELLA VERA 3 Figure 16.2 Subtitle 222 Christopher Taylor macro-phase 1 a man can be seen walking through the outback on a hot day, passing in front of a mountain, wading through a stream, and climbing a rocky outcrop. When he gets to the top, macro-phase 2 commences as he spies a bottle of Bundaberg ginger beer nestling in a bed of ice; he takes the bottle, opens it and drinks it with enormous gusto. These actions are accompanied by a voice off describing the product in a broad Australian accent. Finally in macro-phase 3 the product is shown in the conventional style with the camera moving backwards and forwards over a picture of the product and the slogan: ‘True brew ginger beer’. The phases are identified as such in that various semiotic modalities can be seen to come together to form a unified whole. For example, in macro-phase 1 the single action of walking through the outback is accompanied by a particular piece of music which ends at the end of the phase. The transition then leads us into macro-phase 2 which is of course connected to 1 but recognised as a distinct unit – that of the drinking of the beer accompanied by a different musical item. There is also connection at an intertextual level. This advertisement forms part of a series, linked very clearly by a number of factors – the same characters, the same music, the same appeal to the outdoors, very similar wording, the same slogan, the same voice off and, of course, the same product. In Kress and van Leeuwen’s (1996) terms, the multimodal discourse of all the ads is that of convivial outdoor beer drinking in hot weather allied to the conventions of drinks advertising. The design is re-proposed throughout the series of advertisements which take the verbal and visual discourse of the original and create new, instantly recognisable multimodal texts using mostly the same resources. In terms of production, different discourses are used in the creation of genrelets of the subgenre ‘Bundaberg ads’ of the main genre ‘drinks ads’. For example, the man’s no-nonsense voice and pronounced Australian accent add texture to the discourse. These advertisements are then distributed on television to a large audience, originally to an audience anchored in the cultural context of Australia, and of beer drinking. Interestingly, the commercial promotes a non-alcoholic product but within the collective semiotic knowledge regarding serious beer drinking. Its discourse is not that of Coca Cola. Figure 16.3 shows the division in macro-phases and micro-phases of the Bundaberg beer advertisement, indicating how the verbal and nonverbal elements combine. Pedagogical Tools for Training Subtitlers 223 MACRO-PHASES: 1. Man walking through outback Musical accompaniment 2. Action with bottle 3. Attention on product Macro-phase 1: 1. Man walking away through grass 2. Side shot of legs (mountain behind) 3. Passes tree under red sky 4. Back view of man 5. Side shot of legs 6. Walking through stream 7. Close-up of feet 8. Emerging from stream Music changes 9. Walking through undergrowth 10. Climbing rock face 11. Reaches summit Macro-phase 2: 12. Man sees bottle 13. Reaches for bottle ‘You can’t escape from the truth’ 14. Opens bottle ‘Naturally brewed from our own family owned company’ 15. Drinks from bottle Macro-phase 3: 16. Close-up of label, panning back 17. Bottle shown and words TRUE BREW GINGER BEER Figure 16.3 Some suggestions for the subtitler – DDPDI As mentioned above, the discourse of the Bundaberg advertisements hanging in the air is that associated with beer drinking and television advertising, particularly as these relate to the original linguistic and cultural setting. The ads are archetypically Australian. This poses a 224 Christopher Taylor dilemma for the subtitler who has to tap into both an abstract discourse defined by local cultural mores, and translation norms for written subtitles. The dilemma hinges on the activating of a localisation process (adapting the text to a local audience’s expectations) as opposed to adopting the opposite strategy of foreignising the text and maintaining the original flavour. It is possible that ginger beer is best advertised in its Australian context, as this appeals to the foreign viewer more than a grafting of a locally connotated (written) text onto the original. On the other hand, as subtitling is an additive process, nothing is lost of the original including the voice and accent, and this may be enough to preserve the Anglo-Saxon context, even if the written text is localised. In fact, it is riskier to foreignise the wording – i.e. give it an Australian feel – in, say, Italian (the extreme case of this would be simply to leave the subtitles out altogether) as the discourse of the subtitles has also to adhere to the linguistic games being played. Thus the play on the key word ‘truth/true’ [verità/vera] needs to be kept; the use of monosyllabic generic terms, rhyme and word play are all up there in the abstract as integral parts of the discourse of advertising. While the discourse of Italian advertising can to a certain extent fit into this slot, the final slogan poses a question. The translation of ‘True brew ginger beer’ with Ginger beer quella vera is dictated by the need to translate (localise) the term ‘brew’ which, unlike ‘true’, is not at all familiar to a mass audience. Does this matter, if the sound and rhythm and rhyme, and particularly the picture, convey the message adequately? No answer will be given to this basically rhetorical question because its purpose is simply to illustrate the kind of thinking that needs to be activated in addressing the problem of localising or foreignising the discourse of the source text. The design of the visual and oral discourse in the Bundaberg texts remains, and in a subtitled version the only alteration to production is in the verbal text. However, the latter is transferred back from the spoken to the written medium, which is itself complicated by the fact that the original was ‘written to be spoken as if not written’. Complications notwithstanding, the actual wording of the titles constitutes a phase of production and a number of factors come into play, which are not found in any other type of text production or translation process. The writing of the subtitles is based firstly on a variegated set of considerations and strategies of a linguistic, semantic and pragmatic nature. A limited number of scholars and experts have written on the subject of subtitling strategies but the works of Ivarsson (1992), Ivarsson and Carroll (1998), Gottlieb (1992, 1997, 2000) and Gambier (1995, 1998) are Pedagogical Tools for Training Subtitlers 225 sufficient to show that a great many notions exist as to how to best perform this craft. Secondly, subtitle production is governed by a whole series of technical constraints ranging from positioning to timing and designing. The art of spotting (i.e. placing the subtitles in the right place at the right moment) is a crucial part of this production process. Turning to distribution, the question of localising/foreignising arises again. Subtitling a film involves projecting the discourse, design and production of the original product into a new culture. An example of the kind of issue that can arise is provided by one of the Bundaberg advertisements in which a young girl on the beach spills some ginger beer from its bottle onto a magazine, and then proceeds to drink from the folded mag. This causes no real concern in Anglo-Saxon cultures but in some other cultures this would be frowned upon, and showing the advertisement would not be conducive to marketing the product. Finally, in terms of interpretation, the words of subtitles represent a distinct (macro) genre and are not interpreted in the same way as written scripted words, words on the screen in the original language or the spoken words. Thus, where localisation is required, and this is often the case in advertising, subtitles can act to soothe sensibilities and avoid tensions. Their interpretation is, however, bound to other accompanying semiotic features, such as a raucous or a sensuous voice, or a musical theme. Lessons for the subtitler – phasal analysis (1) Phasal analysis allows us to see how (multimodal) texts are constructed, continuously and discontinuously, and thus where to look for patterns at a lexico-grammatical and semantic level. These patterns must be recognised and, if possible, respected. (2) Within the ‘artificially produced situation’, subtitlers can distinguish combinations of field, tenor and mode and attempt to relate them to the kind of situation they wish to create for the target audience. (3) Field: does the translator’s knowledge of the world and range of schemata equip him/her to effectively relay the meaning of the text, or is more care required in foreignising or localising the text? (4) Tenor: this is the most delicate area for the subtitler. Scholars like Kovač ič (1998) have shown that in subtitling there is a strong tendency to sacrifice the interpersonal element, those items that particularly distinguish spoken language, as exemplified in the description of the experiment conducted in Trieste (see above). It is at least advisable to maintain a sufficient number of these features. 226 Christopher Taylor Kovač ič herself provides the example of the mother figure in the televised rendering of Eugene O’Neill’s Long Day’s Journey into Night. The mother, in the midst of feverish discussions of a subversive political nature between earnest young men at table, continues to repeat platitudes about her hair, of no consequence to the unravelling of the plot. However, the character of the mother is important as a kind of counterpoint to the serious revolutionary action, and is highlighted by her inane repetitions. Similarly, the stuttering of the character Ken, played by Michael Palin, in the film A Fish Called Wanda, while not adding to the storyline, provides an important angle on the character and contributes to the humour of the text, thereby deserving of the translator’s attention through Gottlieb’s strategy of ‘imitation’ (1992: 166), that is, the phonetic transcribing of the stutter. The dynamic force of repetition, hesitation, discourse markers, and hedging devices should not be underestimated. (5) Mode: in a counter-tendency to the omission of interpersonal elements, the spoken-to-written transformation often results in overexplicitness in the information load. Baker (1995) conducted research on written texts proving, through a comparison of parallel texts, that translated material tended to be more explicit than equivalent texts written in the original language. For example the ‘that’ relative, which is often expendable, is found significantly more often in translated texts. Abbreviations of the ‘don’t’, ‘wouldn’t’ kind are significantly more frequent in original texts. Thus the presentation of information in translations has proved to be slightly different over a large amount of material. The question of whether the same occurs in subtitles is inevitably connected to the ‘written to be spoken as if not written’ conundrum. In theory the writer of the text to be translated will have made efforts to make that text seem natural or original, even though we have seen that this objective is difficult to achieve. In any case, subtitlers must be aware of this and also consider that their subtitles are, in a sense, ‘written to be read as if not written’. (6) Subtitlers must make every effort to be faithful to the original text in terms of ‘experience relationships’ or the ideational element, ‘interaction relationships’ or the interpersonal element, and the medium relationships, and thus recognise where the ideational, interpersonal and textual features show consistency and congruity. (7) The experience, interaction and medium relationships are also present in the non-linguistic modalities (visual, musical, gestural) Pedagogical Tools for Training Subtitlers 227 and the consistency and congruity must be identified as an integrated whole. Phasal analysis, as adapted by Thibault and Baldry (2000), performs this task. (8) Subtitlers should work from the transcribed text of the original film, not the original script. The way actors react to artificially produced situations often results in dialogue that moves closer to real language use. As Zaber (1980: 8) states: ‘The kind and degree of training and the sophistication an actor brings to his art are further filters on the script performed’. These suggestions, provided as debating points rather than prescriptive rules, mark the end of an initial phase in theoretical research into screen translation didactics. Much hard work of a more practical nature is now required to make sure these ideas are converted into useful tools for the screen translator. References Baker, M. (1995) ‘Corpora in Translation Studies: an overview and some suggestions for future research’. Target 7: 223–43. Baldry, A. (ed.) (2000) Multimodality and Multimediality in the Distance Learning Age. Campobasso: Palladino Editore. Fries, P. (2002) ‘Some aspects of coherence in a conversation’. In P. Fries, M. Cummings, D. Lockwood and W. Sprueill (eds). Relations and Functions within and around Language (pp. 346–88). London, Continuum. Gambier, Y. (ed.) (1995) Audiovisual Communication and Language Transfer. Translatio, Nouvelles de la FIT – FIT Newsletter XIV(3–4). Gambier, Y. (ed.) (1998) Translating for the Media. Turku: University of Turku. Gregory, M. (2002) ‘Phasal analysis within communication linguistics: two contrastive discourses’. In P. Fries, M. Cummings, D. Lockwood and W. Sprueill (eds) Relations and Functions within and around Language (pp. 316–45). London, Continuum. Gregory, M. and Carroll, M. (1978) Language and Situation: Language Varieties and Their Social Context. London: Routledge and Kegan Paul. Gottlieb, H. (1992) ‘Subtitling – a new university discipline’. In C. Dollerup and A. Loddegaard (eds) Teaching Translation and Interpreting: Training, Talent and Experience (pp. 161–70). Amsterdam and Philadelphia: John Benjamins, Gottlieb, H. (1997) Subtitles, Translation and Idioms. Copenhagen: University of Copenhagen. Gottlieb, H. (2000) Screen Translation: Six Studies in Subtitling, Dubbing and Voice-over. Copenhagen: University of Copenhagen. Halliday, M.A.K. (1994) An Introduction to Functional Grammar. London: Edward Arnold. Ivarsson, J. (1992) Subtitling for the Media. A Handbook of an Art. Stockholm: TransEdit. 228 Christopher Taylor Ivarsson, J. and Carroll, M. (1998) Subtitling. Simrishamn: TransEdit. Johnson-Laird, P.N. (1980) ‘Mental models in cognitive science’. Cognitive Science 4(1): 71–15. Kovač ič, I. (1998) ‘Six subtitling texts’. In L. Bowekr, M. Cronin, D. Kenny and J. Pearson. (eds) Unity in Diversity? Current Trends in Translation Studies (pp. 75–82). Manchester: St. Jerome. Kress, G. and van Leeuwen, T. (1996) Reading Images: the Grammar of Visual Design. London: Routledge. Minsky, M. (1975) ‘A framework for representing knowledge’. In P. Winston (ed.) The Psychology of Computer Vision (pp. 211–77). New York: McGraw Hill. Sanford, A.J. and Garrod, S.C. (1981) Understanding Written Language. New York: Wiley. Schank, R.C. and Abelson, R.P. (1977) Scripts, Plans, Goals and Understanding. Hillsdale, N.J.: Erlbaum. Thibault, P. and A. Baldry (2000) ‘The multimodal transcription of a television advertisement: theory and practice’. In A. Baldry (ed.) Multimodality and Multimediality in the Distance Learning Age (pp. 311–85). Campobasso: Palladino Editore. Zaber, O. (1980) The Languages of Theatre: Problems in the Translation and Transposition of Drama. Oxford: Pergamon Press. 17 Teaching Subtitling in a Virtual Environment1 Francesca Bartrina Introduction The crucial role played by audiovisual translation (AVT) in contemporary international communication invites translator trainers to contemplate the different possibilities available when training translators for the modern mass-communication market. The acquisition of new skills in the use of Information and Communication Technology (ICT) is a challenge that instructors of subtitling and trainees must face up to. In online multimedia courses, the use of digital technology is an imperative, but professional computer programs are expensive and educational licences are often unavailable. In this situation, how can students get the hands-on training they need? In response to this situation, some academic institutions try to generate their own in-house solutions. In the Catalan/Spanish context, two universities have answered these challenges by creating their own programs. The Universitat Autònoma of Barcelona (UAB) has developed the programs Subtitul@m for subtitling and REVOice for dubbing and voice-over: ‘Both are pioneering programs that give students the opportunity to simulate real working conditions and become familiar with the various procedures involved in the different AVT modes’ (Díaz Cintas and Orero, 2003: 373). In 2001 Richard Samson in the Translation School of the Universitat de Vic, in Barcelona, created the Poor Technology Group (PTG) ‘for people who work in the field of education and who seek innovative low-budget digital solutions to vocational (professional training) educational problems’ (www.uvic.es/fchtd/especial/en/ptg/ ptg.html). Ever since, the group and its members have been developing solutions to a variety of training issues, including dubbing and voiceover. They faced the demands of online subtitling instructors who 229 230 Francesca Bartrina considered the subtitling practioners as a ‘highly specialised professional sector that uses exceedingly expensive software’. Samson has created the Digital Video Subtitling Compilation (DVSC) that can be either distributed to students on CD-ROM format or downloaded as a demo. The programs included in this project are free and available on the following websites: ● ● ● Subtitle Workshop (Subtitle editor): www.urusoft.net/downloads.php Direct X (Windows multimedia library files): www.microsoft.com/ windows/directx Digital Video Subtitling Compilation: www.uvic.es/fchtd/especial/ en/ptg/ptg.html When teaching subtitling online, it is essential to work with digital subtitling programs because they emulate the working conditions of professional subtitlers in the contemporary market. In the real world, subtitlers are exposed to different subtitling software packages in each of the companies they work for. We need to train our students to be able to use as many subtitling programs as are available on the market.2 In what follows, I discuss some aspects of teaching subtitling in a virtual environment. Quality in teaching interlingual subtitling: the Catalan and Spanish contexts Subtitling is a way of translating what is being said in an audiovisual text, with two characteristic features. First, there is a change of medium, from the oral to the written form. Second, the oral message of the source audiovisual text is also present in the translated product. The specificity of subtitling can be dealt with in an online campus taking into account the following requisites. Spain is a dubbing country and the high quality of some Spanish and Catalan dubbed programmes has been stressed. In his book, Ávila (1997: 37) praises the 1940s to the early 1970s as the Golden Age of Spanish dubbing and deplores the declining levels of quality nowadays. The Catalan Channel TV3 (1997: 209–10), in its Criteris lingüístics sobre traducció i doblatge, pays tribute to the translators of the best dubbed films in the first 12 years of broadcasting of the channel. Unfortunately, nothing similar can be said in the case of subtitling. Until recently, the poor quality of Spanish and Catalan subtitling has led us to pose the question: do Spanish audiences prefer dubbing because of the low Teaching Subtitling in a Virtual Environment 231 quality of subtitling or is the low quality of subtitling due to its marginal status? (Díaz Cintas, 2003: 69). The popularisation of DVD has increased the availability and the amount of Spanish and Catalan subtitling. This increase does not necessarily mean a concentration of work in the Spanish industry, because due to the phenomenon of globalisation not all Spanish and Catalan subtitling is done in Spain. However, we have to prepare our students for the global market in order to jump from online campuses to the virtual professional environment. We have to train our students to be competitive, which means that they have to be able to offer their services as subtitlers to international subtitling companies located around the world. The use of digital subtitling software in online campuses is the best way to prepare for it. At the same time, the dialogue between universities and subtitling companies must be constant and fluid. 3 The audience In subtitling, the oral and written messages are received simultaneously, allowing for comparison between Source Text (ST) and Target Text (TT). In his seminal work, Ivarsson (1992: 49) states that: ‘It is, moreover, unnecessarily irritating to those who have some knowledge of the language when a person says something and it is not translated. The natural reaction is: Why ever did they skip that?’. Such remarks made by the audience make this kind of translation particularly ‘vulnerable’, as Díaz Cintas has emphasised in several of his works (2001, 2003). Viewers who are familiar with the Source Language of an audiovisual text tend to notice some of the discrepancies between the target subtitles and the source dialogue and often criticise subtitlers for their inaccuracy. When subtitling, professionals are extremely conscious of what their audiences expect to be translated. They frequently choose the translation solutions that are most similar to the ST. As trainers, two ways are open for us here. The first one is to consider that, following audience expectations, similarity to the ST is a quality criterion in subtitling. We can encourage students to follow the order of the verbal elements of the ST and to go for the solution which is more similar to the original. The second one is to encourage them to choose the most fluent solution, in terms of accessibility, as we will see below, encouraging an experience for the receivers of subtitles that will hopefully change this prejudice. Probably, the best solution is to find a balance between the two positions. 232 Francesca Bartrina Unquestionably, subtitling directly affects the audience’s perception of the audiovisual programme. Gambier (2001: 102–3) has stressed the fact that audience perceptions of subtitling depend on sociological and audiovisual variables such as age, reading habits, knowledge of the Source Language (SL) and Target Language (TL), genre of the audiovisual programme, channel, broadcasting time and rhythm of the audiovisual product, among others. We could argue that the impact of DVD has challenged the way in which viewers perceive quality and subtitling. Audiences’ empowerment has been increased with the possibility of comparing dubbed and subtitled translations of the same audiovisual programme. This situation has already had consequences in the profession with clients demanding that dubbed and subtitled versions of the same film be similar. It seems that for DVD subtitling similarity to the dubbed version and literal solutions are demanded. Another issue to be considered is the channels at the audience’s disposal to express their views about subtitling. In our field, there is a strong demand for audience studies in order to know what the viewers really think about AVT. However, we should remember that most audience studies carried out in Spain are done to attract publicity agents. Mayoral (2001: 44–5) discusses the constraints and limitations that scholars in Translation Studies have to face when embarking on research of this nature. The question arises as to what kind of subtitling quality awareness can be developed in continuous online assessment, and we need to find a balance between homogeneity and diversity. Students need to be exposed to a range of different standards that are operative today in the cinema, TV, video and DVD subtitling markets, but they also need training in specific standards in order to be prepared for academic grading. We therefore have to study various subtitling practices implemented in the market, without forgetting to use a unifying code, a unique set of subtitling standards as proposed by Ivarsson and Carroll (1998) or Karamitroglou (1998). Gambier (2001, 2003) has written extensively on the expectation of quality in screen translation. For him, the key word for quality in AVT is ‘accessibility’, a concept including many features that can easily be adapted to the context of subtitling: acceptability (grammar, style, terminology), legibility (position, subtitle rates), readability (reading speed rates, shot changes), synchronicity (what is read is what is shown), relevance (what information is to be deleted or added) and translation strategies when dealing with cultural items. It can be concluded that Teaching Subtitling in a Virtual Environment 233 these features are the criteria for quality in subtitling, and they can be considered the basis for fluency in this kind of translation. In my opinion, the concept of accessibility is extremely useful when teaching subtitling. It is an eclectic term in the sense that it tries to cover what has been said in general definitions of good subtitling practice, such as that of James (1998: 245): ‘to produce subtitles which are accurate, credible, easy to assimilate and which flow smoothly’. And it also includes specific demands on subtitling such as: ‘written subtitles should be made to ‘sound’ like their spoken equivalents’ (Brondeel, 1994: 29). Students should be told at the beginning of the course that this is the quality standard upon which their work will be assessed. Teaching subtitling online When teaching online, a great deal of activity on the virtual platform becomes crucial. Trainers have to use all the online resources at their disposal to keep students motivated. When teaching subtitling, a good method can be the design of a set of activities aimed at familiarising students with the whole process and the spatial, temporal and textual parameters. These activities must be an integral part of a curriculum that is learning-centred and that involves high-level cognitive processing.4 These activities can be found within the virtual platform, together with the audiovisual material. Sometimes it may be more convenient to send a CD to the students with the audiovisual material, because of its large volume. Some activities should be aimed at encouraging students to find their own audiovisual material and to locate scripts using different Internet resources. In distance learning, the relationship between student and teacher has to be reinforced with specific tools like forums, chats and frequent interaction via e-mail.5 A timetable for students’ work should be made explicit at the beginning of the course. If students carry out activities frequently (every one or two weeks) and trainers answer promptly with their remarks, the motivation of the whole virtual group is bound to increase. It is also very important to work with professional deadlines: subtitlers need to work fast and accurately. The literature on the subject stresses the importance of training translators to be able to work on the whole process: timing or spotting, translating, adapting, and editing. If the same professional carries out the whole process, the quality of the subtitling is more likely to be guaranteed. Using a digital subtitling program, students can work simulating a 234 Francesca Bartrina professional context. This experience will boost their employability because they will be the sort of professional the market needs. Viewing Students working with a dialogue list need to check the written script against the text that is actually spoken in the programme. However, Gottlieb (1996: 284) has argued for the need to teach subtitling without scripts, mainly because: Without scripts as translatory crutches students remain focused on the fact that subtitling is not just a matter of translating some lines from a script and shaping them into neat blocks. Subtitling is a craft in which one recreates the foreign dialogue in one’s own (written) language, as an integral part of the original film whose visual content co-interprets the meaning of the lines as they are spoken. Subtitling without scripts can help students appreciate the specificity of the audiovisual mode and reinforce the fact that the spoken text and the image are indissoluble. They are encouraged first to understand and secondly to translate and do the timing simultaneously. Subtitling without a script can make timing easier. However, in the professional world, when translators are given a good postproduction script with an adequate glossary of difficult terms the quality of the audiovisual translation tends to improve considerably. Also, the use of scripts is more convenient and less time-consuming for teaching purposes. When viewing the audiovisual text, students may be guided in different activities so that they are aware of all the non-spoken information that should be translated: signs, notices, etc. They should decide whether songs need to be translated or not, and they should also have to make decisions on questions of politeness and the way the characters address each other, formally or informally (for example, in Spanish Tú/Usted/ Ustedes and Catalan Tu/Vostè/Vostès). Timing, spotting or cueing Digital subtitling programs allow students to work with an alreadymade timing and to practise with strict constraints. In addition, they also allow students to create the timing themselves and to learn the technique. In the market, subtitlers have to learn to do both. If they work for the DVD industry, they often receive a document with an already-made timing (known in the profession as ‘master list’ or ‘genesis file’), but students also have to be prepared to do the spotting themselves Teaching Subtitling in a Virtual Environment 235 and a digital subtitling program allows them to practise this skill. Activities should also be designed to allow students to change, modify and improve an already-made timing. The aim of these activities is to stimulate students to practise spotting, taking into consideration the start and finish of characters’ utterances, the semantic units of the dialogue, the rhythm of speech, the changes in camera shots and sound bridges. Translating Timing constraints determine the priorities when subtitling. Students have to translate what is being said on the screen, taking into account the information given by the iconic dimension of the image, and taking advantage of the tools given by the virtual context: online dictionaries, encyclopaedias and databases, audiovisual translators’ chats and forums, associations’ websites, electronic journals and a multitude of suggestions in, for example, terminology management systems and others. As mentioned by Koby and Baer (2003: 216), when teaching the use of such tools, our main pedagogical goal must be the acquisition of conceptual knowledge: In a computerized translation classroom or lab, the overarching goal is to enable students to acquire both declarative knowledge (i.e. specific technical skills in an application) and procedural knowledge (i.e., the ability to recognize and contextualize technical issues within a conceptual framework) to enable them to work effectively in an increasingly technologized translation industry, and to deal with change proactively through their understanding of the nature and underlying principles of the technology. Editing This is an important stage of the quality control. Digital subtitling programs make it easy for students to check their work – conventions of written language and technical corrections, if needed – improving their impact on the final product since they are able to control the whole subtitling process. Students should be encouraged to edit both their own work and the work of other colleagues. Mayoral (1999: 3) reminds us that: ‘Students must be trained for teamwork, sharing translation tasks not only with other translators but also with professionals in other fields (actors, producers, multimedia technicians, editors, etc.)’. Teamwork is a highly appreciated skill in the professional world. The 236 Francesca Bartrina teacher may work here as a facilitator, distributing students’ versions and contributing to group debates. Priorities in subtitling In the context of an online campus, we should prepare and develop activities aimed at practising the following skills. Synthesis, redundancy and synchrony Synthetic reductions are inevitable in subtitling. They can be partial (condensation or concision) or total (elimination or deletion). Synthesis is related to redundancy, a very important concept in subtitling, because the information not given by the subtitles may be supplied by other elements present in the audiovisual text: the image and/or the sound. Gottlieb (1994: 273) distinguishes between: (1) Intersemiotic redundancy, which enables the viewer to supplement the semiotic content of the subtitles with information from other audiovisual channels – notably the image, and prosodic features in the dialogue. (2) Intrasemiotic redundancy, in the dialog. Especially with spontaneous speech, not only the informative content, but also the verbal style and characterization of the speaker are better served with some reductions in the subtitles. Gottlieb’s proposed distinction is very important as it helps students to be aware of how synthesis works in subtitling and some activities should be prepared in order to practise this skill. It is important to insist upon all decisions being taken in relation to synchrony. Ivarsson (1992: 46–51) establishes two main types of synchrony: between image and subtitle, and between sound and subtitle content. In his own words (ibid.: 49): ‘When items are enumerated they should be written in the same order as in the original so as to confuse the viewers as little as possible. (Strangely enough, viewers are much more irritated by a change in the order than by the omission of one or two items)’. Spatial, temporal and linguistic parameters It is important to ask students to follow different typographical guidelines to emulate the reality of the professional world. Differences are visible in programmes commercialised by the very same distributor, company or TV channel. Spatial and temporal parameters given to Teaching Subtitling in a Virtual Environment 237 students may follow the international standards of subtitling (Ivarsson and Carroll, 1998; Karamitroglou, 1998) and the specific guidelines applied by professional subtitling companies. Students should be familiar with different parameters but we have to make clear to them which parameters must be used in each activity. From online campuses to the virtual professional environment We may have different answers to the question ‘What is quality in subtitling?’ depending on the group of people we want to address: subtitling companies, producers, distributors, broadcasters or viewers. In this context, how can we train translators to be good professionals? What sort of skills do they need to have? In my opinion, the following competences are required: ● ● ● ● ● ● ● ● Excellent command of the two relevant languages and cultures at work. Knowledge of the semiotics of the audiovisual text.6 Ability to summarise. Documentation skills. Capacity for self-evaluation. Ability to work fast and accurately.7 Skills to negotiate with other professionals. Capacity to co-ordinate a project. It is unquestionable that the use of digital software can help to develop the following skills: ● ● ● ● Quick adaptation to changing subtitling software. Adaptability to real world constraints and priorities. Control of all the stages in the subtitling process. Professional documentation skills for online resources. Conclusion As trainers, we have a duty to make explicit which are the quality criteria we follow when assessing and marking students’ work. By working in a virtual environment with the help of digital subtitling programs, we are empowering our students and giving them the tools they need to improve the quality of the subtitled product. In the professional 238 Francesca Bartrina world, clients have the power to impose their own quality criteria, even though subtitling companies may have their own in-house guidelines. Following Gambier (2003), we also advocate accessibility as the main quality parameter of subtitling, a concept that may include the demands of clients, subtitling companies, viewers, scholars and subtitlers. Notes 1. The author wishes to express her thanks for the research grant awarded by the Generalitat de Catalunya, 2003MQD 00062. 2. Mayoral (1999: 3) considers that: ‘We should also be training students in ‘teletranslation’ (translation at a distance), since the professional market has been globalised to the extent that a translator can work for a client in any part of the word. This requires competence in the use of the Internet and ability to work in teams across distance’. 3. I agree with Díaz Cintas and Orero (2003: 373) when they state that: ‘Universities are always behind as far as technological development is concerned, and maintaining contact with the industry can help to keep lecturers updated and informed. Conversely, subtitling companies can benefit from links with university programmes because they can have a greater say in the design of the academic curriculum and offer advice on skills they most need when recruiting new staff’. 4. Some guided subtitling activities can be found in Bartrina and Espasa (2003: 30–8). 5. The importance of interaction in e-learning and the experience of running an Open Distance Learning MA in Translation Studies has been analysed by Millán-Varela (2001). 6. Future subtitlers should be familiar with Film Studies. At the University of Vic, students of subtitling can follow courses in the Escola de Cinema i Audiovisuals de Catalunya thanks to a specific agreement between the two institutions. Our students also have the opportunity to subtitle into English the audiovisual programmes made by final-year students of cinema for international festivals. 7. We should encourage students to try and improve the working conditions of the professional market: educating the clients, belonging to Associations of Translators. But, at the same time, students have to learn to work under adverse conditions. References Ávila, A. (1997) El doblaje. Madrid: Cátedra. Bartrina, F. and Espasa, E. (2003) ‘Traducción de textos audiovisuales’. In M. González Davies (ed.) Secuencias. Tareas para el aprendizaje interactivo de la traducción especializada (pp. 19–38). Barcelona: Octaedro. Brondeel, H. (1994) ‘Teaching subtitling routines’. Meta 39(1): 26–33. www. erudit.org/revue/meta/1994/v39/n1/index.html Teaching Subtitling in a Virtual Environment 239 Díaz Cintas, J. (2001) La traducción audiovisual. El subtitulado. Salamanca: Almar. Díaz Cintas, J. (2003) Teoría y práctica de la subtitulación: inglés–español. Barcelona: Ariel. Díaz Cintas, J. and Orero, P. (2003) ‘Course profile: postgraduate courses in Audiovisual Translation’. The Translator 9(2): 371–88. Gambier, Y. (2001) ‘Les traducteurs face aux écrans: une élite d’experts’. In F. Chaume and R. Agost (eds) La traducción en los medios audiovisuales (pp. 91–114). Castelló: Universitat Jaume I. Gambier, Y. (2003) ‘Introduction. Screen transadaptation: perception and reception’. The Translator 9(2): 173–89. Gottlieb, H. (1994) ‘Subtitling: people translating people’. In C. Dollerup and A. Lindegaard (eds) Teaching Translation and Interpreting 2 (pp. 261–74). Amsterdam and Philadelphia: John Benjamins. Gottlieb, H. (1996) ‘Theory into practice: designing a symbiotic course in subtitling’. In C. Heiss and R.M. Bollettieri Bosinelli (eds) Traduzione multimediale per il cinema, la televisione e la scena (pp. 281–95). Bologna: CLUEB. Ivarsson, J. (1992) Subtitling for the Media: A Handbook of an Art. Stockholm: TransEdit. Ivarsson, J. and Carroll, M. (1998) ‘Code of good subtitling practice’. Language Today, April: 1–3. Also found at: www.transedit.st/code.htm James, H. (1998) ‘Screen translation training and European cooperation’. In Y. Gambier (ed.) Translating for the Media (pp. 243–58). Turku: University of Turku. Karamitroglou, F. (1998) ‘A proposed set of subtitling standards in Europe’. Translation Journal 2(2). www.accurapid.com/journal/04stndrd.htm Koby, G.S. and Baer, B.J. (2003) ‘Task-based instruction and the new technology’. In B.J. Baer and G.S. Koby (eds) Beyond the Ivory Tower. Rethinking Translation Pedagogy. ATA Scholarly Monograph Series, Vol. XII (pp. 211–29). Amsterdam and Philadelphia: John Benjamins. Mayoral, R. (1999) ‘Notes on translator training’. Intercultural Studies Group Innovation in Translator and Interpreter Training. www.fut.es/~apym/symp/ mayoral.html Mayoral, R. (2001) ‘El espectador y la traducción audiovisual’. In F. Chaume and R. Agost (eds) La traducción en los medios audiovisuales (pp. 33–46). Castelló: Universitat Jaume I. Millán-Varela, C. (2001) ‘Building bridges: an open distance learning MA Translation Studies’. In S. Cunico (ed.) Training Translators and Interpreters in the New Millenium (sic) (pp. 124–35). Portsmouth: University of Portsmouth. TV3, Televisió de Catalunya (1997) Criteris lingüístics sobre traducció i doblatge. Barcelona: Edicions 62. 18 Subtitling: Language Learners’ Needs vs. Audiovisual Market Needs Annamaria Caimi Introduction In the context of globalisation, the question of audiovisual translation (AVT), generally taken to mean dubbing and/or subtitling, is fundamental for linguistic education and responds to new multicultural needs with reference to Europe and, more extensively, to the international world. The amazingly powerful role played by TV, cinema, video and DVD viewing in the twentieth century has brought about an upsurge of academic interest in multimedia translation, which is moving into the twenty-first century with renewed vigour. This chapter focuses on the important role that subtitled films and TV programmes have in the field of language learning, with a view to bridging the gap between producers’ budgetary constraints and linguists’ demand for quality subtitling, in order to serve the needs both of the ever growing population of second language learners and of film marketers. As far as research on subtitling is concerned, the range of publications is impressive. We are particularly indebted to Gottlieb, author of an international annotated bibliography on interlingual subtitling for cinema, TV, video and DVD, covering seven decades, from the invention of film subtitling in 1929 to 1999. The bibliography is published in the journal RILA (Caimi, 2002: 215–397), and Gottlieb (2002: 222) writes in his introduction that it is a ‘reader-friendly guide to the growing literature on interlingual subtitling’. 240 Subtitling: Language Learners and the AV Market 241 The linguistic perspective: cognition and simplification in subtitling Audiovisual translation researchers know that, linguistically, subtitles must be concise and easy to understand and the text must be carefully edited and split: the spoken word directly transcribed usually exceeds the space and time available, since subtitles should not stay on the screen for longer than six seconds. Yet, they should convey the relevant information and linguistic nuances contained in the original dialogue to satisfy second language audience needs. Viewing subtitled films is a cross-linguistic communication experience, where the message is simultaneously conveyed by the two most common channels of communication: speech and writing. Consequently, viewers have to practise reading and listening skills simultaneously, backed up by the visual, animated input of the storyline of the film. In an entertaining audiovisual situation, the measurement of listening and reading comprehension abilities may be considered from different perspectives, as it is the result of integrated multifunctional activities. The linguistic perspective I have chosen to highlight in this chapter is restricted to the strategy of simplification, almost regularly applied in subtitling, due to space and time restrictions. Subtitlers tend to encode important information to the detriment of details which may deprive the message of relevant, though not fundamental, clues. There are good reasons for resorting to simplification: it facilitates the transfer from one language to another and from the spoken to the written media. It also speeds up the viewers’ reading activity. When resorting to simplification and reduction of the Source Text (ST), translators in general must be aware of the psycholinguistic and sociolinguistic issues which are likely to influence the target viewers. This awareness should guide the linguistic choices which must be in keeping with the socio-cultural background of the target audience (Pavesi, 2002: 128; Laviosa, 1998; Baker, 1993 and 1996; Blum-Kulka and Levenston, 1983). Simplification influences all translation processes and is of paramount importance in subtitling, where the transfer from Source Language (SL) to Target Language (TL) is subsidiary to the dialogue of the soundtrack. Nowadays, subtitling has achieved general recognition as a type of AVT. Decoding and recoding subtitles involves cognitive processes that interact and interfere in the multifunctional action of translators – who transfer the film dialogue into subtitles – as well as in the combined effort of the film viewers – who have to read, understand and interpret them. Subtitling, which is almost always seen as a language mediation service, is increasingly being perceived as an attractive 242 Annamaria Caimi educational option (Díaz Cintas and Fernández Cruz, 2008; Pavesi and Perego, 2008) in the field of language learning and teaching methodology. In what follows, I analyse some examples of written dialogue and their corresponding subtitles from a cognitive perspective. My aim is to add some conceptual notions to the verbal changes brought about by processes of simplification, highlighted by cognitive insights. Setting the scene Let us first set the scene by sketching the plot of the film Billy Elliot, in order to frame the cues of the dialogue in its geographical, historical and socio-cultural context. The basic storyline is about an 11-year-old boy in a miner’s family who discovers that he prefers to join the girls for ballet classes rather than attending boxing classes. When his father finds out that the money he regularly saves to send Billy to boxing lessons has been spent by his son to learn ballet, he is furious. Billy’s father and Tony, Billy’s older brother, do not think ballet is a proper activity for a boy, and are prepared to back up their opinion with violence. The small family (father, two sons and an old grandmother) lives in Everington, a depressed coal town with ugly brick houses and giant slag heaps, which conveys the poverty and monotony of the bleak northern British setting. Billy’s father and older brother, both miners, are on the picket lines during the huge miners’ strike, which in the early 1980s upset the coal-mining region of Northern England. They speak Geordie, a dialect that is typical of people from Tyneside. The expressions they use are often colourful, and all the examples below are used by male characters. In Example 1, Billy is saying goodbye to his gay friend Michael. Billy’s father is watching them with disapproval and Tony, Billy’s brother, reacts by saying to him: Example 1 Tony: Stop being an old fucking woman! Italian La pianti di comportarti come una suocera? Back translation (Can you) stop behaving like a mother-in-law? Tony is a miner like his father and construes his message according to his sociolinguistic parameters. He does not speak standard English. He speaks the dialect of the coal-mining region of Northern England, where he lives and like, most young adults, his language is unrestrained and aggressive. Subtitling: Language Learners and the AV Market 243 The subtitler chooses to transfer this cue by replacing the vulgar expression ‘old fucking woman’ with the term suocera, which preserves the negative connotation, because the cliché of the mother-in-law, seen as an annoying person, is well known to Italian speakers. This means that the perspective of the subtitle, which is the perspective of the translator, foregrounds the negative connotation of the message, but removes the vulgarity of the English expression. Two different languages are concerned and the devices that involve conceptualising one domain of experience in terms of another depend on the knowledge of the two different cultures. In this example, where Tony speaks a type of dialectal English, the Italian translator could have resorted to a similar regional or working-class dialect, but he prefers to avoid dialectal terms and instead simplifies the expression by using core vocabulary. The language of miners or factory workers is usually rich in metaphorical expressions. The translator identifies the source domain suocera [mother-in-law] because it is in keeping with the target domain conveying the Italian concept of suocera as ‘a pain in the neck’. If we highlight the basic cognitive notions which motivated Tony’s message and assess to what extent the source conceptualisation may be transferred into the TL conceptualisation of the same metaphorical pattern, we understand that asymmetry of interpretation in the two languages is motivated by the fact that translation from L2 to L1 (subtitlers usually translate into their mother tongue) is accomplished directly on the basis of transferring the cultural conceptualisation of the situation. The crucial point is to find lexical links between words which convey the same conceptualisation of the situation in the two languages. In this subtitle, the translator decides that an expression in Italian which is compositionally parallel to one in English cannot be found. He delves into Italian metaphorical devices related to women’s most salient defects and chooses the one which, according to him, is conceptually near enough to the phrase pronounced by Tony, although compositionally incomplete. The aim is to give all types of viewers the possibility of quick understanding in order not to make them lose the thread of the story. Example 2 George: I haven’t seen hide nor hair of Billy for months. Italian Back translation Neanche l’ombra del tuo Billy da mesi. I haven’t even seen the shadow of Billy for months. 244 Annamaria Caimi George, Billy’s boxing trainer, tells Billy’s father that the boy is no longer attending his lessons. Perspective and foregrounding connect linguistic coding closely to visual perception in the English expression, which is rendered into Italian with a corresponding idiomatic form. The verb form ‘I haven’t seen’ has been eliminated because it does not affect the meaning of the cue. Example 3 George: He’s like a fanny in a fit. Italian Back translation Si agita come una femminuccia. He flaps about like a right wet. The original dialogue and the translation, both give prominence to the feminine qualities attributed to Billy in a derogative way, but the vulgar sexual reference of the English is again absent from the Italian version. The aim is always to facilitate comprehension by using everyday language. The learner is interested in what is meant, not in Geordie dialect expressions, and this linguistic choice guarantees comprehension. In the following example, George offers himself to help Billy’s father to persuade Billy to give up ballet: Example 4 George: Send him round to my house. I’ll knock sense into him. Italian Back translation Digli di passare da me. Lo rimetto subito in riga Send him round to my place. I’ll sort him out straight away. We can observe parallel interpretations in the two versions but the English foregrounds the language of boxing appropriate to George’s idiolect, whereas the Italian subtitle, though sharing the same conceptual perspective, is rendered into standard, idiomatic Italian. In Example 5, Billy’s father complains about the sacrifice he is making to save money to send Billy to boxing lessons. Interpretation, perspective and foregrounding are crucial to an understanding of colloquial expressions both in English and in Italian. The difference between the two versions rests with the vulgar expression used by the miner, which is chastened by the subtitler who has decided to use mi faccio il mazzo Subtitling: Language Learners and the AV Market 245 Example 5 Billy’s father: I’m busting my ass for those 50 pence. Italian Back translation Io mi faccio il mazzo per quei 50 pence. I’m working my guts out for those 50p. [I’m working my guts out] instead of the more vulgar mi faccio un culo così [I’m busting my ass]. The tendency to avoid transferring vulgar expressions is typically aimed at conveying everyday colloquial language rather than ‘colourful’ idioms. I conclude this sketchy analysis of male dialogues with a cue of Billy’s brother, addressing Mrs Wilkinson, the ballet instructor: Example 6 Tony: What are you going to do? Make him a scab for the rest of his life? Italian Back translation Sta cercando di farne un crumiro? Are you trying to make him a black-leg? Although the verbose original is fitted into a one-liner, the core concept is preserved. Here the subtitler has implemented the strategy of simplification by deleting, in hierarchical order, the redundancies connected with the young miner’s rough way of speaking. He reduces the cue by crossing out the reference to ‘for the rest of his life’, which is elicited by Tony just to reinforce the core concept of the message. The term ‘scab’ is an informal, disapproving way of referring to a strike-breaker, who, according to Tony’s social knowledge of the world, can be associated with a person who follows his artistic aptitude, like his brother. In both cases a person to be ashamed of. The dialogue exchanges we have examined above stress the harshness of Billy’s environment based on miners’ macho culture and his family’s resistance to dance, as opposed to the redemption of masculinity through the desire to succeed in an artistic job, despite non-conformity. The examples show that simplification is a complex strategy entailing a hierarchy of choices, such as reduction, condensation and decimation, on the basis of psychological, socio-cultural and linguistic priorities. These help the translator in singling out the chunks of text which must be kept and those which can be eliminated or condensed.1 246 Annamaria Caimi The teaching/learning perspective: subtitling in second language acquisition According to Krashen (1981), language acquisition consists of the spontaneous process of rule internalisation that results from natural language use, while language learning is the process of developing conscious L2 knowledge through formal study. I belive Krashen’s conceptual distinction is appropriate because, in second language learning, subtitled films, TV films and TV programmes occupy an intermediate position between spontaneous exposure in natural settings and formal instruction. Following Krashen’s double perspective, viewers/learners may be divided into two types: informal learners (Ellis, 1985: 288), who are more concerned with communicating, and formal learners, who are more interested in acquiring grammatical rules to reach proficient communicative standards. The first type is widespread among speakers of lesser-used languages who tend to live in subtitling countries such as Greece, Cyprus and Portugal in southern Europe, and Wales, the Netherlands, Denmark, Finland, Norway and Sweden in the north. Most of them are unintentional learners because they live in regions where the established practice is to broadcast original programmes subtitled. The usual explanation for such practice is that subtitling is cheaper than dubbing, but the situation is somewhat more complex and includes a number of socio-cultural considerations such as local conventions, audience profile and the like. Informal/unintentional viewers/learners are interested in developing lexical resources and pragmatic skills, as their main concern is to enjoy films, television programmes or live events. If by convention these programmes are screened or broadcast in the original, with the support of subtitles in the TL, viewers’ exposure to the second language works to their advantage, because they can profit by all the beneficial influences arising from the audiovisual environment. In fact, the effectiveness of subtitles is crucially dependent upon the hidden semiotic connections between text and image, which affect the meaning of the visual-linguistic message and the way in which the spoken/written message is ultimately received. Due to the subsidiary presence of first-language written cues, this type of audiovisual setting may be considered a semi-natural way of second language acquisition, whose benefit can be evaluated according to the type or category of viewers. Both verbal and nonverbal cognition are interwoven throughout multimedia subtitled products, emphasising the powerful role of simultaneous interdependence of linguistic and non-linguistic knowledge Subtitling: Language Learners and the AV Market 247 and mental imagery. In fact, building referential links between accurate mental representations of the word forms and mental images of relevant pictures is useful in understanding and remembering words (Sadoski and Paivio, 2001: 166). Fixing image–form connections is an internal process, which accounts for how learners handle input data and how they use L2 resources in the production of L2 messages. The input is shaped to make it learnable and each individual learner works on the input according to his/her L2 experience, matched to his/her knowledge of the world, to turn input into intake (Ellis, 1985: 166). The reception of subtitled audiovisual programmes is an invaluable device for internalising and automating L2 knowledge. If a viewer is practising this learning situation every day, the reception process becomes a strategic device that motivates the spontaneous use of existing communicative resources. If communication strategies through viewing subtitled programmes are not practised every day, irregular exposure to L2, although facilitated by AVT, is not enough to reach production ability (speaking/writing). It triggers passive/receptive skills, whose effectiveness is directly proportional to the mastery of L2 grammatical data suited to the viewer’s stage of development. People learn a second language by being exposed to it, but they have to be taught or corrected for their mistakes, because the linguistic rules that are not shared by their first language cannot develop unconsciously. Second language learning is a process which involves active mental processes and not simply the forming of linguistic habits. When intentional or unintentional learners have to tackle a second language in a multimedia situation, the effectiveness of their response depends on the way they are able to exploit their cognitive and affective variables. In other words, it depends on how they combine the mental processes they make use of (thinking, remembering, perceiving, inferring, recognising, classifying, generalising, monitoring, analysing, evaluating and memorising) with their personality, attitudes, motivation and emotions. Learners’ proficiency in a foreign language, as the result of what has been taught and learnt during a period of instruction, that is language learners’ achievement, may be contrasted with language aptitude, which is the natural ability to learn a language, not necessarily including intelligence, motivation or interest. Films are always popular sources of foreign language teaching materials, and there is no doubt that they are valuable tools for general unintentional language acquisition – based on the gradient of mere language aptitude – as well as for formal 248 Annamaria Caimi intentional learning goals. In fact, a good standard of subtitling encourages formal students to: Understand more about the theme of the film. Recognise degrees of information. ● Consider and recognise different registers. ● Assess the complexity of the text. ● Appreciate the storyline. ● ● In the third millennium, language acquisition is pivotal to the ongoing process of globalisation but as a collective practice, it always has to come to terms with individual potential or limitations. The marketing perspective: cost-effectiveness vs. subtitling accuracy The quality of AVT often leaves much to be desired and this is a drawback for both untutored/naturalist acquisition, and tutored/classroom acquisition. Audiovisual programmes in a foreign language are in need of language mediation – for example subtitling – which demands continual improvement in quality. In order to boost an international policy of interlingual and intercultural promotion, the actors of the subtitling process (professional screen translators, multilingual subtitling companies, film producers and distributors) should reach a trade-off between what is considered an extra cost for the audiovisual product, that is subtitling, and the most cost-effective way of marketing subtitled programmes. Marketers must comply with the producers’ and distributors’ need for a desired level of transactions with a target market, and analyse the demand for subtitled films in different areas of the audiovisual landscape. The marketing task is to find ways to connect the benefits of the product with the viewers’ needs and interests, by influencing the level of demand in a way that will help the company achieve its objectives. This task is carried out by marketing research and planning. Within marketing planning, marketers must make decisions on target markets, market positioning, product development, pricing, channels of distribution, physical distribution, communication and promotion (Kotler, 1988). As we are moving into a more complex society where communication at intercultural and international level is growing rapidly, it is of paramount importance to safeguard the communication possibilities of the languages in use by multimedia products. Serious reflections on all the Subtitling: Language Learners and the AV Market 249 linguistic constraints that audiovisual translators must overcome should be shared by linguists and language teachers, as well as by film marketers. Linguists and language teachers are concerned with linguistic and cultural differences that exist between SL and TL. These differences are evidenced by a series of variables such as experience of translation practice, TL and target culture competence, general linguistic and extralinguistic knowledge, typological knowledge, age of the translator and level of pedagogic practice. Marketers should be aware of the importance of the tasks carried out by film translators in order to increase users’ options for consumption, and to value their work, not only by using emotional appeal in marketing, and advertising the storyline of subtitled products, but also by highlighting their cross-cultural and cross-linguistic value. The international community of viewer-learners is offered a dense network of multilingual, multimedia optional products which are functional only if the quality of the translation is good. The notion of quality cannot be separated from the notion of risk and uncertainty. Every economic transaction entails a division of work between seller and buyer (you produce a film, I watch it), but such division of work pertains to different types of activity: one is risk-taking, which is predictable and can be planned for only on a statistical basis; the other is represented by envisaging uncertainties that cannot possibly be found by statistics. The risk part of transactions is increasing significantly, but uncertainties are greater and unpredictable in our society, requiring increased capacity for learning and adaptation. The demand for higher quality in film translation can be seen as an attempt by the viewer-learners to force producers to improve their products, to manage more consciously the risk and uncertainty of their offers. From a political perspective, we should resort to legal constraints to attain a balance between cultural and linguistic values and economic profit. For example, the documents of the European Commission, or of other highly influential international institutions that promote interlingual and intercultural education, should mention the importance of optimal quality in AVT and set guidelines to safeguard language correctness and appropriateness, in order to guarantee viewer’s satisfaction and keep language accuracy at a high level. This is a fundamental prerequisite because the general laws of the market, no matter what the marketed product may be, are profit-oriented rather than quality-oriented. Since screen translation is an extra cost, which is often charged to local distributors, the respect for language correctness should be suggested, if not imposed by law. 250 Annamaria Caimi A second prerequisite is the development of highly advanced courses on AVT with the aim of honing the students’ linguistic as well as technical skills (see Bartrina, this volume). In addition to learning all the translation strategies appropriate to the transfer of dialogue into subtitles, without leaving aside the question of cultural adaptation, students should learn to match the rhythm of the film, to determine the in and out time for each subtitle, to maintain synchronisation between text and sound. They should also learn how to edit the subtitles and to correct and polish the language where necessary. In other words, at the end of the course they must become professional subtitlers, able to work with sophisticated subtitling equipment and as part of a team. A third prerequisite concerns training programmes for film producers and distributors, which should include lessons on the function and the benefits of subtitling for international sales and distribution. People involved in the communication business should prepare a subtitled version for all communication formats, including cinema. Every major advance in subtitling technology, which covers everything from offline PC-based subtitle preparation systems to multilingual, fully automated transmission solutions, should be catapulted into the twenty-first century as the unavoidable framework for linguistic and cultural values. This will give due emphasis to the role of languages as the most important means for international exchange and mutual comprehension, the most important means of communication. Note 1. See Pavesi and Tomasi (2001) for an in-depth study on the translation universal of simplification. References Baker, M. (1993) ‘Corpus linguistics and translation studies: Implications and applications’. In M. Baker, G. Francis and E. Tognini-Bonelli (eds) Text and Technology: In Honour of J. Sinclair (pp. 233–50). Amsterdam and Philadelphia: John Benjamins. Baker, M. (1996) ‘Corpus based translation studies: The challenges that lie ahead’. In H. Somers (ed.) Terminology, LSP and Translation. Studies in Language Engineering in Honour of Juan C. Sager (pp. 175–86). Amsterdam and Philadelphia: John Benjamins. Baldry, A. (ed.) (2000) Multimodality and Multimediality in the Distance Learning Age. Campobasso: Palladino Editore. Subtitling: Language Learners and the AV Market 251 Blum-Kulka, S. and Levenston, E. (1983) ‘Universal of lexical simplification’. In C. Faerch and G. Kasper (eds) Strategies in Interlanguage Communication (pp. 119–39). London and New York: Longman. Caimi, A. (ed.) (2002) Cinema: Paradiso delle Lingue. I Sottotitoli nell’Apprendimento Linguistico. Special Issue of RILA (Rassegna Italiana di Linguistica Applicata) 34(1–2). Díaz Cintas, J. and Fernández Cruz, M. (2008) ‘Using subtitled video materials for foreign language instruction’. In J. Díaz Cintas (ed.) The Didactics of Audiovisual Translation (pp. 201–14). Amsterdam and Philadephia: John Benjamins. Ellis, R. (1985) Understanding Second Language Acquisition. Oxford: Oxford University Press. Gottlieb, H. (2002) ‘Titles on subtitling 1929–1999’. In A. Caimi (ed.) Cinema: Paradiso delle Lingue. I Sottotitoli nell’Apprendimento Linguistico. Special Issue of RILA (Rassegna Italiana di Linguistica Applicata) 34(1-2): 215-397. Kotler, P. (1988) Marketing Management. Englewood Cliffs: Prentice Hall. Krashen, S. (1981) Second Language Acquisition and Second Language Learning. Oxford: Pergamon. Laviosa, S. (1998) ‘Universals of translation’. In M. Baker (ed.) Encyclopedia of Translation Studies (pp. 288–91). London: Routledge. Pavesi, M. and Tomasi, A. (2001) ‘Per un’analisi degli universali della traduzione in testi scientifici: L’esplicitazione e la semplificazione’. In C. Bettoni, D. Calò and A. Zampolli (eds) Atti del secondo convegno dell’AITLA (pp. 129–49). Perugia: Guerra. Pavesi, M. (2002) ‘Sottotitoli: dalla semplificazione nella traduzione all’apprendimento linguistico’. In A. Caimi (ed.) Cinema: Paradiso delle Lingue. I Sottotitoli nell’Apprendimento Linguistico. Special Issue of RILA (Rassegna Italiana di Linguistica Applicata) 34(1–2): 127–42. Pavesi, M. and Perego, E. (2008) ‘Tailor-made interlingual subtitling as a means to enhance second language acquisition’. In J. Díaz Cintas ( ed.) The Didactics of Audiovisual Translation (pp. 215–25). Amsterdam and Philadelphia: John Benjamins. Sadoski, M. and Paivio, A. (2001) Imagery and Text – A Dual Coding Theory of Reading and Writing. Mahwah, N.J. and London: Lawrence Erlbaum. Index A Fish Called Wanda, 226 accent, 134, 137, 224 accessibility, 1, 5–6, 12–13, 86, 149, 163, 231–3, 238 UK, 6, 12, 16 USA, 6, 12, 16 see also audio description; blind; deaf; subtitling for the deaf and hard-of-hearing adaptation(s), 10–12, 74, 76, 81, 95, 115–17, 119–20, 122–8, 153, 192, 237, 249–50 Spain / Germany, 12, 115ff Adriana Lecouvreur, 68 Albert Herring, 10, 71, 74–7, 79, 81 American productions, 88, 98, 119 anime, 49 audience design, 151–3 audience reception, 3, 231–3 Italy, 11, 97ff Norway, 87–9, 94, 208–9 Spain, 100, 232 see also dubbing; subtitling audio describer, 13, 170, 172–80, 184 audio description (AD), 2, 6, 11, 180, 183–4 decision making, 174–6 impartiality, 173–4 targets, 16 theatre, 13, 170ff see also blind; partially sighted audio subtitling, 7 audiovisual market, 12, 15, 30, 112, 151, 200, 240ff Billy Elliot, 242 blind, 2, 5–7, 170, 172, 180 see also accessibility; audio description; partially sighted body language, 6, 92, 203, 205, 210–11, 217 Braveheart, 25 British Broadcasting Corporation (BBC), 1, 12, 133, 135, 140–3, 145–7 Britten, Benjamin, 10, 62, 71, 74–6, 78–9 see also opera Bohème, La, 66 broadcast interpreting, see interpreting captioning, 56–8, 64, 91, 155 closed, 5 see also subtitling for the deaf and hard-of-hearing; teletext Carmen, 63, 73, 124 cartoons, 2–4, 102, 106, 215 Casa de Bernarda Alba, La, 120–1 channel acoustic, 37–8 visual, 23, 37–8, 154, 197, 198 cinema, 2, 4, 6, 10, 22, 48–9, 58, 88, 105, 115, 118–19, 123, 125, 159, 217, 232, 240, 250 audience(s), 94 see also film Club Dumas, El, 122 coherence, 47, 159, 162, 167 cohesion, 22, 25, 47, 160, 162 cohesive devices, 87 commentary, 24, 131, 133, 173–4 condensation, 50, 86, 96, 236, 245 constraint(s), 29, 34, 36, 43–4, 47, 57, 79, 92, 159, 232, 237, 240, 249 financial, 5 linguistic, 21, 26, 249 technical, 21–2, 50, 151, 225 textual, 21, 26 see also dubbing; subtitling; voice-over convention(s), 32–3, 64, 67, 80, 110–11, 152, 164, 202, 207, 235, 246 252 Index 253 cooperative principle, 198–200, 202, 207 see also maxims (Gricean); pragmatics cueing, 31, 37, 43, 45–6, 67, 234 see also spotting; timing cultural reference(s), 10, 79, 100, 102 , 188–9, 223 Curious Incident of the Dog in the Night, The, 178 cut(s), 41, 43, 46, 65, 216 deaf/Deaf, 5, 153–67 see also accessibility; hard-ofhearing; subtitling for the deaf and hard-of-hearing deletion, 25, 236 Descriptive Translation Studies, 9, 36 dialect(al), 11, 26, 88, 109, 156, 242–4 see also idiolect(al); sociolect(al) dialogue list, 91, 95, 135–6, 234 digital, 1, 6, 13, 22, 59–60, 135, 166, 229–31, 233–5, 237 technology, 3, 6, 15, 165, 229 discourse, 25–6, 75, 134, 151, 216–26 distributor(s), 3, 159, 236–7, 248–50 documentary, 91, 95, 135–6, 142, 144, 146, 215 dubbese, 102, 108–12 Italy, 102ff see also audience reception; dubbing dubbing, 4, 13, 97ff, 125, 130–4, 167, 229, 240 constraints, 92 countries, 2–3, 11, 32, 36–7, 85, 93, 95, 97–8, 119, 230 industry, 98–9, 111–12, 119 versus subtitling, 10–11, 85ff, 151–2, 246 see also revoicing; voice-over Dumas, Alexandre, 67 DVD, 1–4, 13, 22–3, 30, 32–3, 38, 48, 56, 58, 65, 67, 69, 112, 118–19, 152, 159, 165–6, 217, 231–2, 240 market, 6, 9, 21, 34, 97, 232, 234 editing, 34, 59, 65, 90, 148, 163, 219, 233, 235 English National Opera, 59, 61, 73 English Patient, The, 9, 39–41, 44 equivalence, 10, 37, 77–8, 183–4, 210, 212 European Commission, 112, 249 TV channels, 1, 95 Union, 6 European Blind Union, 6 European Captioning Institute (ECI), 9, 33–4 expletive(s), 65 see also swear word(s); taboo language explicitation, 161 exposure time, 86–8 fansub(s), 49 see also (amateur) subtitling feedback effect, 25–6, 29 Fellowship of the Ring, The, 9, 50, 57 film(s), 2–4, 6, 23, 26, 36–8, 85, 90, 95, 99, 106, 117, 131, 156, 161, 164, 170–1, 197, 210–11, 217, 219, 246–8 dialogue, 25, 133, 212, 219, 227, 241 family, 94, 96 festival(s), 1, 92, 118–19 genres, 47, 86, 123, 125, 215 industry, 1, 8, 49, 85, 92, 98, 118, 240, 248–50 Studies, 130, 133, 238 text(s), 14–15, 216, 218 transfer norms, 11–12, 115, 118, 124 translation, 15, 56, 200, 216, 249 see also adaptation(s) frame(s), 46, 58, 214, 216, 221 globalisation, 4, 14, 98, 142, 231, 240, 248 Gregory Hines’ Show, The, 94 Grice, Paul: see maxims 254 Index hard-of-hearing (HoH), 13, 154–8, 164–7 see also accessibility; deaf/Deaf; subtitling for the deaf and hard-of-hearing humour, 10, 74, 77–9, 81, 100, 123, 181, 226 joke(s), 67–8, 100 ICT, 229 idiolect(al), 26, 244 see also dialect(al); sociolect(al) implicature(s), 198–201, 203, 205, 209, 212 see also pragmatics; Speech Act Theory insert(s), 23, 33, 52, 91, 109 Internet, 1–4, 9, 13, 15, 49–50, 56, 68, 106, 187, 191, 203, 233, 238 interpreting, 4, 11–12, 50, 132, 140–1, 145–7 broadcast interpreting, 12, 140–8 Japan / UK, 12, 140–7 simultaneous, 141–3, 145–6, 170 war in Iraq, 144–6 intersemiotic, 116, 162, 164, 236 see also semiotic; polysemiotic intonation, 25, 75, 86–7, 93, 183 irony, 33, 161, 168, 201 Japan Broadcasting Corporation (NHK), 12, 140, 142–7 kinesic, 25, 214, 221 language transfer norms, 115ff see also film transfer norms; norms language learning, 7, 15, 97, 240, 242, 246–7 learning, 15, 233, 238, 248–50 legibility, 22, 60, 232 lip synchronisation, see synchronisation localisation, 13–14, 186–92, 224–5 Human Computer Interaction (HCI), 189–90, 192 Long Day’s Journey into Night, 226 Man Who Mistook his Wife for a Hat, 171 Manhattan Murder Mystery, 40, 44 maxim(s) (Gricean), 198–203, 207–8, 212 see also implicatures; pragmatics; Speech Act Theory Meet the Parents, 205–6 mother tongue, 31, 100, 134, 155, 157, 166, 186, 243 multimedia, 15, 229–30, 235, 240, 246–9 multimodal, 14, 163, 214, 216–22, 225 multilingual, 9, 14, 191–2, 249–50 Murder One, 95 music, 4, 10, 14, 38–9, 47, 71–81, 105, 108, 153, 164–5, 167, 214, 217–18, 220–3 narration, 4, 6, 131, 135 narrative, 14, 25, 73, 90, 123, 125, 163, 173–5, 177, 182, 212 noise(s), 38, 47, 50, 155, 160 norm(s), 9, 12, 22, 36–7, 39, 41–4, 46–7, 100–1, 103, 109–10, 115, 118, 124–6, 160, 188, 224 Norwegian Broadcasting Corporation (NRK), 87, 89 Notting Hill, 9, 39–43, 218 omission(s), 37, 40–1, 43–4, 46, 86, 160, 201, 209, 226, 236 online, 15, 118, 187, 190, 217, 229–33, 235–7 see also training opera(s), 1, 10, 58–69, 71–81 French, 72, 74, 76–80 German, 61, 65, 72, 77 Italian, 60–2, 65–6, 68, 72 Kobbé, G., 73–4 libretto adaptation, 71–81 see also surtitling paralinguistic, 5, 156, 161–2, 164, 167 partially sighted, 2, 6, 170, 172 see also accessibility; audio description; blind Perfect Storm, The, 39–40 Index 255 phasal analysis 219–23, 225–7 see also multimodality pivot language, 116 translation, 134 polysemiotic, 156 see also intersemiotic; semiotic Polysystem Theory, 12, 115 pragmatics, 14, 197–8 see also cooperative principle; maxims; Speech Act Theory Psycho, 23 Pullman, Philip, 175 pun(s), 74, 78, 102 punctuation, 62, 159, 161 quality, 15, 53, 56, 85, 92, 211, 232, 234, 238, 248–9 audio description, 172–3, 178 Catalan, 230–1 dubbing, 11, 99–100, 106, 111 interpreting, 12, 142–3, 146–7 localisation, 189 maxim of, 199–200 subtitling, 9–10, 34, 39, 49–51, 230–3, 237, 240 training, 230–5, 237–8 see also (amateur) subtitling readability, 21–3, 26, 86–7, 96, 210, 232 reading speed(s), 22, 24–5, 34, 58, 86, 88, 160, 197, 232 redundancy, 25, 156, 160–2, 236 reduction, 26, 30, 160, 236, 241, 245 register, 75, 77, 88–9, 159, 248 relevance, 156–7, 163, 199, 207, 232 revoicing, 4, 10–11, 95, 131 see also dubbing; voice-over Royal National Institute for the Blind ( RNIB), 170, 180 Royal Opera House, The (ROH) 59, 62, 64 Sacks, Oliver, 171–3, 184 second language, 52, 154, 160, 166, 240–1, 247 acquisition, 7, 246 segmentation, 24, 37 semiotic(s), 13–15, 99–100, 203, 212, 214, 216–22, 225, 236–7, 246 see also intersemiotic; polysemiotic shot change, 23, 42, 90, 232 sign(s), 47, 103, 109, 156, 161–2, 234 sign language, 16, 154–5 sitcom(s), 2–4, 88, 98, 102, 106, 110 slang, 31, 65, 210 soap opera(s), 2–3, 98, 102, 111, 162, 167, 214–16, 220 sociolect(al), 11, 88, 156 see also dialect(al); idiolect(al) software, 58–60, 69, 183, 190, 237 song(s), 33, 44, 64, 71–2, 76, 81, 102, 234 sound effect(s), 28, 153, 157–8, 161, 164 soundtrack(s), 4–5, 9–12, 22–3, 25–6, 28, 30–1, 87, 93, 132, 152, 170, 217, 221, 241 Speech Act Theory, 14, 197–8, 202 see also maxims; pragmatics spotting, 31, 37, 48, 225, 233–5 see also cueing; timing subtitle(s) / subtitling accuracy, 248–9 amateur, 9, 49–51, 54–7 audio subtitling, 7 constraints, 9–11, 21–2, 26–7, 29, 36, 86, 90, 234–5 countries, 11, 13, 85, 90, 97, 246 for the deaf and hard-of-hearing (SDH), 2, 5–6, 13, 16, 28–9, 151–3, 163, 165–7 Greece / Spain, 9, 36ff guidelines, 237–8, 249 interlingual, 6, 13, 21, 151–2, 156, 163, 165–7, 230, 240 intralingual, 6, 12–13, 152–3, 156, 158, 163, 165–7 live, 49, 58 master, 9, 234 multilingual, 9, 30, 248 Norway, 197ff software, 3, 57, 198, 230–1, 237 see also training supertitles, 58 256 Index surtitles / surtitling, 1, 9–10, 58–69, 71, 73–4, 81 see also opera swear word(s), 89, 103, 110 see also expletive(s); taboo language synchronisation / synchrony / sync, 21, 37–9, 43, 45–6, 90, 132, 160, 236, 250 lip, 4, 92, 94, 103, 110, 131 taboo language, 89, 103, 110–11 teletext, 5, 22, 58, 165, 167 see also captioning; subtitling for the deaf and hard-of-hearing television (TV), 1–7, 10–11, 13, 16, 22, 38, 47, 49, 58, 65, 69, 85–91, 93–5, 98, 101, 111, 118–19, 132–3, 136, 141, 158–9, 164–6, 215–18, 220, 222–3, 246 channels, 1, 95, 137 template(s), 9, 30–4 theatre, 1, 10, 13, 58, 61, 65, 69, 72–3, 81 timing, 22–3, 28, 30–3, 74, 90, 225, 233–5 see also cueing; spotting training, 14, 157, 227, 229, 232–3, 238, 250 audio description, 172, 177–8 interpreting, 143 online, 229ff subtitling, 7, 14, 31, 47, 214 Catalonia / Spain, 229–31 transadaptation, 152 Twelfth Night, 171 Un posto al sole, 111, 214 Vertigo, 24, 27 VHS, 3–4, 33, 48, 119, 159 voice-over, 4, 9, 11–12, 92–3, 95, 130–7, 144–5, 229 constraints, 137 war in Iraq, 135–7 see also dubbing; revoicing Weird Science, 24, 28 Willy the Fresh Prince of Bel Air, 110