See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/285523525 2015 Papers in Arabic-English Translation Studies Book · January 2015 CITATIONS READS 2 57,039 1 author: Mohammed Farghal Kuwait University 126 PUBLICATIONS 523 CITATIONS SEE PROFILE Some of the authors of this publication are also working on these related projects: Translation of Adverbial/Determiner Negation View project All content following this page was uploaded by Mohammed Farghal on 02 December 2015. The user has requested enhancement of the downloaded file. Papers in Arabic/English Translation Studies 1 An Applied Perspective Mohammed Farghal et al. Published by The Jordanian Translators’ Association (JTA) This volume consists of 15 self-contained papers dealing with a variety of applied topics in Arabic/English Translation Studies (TS). The aim is to bring together and update some scattered material in one volume, which is hoped to inaugurate a series of other volumes in a scantily researched area. This will definitely serve students, researchers, and practitioners who usually experience difficulty in locating academic papers dealing with Arabic/English TS. The reader can choose to read any article independently of the others according to his/her own interests. To avoid repetition in the bibliography, a unified list of references is provided separately at the end of the volume. Professor Mohammed Farghal Kuwait University In this volume, one thing is certain: Farghal and his colleagues have gone a very long way towards achieving the goal of encouraging and guiding practitioners to study their own translational behavior and to determine what works best. Like all professional activities, translating is a complex phenomenon, and there is no single ‘right’ approach. But reflection is the name of the game in what we do as translators or interpreters: We do indeed reflect on different versions, different modes and different models, comparatively assessing the merits and demerits of a particular strategy, and in the process reshaping past and current experiences in a manner that can only lead to improved practices. Professor Basil Hatim American University of Sharjah To the Memory of Professor Lewis Mukkattash & Professor Abdullah Shakir i Acknowledgments First and foremost I would like to thank my former as well as my present graduate students who have been a source of inspiration for me over the years and who have contributed to this volume both directly and indirectly. Several of them are now university professors of Linguistics and/or Translation Studies in Jordan and abroad. My thanks also go to my colleagues at Yarmouk University and Kuwait University whose scholarly discussions and commentaries have contributed to the shaping of many of the ideas in this volume. In particular, I would like to mention my teacher and colleague the late Professor Lewis Mukattash and my colleague the late Abdullah Shakir for whom this volume is dedicated. Their love for linguistics and translation will always be a source of inspiration for us and their memories will certainly stay with us. I am obliged to Professor Basil Hatim for kindly accepting to write the foreword to this book. I am also grateful to Mr. Omar Irshaidat for helping with many technical matters. Last, but not least, a warm word of thanks goes to my close friend and colleague Professor Abdullah Shunnaq for suggesting Jordanian Translators’ Association as an outlet for publishing this volume. ii Table of Contents Dedication Acknowledgments List of Abbreviations Preface List of Phonetic Symbols Foreword 1-5 1. Basic Issues in Translator Training 7-16 Mohammed Farghal 2. The Linguistics of Translation 17-40 Mohammed Farghal & Ali Almanna 3. The Translation of English Passives into Arabic 41-54 Mohammed Farghal & Mohammed Al-Shorofat 4. Translating Modals by Zero Equivalents: The Case of Macbeth 55-64 Mohammed Farghal & Alban Beqri 5. Semantic and Syntactic Hurdles in Machine and Student Translation 65-74 Mohammed Farghal & Adnan Gergeos 6. Translational Miscues: Poetry as an Example 75-88 Mohammed Farghal & Rula Naji 7. Audience Awareness and Role of Translator 89-98 Mohammed Farghal & Abdullah Shakir 8. Reader Responses in Quran translation 99-114 Mohammed Farghal & Mohammed Al-Masri iii 9. Coherence Shifts in Quran Translation 115-126 Mohammed Farghal & Noura Bloushi 10. Pragmalinguistic Failure: The Case of Arabic Politeness Formulas 127-144 Mohammed Farghal & Ahmed Borini 11. Media Translation: The Case of The Arabic Newsweek 145-164 Mohammed Farghal & Mashael Al-Hamly 12. Collocations: An Index of L2 Interlingual Transfer Competence 165-174 Mohammed Farghal & Abdullah Shakir 13. Explicitation vs. Implicitation: Discourse Markers as an Example 175-188 Mohammed Farghal & Abdullah Samateh 14. Lexical Reduction in Scientific Translation 189-202 Mohammed Farghal & Mashael Al-Hamly 15. Major Problems in Student Translations of English Legal Texts into Arabic 203-209 Mohammed Farghal & Abdulla Shunnaq List of References 211-227 iv Preface This volume consists of 15 self-contained papers dealing with a variety of applied topics in Arabic/English TS. The aim is to bring together and update some scattered material in one volume, which is hoped to inaugurate a series of other volumes in a scantily researched area. This will definitely serve students, researchers, and practitioners who usually experience difficulty in locating academic papers dealing with Arabic/English TS. The reader can choose to read any article independently of the others according to his/her own interests. Where employed in this volume, square brackets indicate literal translation, which is meant to give a rough idea about the propositional content of some Arabic texts. Boldface typing is mainly used to highlight the study items within the textual data, which is given in Arabic script in some articles and in transliteration in others. To avoid repetition in the bibliography, a unified list of references is provided separately at the end of the volume. The textual data is mainly drawn from published as well as student translations. It covers a variety of discourse types/genres including the literary, the religious, the journalistic, the legal, etc. The aim is to offer insights relating to translation activity in its professional as well as its pedagogical sphere, hence the applied perspective of this volume. The discussion of the textual data is not meant to offer final solutions but rather to engage the reader in what we may call ‘translational argumentation’, where the relevance of theoretical constructs to actual translation practice is highlighted. v List of Recurrent Abbreviations TS Translation Studies SL Source Language TL Target Language L1 First Language L2 Second Language vi List of Arabic Phonetic Symbols /b/ voiced bilabial stop /m/ bilabial nasal /f/ voiceless labio-dental fricative /ð/ voiced interdental fricative /ð/ voiced interdental emphatic fricative /ө/ voiceless interdental fricative /d/ voiced alveolar stop /t/ voiceless alveolar stop /ḍ/ voiced alveolar emphatic stop /ṭ/ voiceless alveolar emphatic stop /z/ voiced alveolar fricative /s/ voiceless alveolar fricative /ṣ/ voiceless alveolar emphatic fricative /n/ alveolar nasal stop /r/ alveolar rhotic liquid /l/ alveolar lateral liquid /š/ voiceless alveo-palatal fricative /j/ voiced alveo-palatal affricate /y/ palatal glide /w/ labio-velar glide /k/ voiceless velar stop /ɤ/ voiced uvular/post velar fricative /x/ voiceless uvula/post velar fricative /q/ voiceless uvular stop /‘/ or /c / voiced pharyngeal fricative /ḥ/ voiceless pharyngeal fricative /’/ glottal stop /h/ voiceless laryngeal fricative /i/ high front short vowel /u/ high back short vowel /a/ low half-open front-to-centralized short vowel /ii/ high front long vowel /uu/ high back long vowel /aa/ low open front-to-centralized long vowel /ee/ mid front long vowel /oo/ mid back long vowel vii Foreword by Basil Hatim In presenting you with this volume of excellent collaborative work in translation theory and practice, I want at the outset to highlight and speak to one particular theme I consider pivotal in the present context. This is ‘research’, and the role research plays in promoting ‘reflective practice’ not only in doing translation but also and equally significantly in developing a theory of doing translation. As the editor of the present book puts it in his introductory contribution (p. 5): “translation theory is intended to refine and sharpen the already existing level of translating theory by bringing to consciousness a set of strategies and principles in practicing and/or prospective translators.” This is just one of the many important claims which this timely volume makes, forcibly and insightfully. There are many reasons for my choice of thematic focus on research and the reflective practitioner, but one particular reason stands out. In reading through the various contributions to this new book by Mohammed Farghal et al, one feature struck me as distinctive, namely how each contribution tells upcoming researchers the full story: from a ‘given’ assumption or set of assumptions accepted so uncritically by both practitioner and theoretician of translation alike for far too long (a ‘problem’, in short), through a questioning which leads to ‘new’ proposals put forward as possible ‘solutions’, then through the analysis of appropriate ‘data’ marshalled in support, to arrive finally at some plausible conclusions and practical applications that translators, students of translation and, indeed, teachers and theoreticians of translation, will find highly insightful (see Candlin, Crompton & Hatim 2015). This ‘reflexivity’, or empowerment of practitioners to develop and execute their own research (Hatim 2012), is one way in which unhelpful dichotomies such as ‘theory’ vs ‘practice’ may be re-assessed, as the present volume does admirably. In short, this is what I would unreservedly call ‘quality research’, and the ‘how-to’ slant which threads its way through the book and which was obviously necessitated by the authors’ concern with their students and with the translation classroom within which they operated, is a most opportune addition to the literature. I, would, therefore, want to recommend this book not only as a handbook on the theory and practice of English/Arabic translation, but also as a research manual, a how-to guide for practitioners trying to find their way, through the maze of reflection, towards becoming effective researchers themselves. To perform its role effectively, one of the aims of a theory of translating (with people like Nida), as opposed to a theory of translation (as Catford would have it, for example), is to start by raising awareness. Essentially, this is to promote the realization that a given problem is ‘real’ and that it is recognized as a problem by the majority of those it affects. Once a problem is defined, solutions can be sought, a process of encouraging the kind of practitioner research that has aptly acquired the name ‘action research’. Up to the 1980s, research in Translation Studies was generally a matter of ‘reflection’, and practitioners looked to theoreticians almost as the sole providers of this body of ‘analytical’ knowledge that could help them expand the ‘craft’ knowledge they already possessed. But then a wind of change came and a generation of young researcher-writers began to occupy the forefront in theorizing practice and practicing theory. That is to say, we saw a reversal of the trend of treating practitioners as mere consumers of research. The practitioner is now viewed as someone who is heavily involved in the determination of the problem or the ‘puzzle’. And here, I beg the reader’s indulgence to allow me relate a personal anecdote. I was on a visit to Jordan in the mind-1980s and had stopped by to see a dear colleague, now departed (the late Professor Lewis Mukattash), when a young professor walked in fresh from an excellent PhD from Indiana (that was Mohammed Farghal) all jubilant with what I think must have been one of his earliest publications, and the title, if memory serves, was the New Toy Effect. It was about how new fashions in applied linguistics can be deceptively alluring. It was then that I saw originality and imagination, qualities that Mohammed Farghal has demonstrated over the years, and that the present volume exhibits most conspicuously. But to return to the here and now, the scene for Farghal et al’s new volume is set with ‘Basic Issues in Translator Training’, a major article in which Farghal focuses on translator training in the Arab World and argues among other things for an important distinction to be drawn between a theory of translating and a theory of translation, with the translation activity governed by what he calls “a principle of relevance – the decision to render a segment (or an aspect of it) or not depends entirely on whether that segment is relevant in any given context”. The paper concludes with a re-statement of the need to view translation as an act of communication, with considerations of comprehensibility and readability uppermost in the translator’s mind. In ‘The Linguistics of Translation’, Farghal & Almanna re-visit decision-making in translation and demonstrate how, instead of the static, mono-level view, this process may best be undertaken at different linguistic levels, including phonology, morphology, syntax, and semantics. Syntax has pride of place when Farghal & Al-Shorafat delve deeper into the intricacies of the English passive and examine ways of dealing with this complex structure in Arabic. This challenges traditional, intuitive views predominant in studies of the translation of English passives, and test these against empirical data consisting of translations of English passive utterances as they naturally occur in an English text. A number of strategies seem to be employed by translators in dealing with this particular structure, including nominalization, adjectivalization, and activization. Staying with syntactic structures, ‘Translating English Modals’ by Farghal & Beqri concludes that, while English possesses a highly grammatical zed 2 system of modality, Arabic employs a diversity of modal expressions which can effectively capture the various shades of meaning encapsulated in English modals. Next, Farghal & Gergeos examine ‘Machine and Student Translation’ and address some subtle English syntactic features that tend to present formidable problems to both machine and student. There is first the issue of recognizing variation in parts of speech, with polysemous and/or homonymous syntactic categories likely to cause serious interpretation problems to both machine and student translators. Thence, we have the issue of structural ambiguity at phrase and sentence level creating interpretation difficulties in translation activity. Third, parenthetical structures in English sentences for their part may trigger reference and focus mishaps in machine and student translation alike. Finally, the complications of word order variation are examined and the damage that may be caused in translation due to such aspects of message construction as marked stylistic inversion and subject deletion is identified. From the world of machine translation, Farghal & Naji take us next into the murky area of ‘verse translation”. The notion of ‘miscue’ is central here. From a translational perspective, miscues are shown to occur at the phonological, morphological, syntactic, and semantic levels, often twisting, or even crippling, the cognitive structures communicated by poets. When culture enters the equation and intercultural mistranslations become a fact of life, the need on the part of the translator to be aware of the culture-bound norms of expression in the TL becomes most acute. With Shakir (May Allah bless his soul), Farghal deals with this issue in ‘Audience Awareness and Role of Translator’. By employing an Arabic political editorial translated into English by their MA translation students, the authors show that to avoid deviations that can mar the translation product in terms of micro- and macro-features, translators need to be vigilant to address features such as lexical appropriateness, tone of presentation, and content organization. From sensitive (poetic or political) texts, we move next to sacred texts, specifically to ‘Coherence Shifts in the Translation of the Holy Quran’. Here, Farghal & Alblushi establish a crucial distinction in Quranic translation, namely reader-focused vs. text-focused coherence shifts. It is suggested that the wide cultural distance between source text and target text may cause reader-focused shifts when opting for literal translation in handling partial and complete referential gaps. On the other hand, the translator’s inadequate language competence in Arabic and the absence of consulting Quranic exegeses sometimes trigger text-focused shifts, thus offering unintended readings. We remain with sacred and sensitive texts in translation and survey Reader Responses in Quran Translation, a paper in which Farghal & Al-Masri deal with the specific issue of ‘referential gaps’ in the Quranic text. The study shows that referential gaps are problematic to process in English translations of the Holy Quran. This is an extremely subtle level of Quranic discourse, and one that, although examined extensively in the exegetical literature by Arab jurisprudents and rhetoricians alike, has not made its mark 3 in work published in English, hence the importance of this particular contribution. The paper empirically shows that what the target reader may think is a correct interpretation of a Quranic verse may turn out to be a distorted reading caused by a referential gap. Thus, in the absence of a sound cultural rapport with Quranic discourse, the target reader may project his/her own knowledge of his/her own religion on a Quranic text and, subsequently, fall victim to a flawed interpretation. Farghal & Borini engage with pragmatic failures and deal specifically with ‘Arabic Politeness’ issues. The paper addresses the translatability of Arabic politeness formulas into English from a pragmalinguistic perspective. The analysis is conducted with reference to the Cooperative Principle (CP), the Politeness Principle (PP), and the Irony Principle (IP), and argues that “deficiency in pragmalinguistic competence usually results in communication breakdown or, at best, distortion of the original message”. The crucial area of ‘Media Translation: The Case of the Arabic Newsweek’ is most cogently dealt with by Farghal & Al-Hamly. This project is a case study with a focus on two genres seen in terms of the type of discourse employed in each, the evaluative vs. the expository type. The analysis of the data shows that the Arabic version of Newsweek suffers from a variety of local and global perceptible errors relating to lexical, discoursal and grammatical usage. In terms of genre analysis, these perceptible features present themselves more frequently and seriously in argumentative/evaluative discourse than in expository/nonevaluative discourse. ‘Collocations’ (Farghal & Shakir) constitute a key component in the lexicon of natural language, and translators/interpreters should, therefore, possess a working syntagmatic competence, alongside their paradigmatic competence. The study further looks into the strategies adopted by the student translators/interpreters in their attempts to render target collocations: reduction, synonymy, compensation, paraphrasing and calquing. Failure to cope with collocations in the SLT results in mitigating the evaluativeness parameter, thus weakening the line of argumentation in the TLT. Terminology is dealt with next by Farghal & Al-Hamly. The specific issue examined is ‘Lexical Reduction in Scientific Translation’, with English Reduced Forms (RLFs) seen as problematic in Arabic translation. The translation data examined shows that professional scientific translators employ various strategies in rendering a variety of English RLFs. While Blended Forms and Complete Form + RLF are the most frequent RLFs in the English corpus, it is found that Translation Alone and Translation + RLF are the most occurring strategies in the Arabic corpus. The issue of Translation Universals, an extremely interesting and new research area that has not attracted sufficient attention in English/Arabic Translation Studies, occupies Farghal & Samateh in Markers in Translation. The rather bold claim by Blum-Kulka regarding TT explicitation usually being a response to 4 ST implicitation provides a launching pad for a truly insightful exploration of how three discourse markers function in Arabic translation. Three types of correspondence are identified: explicitation to explicitation, explicitation to implicitation, and explicitation to zero equivalent. The paper concludes that due to the syndetic nature of Arabic discourse, there are cases of discourse markers whose sole function is not so much to explicate as to improvise, and in the process to create, smooth and cohesive discourse. Finally, this tour de force in the study of translation, homes in on ‘Legal Texts in Translation’ dealt with by Farghal & Shunnaq, in terms of syntactic, layout, and tenor problems posed in the process of translation. It is revealed that syntactically parenthetical non-finite clauses in Arabic legal discourse seem to be the single most difficult feature to handle. Layout features such as indentation, and tenor features relating to the formality of legal lexis can also be problematic. To conclude, the question which many of us in teaching and researching translation constantly ask, and which this volumes attempts in a major way to answer is: Can the translator or interpreter be trained to become a reflective practitioner in the sense of reflective practice advocated by Donald Shon (1983), for example, and summed up in the following terms by Boud et al (1985: 19) "Reflection is an important human activity in which people recapture their experience, think about it, mull it over and evaluate it. It is this working with experience that is important in learning." One thing is certain: Farghal and his colleagues have gone a very long way towards achieving this goal of encouraging and guiding practitioners to study their own translational behavior and to determine what works best. Like all professional activities, translating is a complex phenomenon, and there is no single ‘right’ approach. But reflection is the name of the game in what we do as translators or interpreters: We do indeed reflect on different versions, different modes and different models, comparatively assessing the merits and demerits of a particular strategy, and in the process reshaping past and current experiences in a manner that can only lead to improved practices. 5 6 Basic Issues in Translator Training: Special Reference to Arab Contexts Mohammed Farghal Abstract 8 The paper addresses the status quo of student translator training programs in the Arab world by looking at the practical and theoretical dimensions of TS as an emerging discipline. It aims to offer a set of principles and guidelines whose presence seems indispensable. First, an introductory word is said about nature of human communication, nature of translation, and translation programs. Second, an important distinction is drawn between a theory of translating and a theory of translation. Third, it is argued that translation activity should always be informed by a principle of relevance – the decision to render a segment (or an aspect of it) or not depends entirely on whether that segment is relevant in any given context. Fourth, translation needs to be viewed as an act of communication governed by considerations of comprehensibility and readability, rather than an act of prescription informed by dogmatic and obsolete views about correctness. Last, translation activity is shown to involve three stages: the pre-translating, the translating and the retranslating stages. 1. Nature of Communication In its essence, translation is an act of interlingual communication which involves the use of language, whether it be in the spoken form (interpreting) or written form (translating). Explaining the nature of human communication, being the raw material for translation activity is, therefore, a prerequisite for embarking on any pedagogical endeavor relating to translation. The production and reception of language (be it spoken or written) is a dynamic, interactive process whereby explicit as well as implicit propositions are smoothly produced and received. The propositional content, or simply meaning, in human discourse embodies two main functions: the affective (phatic) function and the referential (informational) function at varying degrees, with a discernible dominance of one over the other in various discourses. This functional and fluid division of labor, so to speak, captures the usually intertwined interactional and transactional functions of human communication in its entirety (Brown and Yule 1983). The expression of propositions in discourse by language users embraces two distinct, though complementary, principles: the Open Principle (OP) and the Idiom Principle (IP) (Sinclair 1991). The OP emphasizes the productive (generative) nature of human communication which enables language users to produce and comprehend novel propositions by utilizing a finite set of rules whose functionalization rests on already learned vocabulary items. By contrast, the IP stresses the parroted (memorized) component of human communication which enables language users to fall back on a huge amount of multiword units (canonically including collocational, idiomatic, proverbial, and formulaic expressions, among others) to produce and receive previously encountered (parts of) propositions. In this way, meaning in interlingual communication evolves out of constructing meaning via gammaticalizing (the OP) or parroting meaning by calling up multi-word units (the IP) based on the presence of a Source Text (SL). By way of illustration, the propositional content of Cats love dozing under palm trees may turn out to be a novel one (being the product of the OP) and can literally translate into an Arabic utterance that may involve a novel proposition, viz. تحب القطط النوم تحت أشجار النخيل. By contrast, the familiar English proverb Birds of a feather flock together (being the product of the IP) can readily be translated into a familiarly 9 corresponding one in Arabic, viz. [ إن الطيور على أشكاالاا تقكVerily the birds on their forms fall]. The translator’s awareness of the garmmaticalized vs. idiomatized expression of meaning constitutes the foundation stone in translation activity as an act of human communication 2. Nature of Translation The senses of the transitive verb ‘to translate’ embodies three different, though relevant and related, acts, viz. (1) express the sense of (a word, sentence, speech, book) in another language, (2) express (an idea, book, etc.) in another, esp. simpler form, and (3) interpret the significance of; infer as (The Concise Oxford Dictionary, Ninth Edition). Examining these senses, one can immediately see that the first sense is restricted to interlingual communication, i.e. it involves the use of more than one language, while the second is confined to intralingual communication which may involve explaining, paraphrasing, etc. As for the third sense, one can argue that it is relevant to both intra- and interlingual communication. In this way, the language user (whether he is functioning within one language or mediating between two languages) can perform an interpretative act. Actually, the three senses above capture much of the insight and pith of the debate and theorizing voiced by different scholars working in the discipline of translation studies. The relatively recent move from ‘translation equivalence’ (Nida 1964; Catford 1965; Newmark 1981; House 1981) to ‘translation resemblance’ (Gust 1991), and later to ‘skopos’ (Schäffner 2003, 1998; Hönig 1998; Vermeer 2000) represents a steady shift from the first sense to the third sense in the partial dictionary entry above. To see the contrast more clearly, let’s quote from Newmark (1982) and Schäffner (1998). In the words of Newmark, the translator’s task is “to render the original as objectively as he can, rigorously suppressing his own natural feelings …” (1982:389). By contrast, Schäffner views the translator as a TT [Target Text] author who is freed from the “limitations and restrictions imposed by a narrowly defined concept of loyalty to the source text alone” (1998:238). It should be clear that the ‘limitations and restrictions’ are embodied in definition (1), while the ‘freedom’ is embraced by definition (3) above. At a more theoretical level, transforming Meaning from one Form to another involves a cognitive and a linguistic process. The cognitive process in intralingual communication consists in generating and processing ideas (cognitive structures) and, subsequently, transforming them into words and utterances (i.e. a linguistic code). While ideas enjoy a high degree of constancy, the linguistic code is fluid and variable. Thus, the same idea can be clad differently in terms of language expression by adopting variegated styles. In interlingual communication, the cognitive aspect is mainly pertinent to processing and interpreting ideas rather than generating them (i.e. it is a matter of text comprehension and interpretation). However, the linguistic code remains fluid and variable, thus enabling the mediator (i.e. the translator) to offer translations that differ in language expression (i.e. form) but essentially relay similar content. At face value, therefore, the 10 content enjoys a high degree of constancy, while the form shows a high degree of variability (Farghal 2003). 3. Translation Programs Translation programs at the undergraduate and postgraduate levels have become a common feature of Arab universities and academic institutes. This recent development is due to the increasing demand for translation practitioners on the job market. Most of these institutions were caught off-guard in terms of the availability of competent translation trainers. Therefore, the task of translation teaching was often assigned to bilingual academics specializing in literature and/or linguistics. One can find translation trainers who neither have a sufficient theoretical background in Translation Studies (ST), nor interest or motivation to familiarize themselves with ST as an adequately established sub-discipline of applied linguistics. These academics believe that their formal training in literature and/or linguistics is self-sufficient for teaching translation, which is, to them, a by-product of such training. It is sad that translation training in such contexts and with such attitudes does not go beyond anecdotal expositions. For instance, one may cite the common belief that translation activity is nothing more than using a bilingual dictionary effectively. To draw on one interesting incident, the chairperson of an English department where an MA translation program is run once assertively banned the use of dictionaries by students sitting for the Comprehensive Examination. He was wondering what would be left of the test if the examinees were allowed to use dictionaries. In addition to the serious lack of competent translation trainers, many of the students admitted to translation programs do not possess adequate language competence in the foreign language (predominantly English), let alone competence in their first language (Arabic). This bitter reality turns most translation courses at Arab universities into language rather translation courses proper. While it is true that translation activity is a sophisticated linguistic exercise that can sharpen one’s language skills in the foreign as well as the native language, adequate language proficiency in the relevant language pair is an indispensable requirement. This requirement cannot be taken for granted based on possession of high school and/or university certification relevant to language skills in the language pair. Based on my personal experience, many translation students (both undergraduates and postgraduates) do not demonstrate adequate English language competence that can live up to the taxing requisites of translation activity. Still worse, some even lack such language competence in their native language (Arabic). One should note that translation activity presents constraints and complications that may not occur in intralingual communication. For example, the high degree of flexibility and freedom available to a student when he writes in English or Arabic is tremendously reduced when engaging in translation between the two languages, due to the formal and semantic bond/contract emerging between the original and the 11 translation product. Consequently, translation programs should base their selection of entrants on entrance examinations that gauge translational competence in the language pair rather than decisions that refer to general language proficiency and/or certification alone. 4. Theory of Translating vs. Theory of Translation To many skeptics, the need for translation theory/theories in translation training is far from being clear. The familiar argument is that, until recently, most competent translation practitioners had never received any type of formal or academic instruction in translation studies. While such a polemic is generally valid, it does not negate the presence of theory in translation activity, at least at the psycho-cognitive level. In other words, the competent practitioner who has not engaged in any kind of formal training progressively develops a set of translation strategies that are subconsciously activated when translating. For example, when encountering a proverbial or an idiomatic expression, he first looks for a corresponding expression in the TL. Only after failing to access one will he opt for rendering sense independently of phraseology. Most importantly, therefore, we need to draw a key distinction between a theory of translating and a theory of translation. First, a theory of translating is essentially subconscious; it consists of a set of practical principles and guidelines which are intuitively implemented in translation practice by practitioners on the market. By contrast, a theory of translation is conscious; it consists of a set of theoretical or abstract principles and guidelines which are formally learned and consciously applied by translators. Second, while a theory of translating is naturally acquired through extensive translation activity wherein the set of principles and guidelines reaches a high degree of automatization in finished translators, a theory of translation is formally learned through exposure to or instruction in ST wherein theoretical claims are tested against naturally occurring or concocted translational data. Thus, a theory of translating is subconscious, intuitive and naturally acquired, whereas a theory of translation is conscious, informed and formally acquired. To give an example, House’s (1981, 2000) important distinction between a covert and an overt translation is part of a theory of translation, while the formally uninformed practitioner’s intuition that a translation may be reader-oriented or text-oriented is the output of a theory of translating. To make the distinction more down-to-earth, an analogy can be drawn between language competence (Chomsky 1964: Hymes 1972; Canale 1983) and translation competence (2000).(1) Native speakers of human languages gradually develop sufficient competence in their languages which enables them to use language effectively prior to engaging in any form of formal training. Similarly, translation practitioners gradually develop sufficient translational competence through extensive translation activity. In both cases, a theory of x-ing (that is, communicating and translating respectively) is subconsciously developed. A native speaker can readily judge the linguistic and social well-formedness of sentences and utterances in various contexts. By the 12 same token, a translation practitioner can readily judge the contextual fitness and naturalness of translations. The intuitive knowledge developed by both native speakers and translators through natural exposure to communicating and translating respectively is subject to further refinement and systematization by formal training and instruction, e.g. language, linguistics and translation classes. Hence, a native speaker who has access to formal instruction in language and/or linguistics will develop, in addition to his subconscious theory of communicating, a conscious theory of communication. Similarly, a translation practitioner who has access to formal instruction in ST will develop, in addition to his subconscious theory of translating, a conscious theory of translation. One should note that asking generalists in linguistics and/or literature to teach translation courses is similar, based on our analogy above, to asking a layman native speaker to teach language courses. I am quite certain that most, if not all, of those specializing in language and/or literature would object strongly to the assignment in the latter case, but only very few would question the assignment in the former case. This unfortunate attitude may be attributed to the common view that translation competence alone (i.e. a theory of translating) is all that is needed for the teaching of translation courses, whereas, rightly in this case, language competence alone (i.e. a theory of communicating) is far from being sufficient for teaching language courses. Consequently, scholars working within ST should struggle hard to convince other fellow scholars that a theory of translation is indispensable and that it is not even enough to be a finished translator, let alone an amateur one, when it comes to giving formal instruction in translation classes. Only then will translation courses build their own legitimate reality. Furthermore, theory/theories of translation alone cannot produce competent translators because an adequate translation competence ought to be taken as a point of departure for formal instruction in ST. The role of translation theory is intended to refine and sharpen the already existing level of translating theory by bringing to consciousness a set of strategies and principles in practicing and/or prospective translators. In this case, the practicing/prospective translator is expected to work with many theoretical options whose practical application manifests itself in a translational decision, which is, in the presence of a theory of translation, both practically and theoretically motivated. In this way, translation theory aims to perfect translation competence rather than create it. In fact, translation theory without translation competence (i.e. practical experience) may be described as blank, while translation competence without translation theory may be described as blind. The importance of translation theory/theories here may be likened to the importance of a latent course of study in mechanical engineering for a practicing mechanic whose entire career derives from his practical experience in difference garages. There is no doubt that our friend will be a better mechanic, despite the fact that it was only a matter of ‘Better late than never’. 5. Translation as Question of Relevance 13 The notion of relevance is introduced as a major parameter of human communication (Grice 1975; Sperber and Wilson 1981; Gust 1996, and Farghal 2004, 2012, among others). Translation, being a form of communication, can be convincingly argued to be a question of relevance. This means that what is supposed to be relayed from the SL into the TL is what is contextually relevant. The general implication here is that a textual and/or discourse segment which is relevant in one context may not be relevant in another. By way of illustration, the phraseology ‘the Custodian of the two Holy Mosques' in reference to the Saudi monarch is essentially relevant to the discourse employed by Radio Riyadh, whereas it is completely irrelevant in a BBC news bulletin where ‘King Abdullah of Saudi Arabia’ or just ‘the Saudi king/monarch’ will be most appropriate. Most frequently, the question of relevance arises in the context of choosing between form and function in the process of translating. It is the translator’s job to decide whether both form and function are relevant or only one of them is relevant in any given translational decision. Translational questions relating to form and function are assessed and resolved in light of three contextual factors, i.e. text type, audience and author. To deem one contextual factor more relevant than the others will show in translational options. For example, the Arabic cognate accusative is a textual feature of Arabic whose formal relevance when translating into English is very low (e.g. compare ‘We discussed the plan in a detailed discussion’ with ‘We discussed the plan in great detail’). Nonetheless, considering the cognate accusative a relevant feature, many translators of the Holy Quran relay this feature formally into English. M. Pickthall offers ‘Therefore we grasped them with the grasp of the mighty, the powerful’ and M. Khan and T. Hillali give ‘We seized them with a seizure of the all mighty, all capable to carry out what He will’ as renditions of the Quranic verse [ فأخكهم اخكه عز كز دق ك رSo he took them with able mighty taking]. Clearly, the authoritativeness and sanctity of the text in question has motivated these translators to consider the Arabic cognate accusative as formally relevant, despite its failing to achieve a good degree of naturalness in English. Sometimes, the question of relevance is guided by the norms of naturalness in the TL, i.e. what is relevant is what sounds natural and acceptable. This means that the audience assumes special importance in terms of relevance. By way of illustration, P. Stewart (1981) considers the mention of ‘the Prophet’ in the Arabic welcoming formula رار كا الن ك،ً أمك،ً[ أمكwelcome, welcome the prophet visited us] in his translation (Children of Gebelawi) of Najeeb Mahfouz’s (1959) Awlad Haritna irrelevant and, consequently, renders it as ‘Welcome! This is a great honor’. Had Stewart deemed the Arabic metaphor in this formula relevant, i.e. by translating it into ‘Welcome! The Prophet visited us’ instead of the rendition above, he would have twisted the implication of intimacy and sincerity in Arabic to that of sarcasm in English, in addition to the low degree of processability of his translation by English native speakers. So, again relevance presents itself as a robust maxim in translation practice. 14 In some cases, the translator’s preoccupation with SL cultural considerations may blur interlingual communication. This occurs when the translator is bent on adopting SL phraseologies at the expense of TL naturalness. Situations of this kind may give rise to communication breakdowns because the discrepancy in relevance between the SL and TL is too great to be worked out on the basis of universal principles. To cite an illustrative example, witness how P. Theroux’s (1987) translation of the Arabic proverb [ العكي صيكي واليك صيكيThe eye sees and the hand is short] in Abdurrahman Munif’s novel mudini-l-malḥ: taqaasiim al-layl wa-n-nahaar ‘Cities of salt: Variations on Day and Night’ into ‘The eye sees far but the hand is short’ and ‘Sight is long but our hand is short’. Regardless of any role that the context may play in improvising a potential interpretation of the English renditions above, one may be able to argue that, at best, these renditions are hard-going and, at worst, incomprehensible by native English speakers. By contrast, considering relevance in light of TL norms would lead to renditions like ‘The reach falls short of the desires’ or ‘The spirit is willing but the flesh is weak’. In this case, the Arabic metaphor is rightly considered an irrelevant formal feature. Finally, the issue of relevance should be related to lexical and referential voids between languages (Rabin 1958; Ivir 1977; and Dagut 1981). In order to deal with translation voids properly, the translator should decide the relevance of gaps in terms of incidental/casual mention versus planned/instrumental mention. While the former does not affect the discourse of the text in question, the latter does so to a great extent. On the one hand, the Arabic religious term الزككاmay incidentally occur in a work of fiction where the technical details of this term are completely irrelevant. Consequently, the translator may relevantly opt for an English cultural substitute (Larson 1982), e.g. ‘charity’ or ‘almsgiving’ in translation. On the other hand, the same term may occur in a religious text where the exact technical details of the term (e.g. the fact that الزككاis compulsory and is strictly quantified in Islam) are relevant. In this case, one should have recourse to other translation strategies (e.g. descriptive translation, transliteration, footnoting, lexical creation, etc.) to bring out relevant details because cultural approximation falters (for more details about translation strategies, see Ivir 1991). 6. Translation as an Act of Communicating Many specialists (or pseudo-specialists) in translation studies and neighboring areas often raise the issue of untranslatability and assertively make it a central point in their discussions and expositions. They claim repeatedly that untranslatability is a major, if not a fatal drawback in translation practice and, subsequently, employ it as an escape-hatch to avoid serious scrutiny and analysis. Their argument usually overlooks the fact that total communication, whether it belongs to intralingual or interligual communication, is a mere desideratum. Thus, when one attempts communicating a spoken or a written message in his own language, he performs the task a varying degrees of success and/or failure. This being the case, the deficit is expected to be greater in translation because it is ‘second-hand’ rather than ‘firsthand’ communication. This inherent quality of both forms of communication should be taken for granted and should never pervade polemics in translation circles. 15 Translation, therefore, needs to be viewed as an act of communicating in its own right. The translator should never lose sight of the fact that he is communicating a message from one language into another. The success of the translation product depends entirely on how meaningful and communicative it is in the TL context. In many cases, translations establish their own usefulness and acceptability independently of the originals. In point of fact, real-life situations involve either the original or the translation, but rarely both. The search for the original and the translation at the same time is predominantly an academic and/or scholarly matter. Even when translation activity is dealt with academically, the translation critic should always bear in mind that translating is not a static but rather a dynamic act of communicating. In this way, priorities in translation practice are supposed to differ from one context to another depending on the skopos of any given translation (Vermeer 2000 and Schäffner 2003). Most importantly, one should remember that an SL text is potentially capable of receiving more than one workable translation. The differences between the TL versions and the SL text may range from linguistic to interpretative features. Comparing translations of the same text with one another should be communication-oriented, that is, the translation critic ought to be aware of the questions of priority and relevance when pitting one translation against another. In the final analysis, it is not a matter of rejecting one translation in favor of another but rather a matter of explaining why translators may have different options in a variety of contexts that are diachronically and synchronically juxtaposed. In this regard, an important distinction is drawn between a translation mistake and a translation error (Pym 1992). A translation mistake may be viewed as a translational decision that cannot be borne out in terms of priority and relevance, whereas a translation error may be regarded as a communicatively-motivated translational option, despite the availability of another/other option(s) that may fare better than the one opted for. In other words, translation mistakes operate within the dichotomy of right or wrong, while translation errors maneuver within a multiplicity of potential versions. A final point in the context of translating as act of communication pertains specifically to practical training in English into Arabic translation. The fact that many Arab translator trainers still think of Arabic in prescriptive terms gives rise to dogmatic arguments regarding lexis and phraseology in Arabic translations (TL texts). Such arguments often ignore the reality that language is a living organism which changes over time and that that translation is an act of communication where the linguistic code functions as a mere carrier of content in translation. Empty arguments over whether translators can use expressions such as عكال الجكوو نك،لعكب وورا دك الةقك تحكت السكً اسك الجليك،[جسكوراplay a role, high quality, build bridges of confidence, under arms, break the ice] and a plethora of other expressions do not get us anywhere. Such expressions have become part of the linguistic repertoire of all educated Arabs (for more on this, see Darwish 2005, who is an example par excellence of this category). It goes without saying that when languages come in contact, they impact one another tremendously in terms of lexis and phraseology, with a bias in the direction of more influential languages, such as English these days. To cite another interesting incident in this respect, I was struck to hear from some students 16 that their translation teacher insisted on having [ وار الخيالكhouse of images] as the only equivalent to ‘cinema’, which a familiar borrowing in Arabic, i.e. السكيناا. One could be creative enough to imagine how an Arabic native speaker would economically tell his interlocutor that ‘he had a flat tyre/puncture’ in Arabic without employing the English borrowing صنشك. It should be made clear to students of translation that borrowing is a legitimate and natural word formation process in human languages, Arabic being no exception. This important process manifests itself in two forms: loan words, e.g. [ راو ككو كا يككوت ص لاككان و اق اطيكك فيز ككاdemocracy, radio, computer, parliament, physics], etc. and loan translations, e.g. ًاطحك وكحا الحك ال كارو ا قك ااوكو،دكه ا [ أصكيradio, computer, skyscraper, the cold war, a white coup], and so on. Both categories of borrowings have become an indispensable component of the Arab translator’s linguistic repertoire which cannot be simply erased by dictates that are completely based on illusions. In point of fact, the sophistry associated with such matters does more harm than good, if any, to translator training which, in the final analysis, aims to drive home the fact that translating is communicating. 7. Translation as a Multistage Process It is not uncommon for some teachers and many students to think of translation as a one-stageprocess which starts with translating the first segment of a text, be it a word, a phrase, a sentence, or a paragraph and ends with rendering the last segment. In this way, translation is viewed as a mechanical exercise involving the transfer of meaning between two languages in small, successive doses. The lack of dynamism in this orientation may result in many translational mishaps such as disconnectedness, unnaturalness, and, at worst, communication breakdowns, among other things. To overcome problems like these, translation activity needs to be regarded as a multi-stage process encompassing three integrated phases: pre-translating, translating, and retranslating. The pre-translating stage is preparatory before pen is put to paper to translate proper. It aims to secure a good understanding of the SL text, be it a news report, an editorial, a legal document, a poem, a novel, or any other type of text and tune oneself with the atmosphere of the text in order to establish a linguistic and cognitive rapport with the discourse in question. This phase is oriented toward translation rather than an ordinary reading situation. Therefore, the translator is required to provide meticulous interlinear notes which are meant to facilitate his work at the second stage. This exploratory mission ranges between moderately easy tasks, e.g. the comprehension of a news report to highly challenging ones, e.g. the unravelling of symbolism in a poem. During this stage, the translator should be forming, abandoning, and re-forming translational hypotheses along the way. For instance, a translational hypothesis relating to the title of a newspaper commentary may be re-formed or even abandoned after reading the first paragraph. Witness how the Kuwaiti newspaper commentary title [ الكواو طكال وصكوThe boy takes 17 after his father] (Al-Watan 2006) may initially lend itself to the translational hypothesis embracing the rendition 'Like father like son'. Only after reading the first paragraph will the translator abandon this hypothesis in favor of one that supports the polemic that the sons born to supposedly Kuwaiti fathers and non-Kuwaiti mothers may take after anyone but their presumed fathers. Thus, a rendition such as 'Like son like mother' or even 'Like son like neighbor' would be needed in order to reflect the content of the commentary whose title ironically tells a different story. Similarly, a hypothesis relating to the translation of a symbolic title of a novel may undergo numerous reformulations along the way before a sound settlement is adopted. Whatever the case is, a good understanding of the SL text remains the first milestone of translation process. Other things being equal, it can be argued that good comprehension begets good translation. The second stage (the translating stage) constitutes the cornerstone in translation activity as it involves the re-encoding of the SL material by phrasing out the source text's meaning/message in TL semiotic signs. At this stage, the translator engages in intensive decision making regarding form and content and, subsequently, the type of equivalence/ resemblance settled for, a process which is always informed by contextual factors including text-type, audience and author. Thus, the notion of equivalence/resemblance, which may be theoretically motivated, becomes a correlative of context. Needless to say, language competence (transfer competence in particular), cultural competence and schematic competence play a pivotal role in producing a workable TL version during the execution of the multi-faceted task at this stage. Lastly, we have the retranslating stage where the translator goes over the entire TL text in search of small corrections and refinements here and there. These may range from simple amendments relating to grammar and diction to more subtle ones pertaining to textuality and discourse. Regardless how competent the translator is, it can be argued that the retranslating stage is essential because it inevitably renders the translation a better one at, of course, varying degrees, depending on the quality of work at the second stage and the level of translation competence on the translator's part. The amendments made at this stage may be thought of as the final touches added to different human states of affair – touches which, though cosmetic in the main, may prove indispensable in the translation profession. 7. Conclusion This article shows that the training of student translators should start with addressing the nature of the raw material of translation activity, i.e. language, by bringing out the fact that human communication is realized by operating two complementary principles: the open principle and the idiom principle. The twinning of these two principles forms the basis for the possibility of offering more than one good translation of the same SL text. It also shows that translator trainer programs at Arab universities still regard translation studies as derivative rather than a discipline in its own right. This erroneous belief has led to giving the 18 assignment of teaching translation courses to generalists in linguistics and/or literature who have no interest in translation studies beyond being bilingual in Arabic and English. To remedy this serious problem, we should make sure that translator trainers possess an adequate knowledge of translation studies before they are entrusted with teaching translation courses. In particular, an important distinction is drawn between a theory translating and a theory of translation. While we explain how a theory of translation is necessary, such a theory is argued to functionalize and perfect translational competence rather that create it. Equally important, the article argues that translation activity is essentially a question of relevance and priority. Thus, contextual factors are of paramount importance when it comes to deciding what is relevant and what is not. Regardless of differing translational decisions along the way, the fitness of a translation is gauged against a principle of communicativeness whereby translation is viewed as an act of communicating rather than an act of prescribing. Thus, translation mistakes, which are described in terms of right or wrong, are differentiated from translation errors, which are critically analyzed in terms potential TL versions. Finally, it is shown that translation activity is a multi-stage rather than a one-stage process. While the translating stage constitutes the backbone of the process, the pre-translating and the retranslating stages are argued to be integral to the process if cohesion and coherence are to be catered to optimally in the translation. It is of utmost importance, therefore, to introduce this procedural parameter into student translator training. The Linguistics of Translation Mohammed Farghal & Ali Almanna Abstract The paper demonstrates through the use of ample textual data that translation involves significant decision-making at different linguistic levels, including phonology, morphology, syntax, and semantics. The translator's awareness of the linguistic mismatches in the language pair constitutes a foundation stone in his work. Hence, this study discusses various strategies of handling linguistic parameters in the hope of bringing them into the consciousness of practicing translators, as well as translation teachers. 19 1. Overview Despite the fact that human languages share general rules in the sense of Chomsky's universal grammar, it remains true that parametric variation between languages involves a lot of mismatches at the different levels of linguistic description. In this respect, Farghal (2012) holds that languages phonologize, morphologize, syntacticize, lexicalize, phraseologize differently within general parameters. This fact rightly motivated Jakobson (1959) to say that translation between languages is a matter of replacing messages in one language with messages in another without getting trapped by surface linguistic features. Krazeszowki (1971:37-48) argues that there are few, if any, congruent structures between languages. One-to-one strict correspondence is, therefore, the exception rather than the rule in translation. In most cases, the translator is confronted with one-to-many or many-to-one correspondences while working with any language pair. Despite the numerous linguistic mismatches between languages, Kachru, (1982:84) claims: “Whatever can be said in one language can be said equally well in any other language”. While Kachru’s statement may be true in a qualified manner, we believe that the disparities between languages are a matter of asymmetric equivalence or resemblance. In this way, similarity can be detected within difference. Newmark (1991: 8) stresses that due to differences in frequency, usage, connotation and the like, the meaning of any lexical item in Language A cannot be identical to that in language B. Such linguistic differences at lexical or phrasal level, for instance, prompt translators to adopt certain strategies to minimize such 'linguistic inequivalences' (Al-Masri 2004: 74). This is in line with Hatim and Mason (1990: 23) who highlight that “translation involves overcoming the contrasts between language systems: SL syntactic structures had to be exchanged for TL structures; lexical items from each language had to be matched and the nearest equivalents selected”. Translators, being charged with such constraints imposed on them by virtue of the differences between the linguistic systems of the interfacing languages, i.e. the lack of a one-toone relationship between lexical and grammatical categories, opt for different strategies, such as addition, omission, paraphrasing, elaboration, adaptation and so on. When discussing linguistic and/or textual considerations in translation activity, one needs to distinguish between obligatory features and optional features. On the one hand, obligatory features involve choices that must be followed by the translator in order to satisfy the rules imposed by the TL system, without which the translation will be ungrammatical. Optional features, on the other hand, represent cases where the translator can exercise real choice by deciding on one translation option rather than another/others. By way of illustration, let us consider the following English sentence along with its Arabic translation: 20 (1) The two black boys quarreled while they were playing in the narrow alley. .) تشاجر الصبيان الزنجيان وهما يلعبان في الزقاق الضيق2( [quarreled the black(dual) boys(dual) and they(dual) were playing(dual) in the narrow alley] Examining the Arabic translation in (2), we can readily see that the translator implemented four obligatory features, viz. using the dual form ( الصببيانfor the two boys), marking an adjective, a pronoun and a verb (يلعببان/همبا/ )الزنجيبانfor the dual number, and marking the adjectives for definiteness (الضبيق/)الزنجيبان. Here the translator has no choice but to follow these adjustments because they are imposed by the language system in Arabic. The violation of any obligatory feature would produce broken or 'pidgin' Arabic. One should note that obligatory features such as these are taken for granted as part of language competence, hence not deserving any further discussion in translation activity. In contrast, it is in the domain of optional features that translators exercise decision-making and flexible choice. That is why translation criticism flourishes in this area apart from obligatory features. To look again at the translation in (2), one can imagine other linguistic options that could have been followed, albeit subject to criticism, as can be illustrated in (3) below: (3) a. الصبيان الزنجيان تشاجرا وهما يلعبان في الزقاق الضيق. [The two black boys quarreled and they were playing in the narrow alley] b. تشاجر الصبيان الزنجيان عندما كانا يلعبان في الزقاق الضيق. [The two black boys quarreled when they were playing in the narrow alley] c . تعارك الولدان الزنجيان بينما كانا يلهوان في الشارع الضيق. [The two black boys fell out while they were loitering in the narrow street] As can be seen, each of the choices in (3) follows a linguistic option which is different from the one adopted in (2). The first rendering (3a) changes the word order from Verb-Subject to Subject-Verb while maintaining the choice of conjunction (coordination) and lexis. The second rendering maintains the word order and lexis while changing the conjunction into subordination. For its turn, the last rendering (3c) extensively changes the lexis ( الولدانfor الصبيان, يلهبوانfor يلعببان, and الشببارعfor )الزقبباقwhile preserving the word order and the choice of the category of conjunction, i.e. subordination albeit a different subordinator (' عندماwhen' vs. ' بينماwhile'). 21 Let us now look at an authentic example from Hemingway's The Old Man and the Sea (1952), along with its Arabic translation in Ba‘albaki’s ( الشيخ والبحر1985): (4) The boy was sad too and we begged her [the fish] pardon and butchered her promptly. ) ولقد ران الحزن على الغالم أيضا ً فالتمسنا من السمكة القتيل5( .العفو والمغفرة ونحرناها [And the sadness overwhelmed the boy, so we begged pardon from the killed fish and slaughtered it] Ba‘albaki, as can be observed, has followed some optional decisions. Firstly, he has rightly changed the word order in order to offer an unmarked structure comparable to the English one. Secondly, he has decided to elevate the style in Arabic by choosing highly formal lexis, viz. ران, الغبالم, فالتمسببنا, and نحرناهبا, thus altering Hemingway’s simple narrative into stilted narrative. Thirdly, the translator has decided to employ an Arabic synonymous lexical couplet, viz. العفبو [ والمغفبرةpardon and forgiveness] in an attempt to offer more natural discourse. Finally, he has committed two lexical errors, viz. using the adjective ' القتيبلthe killed' to post-modify the fish and ّ 'chopped' when employing the Arabic verb ' نحبرslaughtered' instead of the correct verb قطب referring to the fish as if it were a sheep or a camel. In fact, fish are not slaughtered the way other animals are; they are just taken out of water before they undergo chopping or anything else, nor are they killed like other animals. Below is a suggested translation that takes care of these critical points: ّ .وقطعناها على عجل ) ولقد شعر الصبي أيضا ً بالحزن فطلبنا من السمكة العفو6( [The boy was sad too, so we begged pardon from the fish and chopped it promptly] It is within the bounds of these translation options that the translation critic can exercise his/her profession by showing how and why one option is preferable to the other options. In the rest of this section, we will look at translation options relating to different linguistic levels, namely phonology, morphology, syntax, and semantics. 2. Phonological Features 22 Phonological features become an important aspect of translation when form comes to the fore in discourse and presents itself as inseparable from content. This is where phonological features emerge as part and parcel of content that need to be taken care of by the translator. The clearest manifestation of phonological features occurs in poetry (e.g. alliteration, rhyme, meter, paralleled repetition, etc.) where defamiliarization and the creation of new paradigms are embodied in such features (Fowler 1996). Hence, translating verse into verse is the most challenging task in translation; it may require, as many believe, a poet translator in order to render the formal properties that improvise poeticness which legitimates the discourse in this genre. A comparison between a verse rendering and a prose rendering of a Shakespearean sonnet is a case in point (Farghal 2012: 208-209) (7) Like as the waves make towards the pebbled shore So do our minutes hasten to their end; Each changing place with that which goes before, In sequent toil all forwards do contend. ) كما األمواج تتجه نحو الشاطئ المهيب8( تتسارع الدقائق في عمرنا نحو المغيب تتبادل األدوار في نسق وتوال عجيب .جاهدة نحو هدفها في تنافس عصيب [Like the waves heading for the awesome shore Minutes hasten in our age towards sunset They exchange roles in wondrous pattern and consecution Toiling towards their target in adverse competition] ) كما األمواج تتجه نحو الشاطئ9( ذي الحصى ال صائق ف عا ا حو اا اا،ت سار كل ت اول الااان د ال تس قاا جام حو األدام ف تنافس اقيق [Like the waves heading towards the pebbled shore Minutes hasten in our age towards their end Each exchanges the place with the one before it Toiling towards the front in true competition] It is true that the prose translation in (9) is more reflective of the content of (7), but it is seriously lacking in poeticness because it ignores phonological features, namely rhyme and meter. When compared with the translation in (8), which differs in small ways as to content in (7) while keeping the same thematic thread, one can appreciate the discrepancy between the two. It is the phonological features that qualify (8) as poetic discourse on the one hand and (9) as commonplace discourse on the other. The mere layout of material in poetry translation would in no way make up for improvising key phonological features. 23 In fact, it is a trade-off between form and content, where form needs to be given priority in poetry translation. One should note that poeticness is a matter of degree in human languages; it is not an allor-nothing phenomenon. Different discourses manifest different degrees of poeticness and, apart from literary discourse, everyday language is full of figurative expressions where phonological features usually occupy a position. For example, such features play a key role in the creation of proverbs which mirror social life in different cultures. These proverbs often function as background for the formation of remodeled expressions (for more details, see Farghal and Al-Hamly 2005). By way of illustration, consider the two remodelings below: (10) A smile a day keeps misery away. (twitter) (11) A laugh a day keeps the doctor away. (Daily Strength/Cyndi Sarnoff-Ross, Oct. 21, 2011) Both remodelings, as can be observed, fall back on the familiar English proverb 'An apple a day keeps the doctor away' in order to communicate fresh messages that have nothing to do with food as such. The tendency for investing existing phraseologies in the creative formation of new ones is mainly motivated by a desire to bring phonological features to spotlight in order to consolidate the message and make it more appealing to the audience. For instance, being a psychotherapist, Sarnoff-Ross in (11) above has succeeded in choosing a title that functions as a semiotic sign which summarizes her entire article. From a translational perspective, the translator needs to fall back on his cultural heritage in order to find a rhythmic phraseology or to create his/her own remodeling which dwells on a similar theme. In this case, a creative translator would offer a title like اضكحاوا تيكحوا 'Laughing makes you healthy', thus remodeling the familiar Prophet Mohammed’s hadith (saying) ' صكودوا تيكحواFasting makes you healthy'. A commonplace title like أمايك الحكح ' لليكحThe importance of laughter for health' would be far less effective and appealing. Again, it is the phonological features that make the difference. Last, phonological features present themselves as a significant issue in borrowing and transliteration, which are important translation strategies. Borrowing, which is a key translation strategy from English into Arabic, manifests itself in two forms: loan words and loan translations. While phonological features are not relevant to loan translations where the concept of the word is borrowed independently of the form (e.g. ،' دكه اradio' and ' ااوكوcomputer'), they are at the heart of the process of loan words where both the form and concept of the word is borrowed. This necessitates taking account of phonetic gaps between English and Arabic when naturalizing a word, e.g. replacing a vowel with another or a consonant with another, viz. راو كوfor 'radio' and كا يكوتfor 'computer'. In some cases, the process is not straightforward, that is, the English sound may be replaced by more than one sound depending on the Arab region. For example, the /g/ sound may variously be replaced with /غ/, /ج/, or /ك/ (/ɤ/, /j/, /k/, respectively). Thus, in terms of phonological representation there may be cases where there are competing forms, e.g. the two authors of this article used different forms of a recurrent 24 word while recently editing a book in Arabic about translation. Subsequently, they had to negotiate the issue and finally settled for ‘ اللغك اإل جليز كthe English Language’ rather than ‘ اللغك اإل اليز كthe English Language’. Sometimes, familiarity and frequency may override well-established norms. For example, Farghal (2011), when translating a Croatian novel 'The Ministry of Pain' by D. Ggresič (2008) from English into Arabic, decided to render the recurrent name Goran as صكوران, being aware of the familiarity of this name in the Arab media among sport circles, thanks to Goran Invaniševič, the well-known Croatian professional tennis player. Surprisingly, however, the reviewer and/or commissioner changed the said name to جكوران, the one which now appears in the published translation without consulting the translator, hence the importance of opening a dialogue between those in charge of translation quality control and the translator (for more details, see Almanna 2013). Competing phonological representations may also involve ideological moves (Farghal 2010; Farghal 2012; Farghal and Al-Manna 2014). Historically, most Christian names designating places or personalities in the Arab Middle East receive Anglicized phonological representations that now compete with more transliteration-oriented representations. For example, the choice between Al-Khalil and Hebron or Al-Quds and Jerusalem may be instigated by the ideology of the translator. One can also notice a tendency to avoid the originally Greek and later on Anglicized phonological representations of names of Arab Muslim medieval scholars such as Averroes and Avicenna in favor of more phonologically faithful forms, viz. Ibn Rushd and Ibn Sina, which may carry ideological moves. When it comes to rendering Arabic proper names into English nowadays, the tendency is to transliterate them by sometimes simplifying phonetic gaps, e.g. Ali, Tareq, and Amman and sometimes preserving them, e.g. Khalid and Dhafir. Likewise, some English names are adjusted phonologically such as دكارfor 'Mary' and صي كfor 'Peter' and some maintain the same pronunciation such as جكونfor 'John' and وكاfor 'Sandy'. In few cases, one might find domesticated phonological representations that occur in the translation of some literary genres (mainly in dramas) such as دfor 'Mary' and صط سfor 'Peter'. 3. Morphological/Word-formation Features English and Arabic represent two contrasting morphologies. While English morphology is predominantly analytic, Arabic morphology is largely synthetic. To explain, an English word like writers can be readily analyzed into a root, the doer morpheme and the plural morpheme, whereas the corresponding Arabic words ' كاتكبwriter' and ' ك ّكاbook' do not lend themselves to such a linear analysis, viz. the doer morpheme and the plural morpheme consist of vowel changes within the abstract triconsonantal root /ktb/ 'a prelexicalized form that has to do with writing', which becomes /kaatib/ 'writer' and /kuttaab/ 'writers', respectively in this case. In the two morphologies, the root functions differently. In English derivation, the root functions as input for prefixes and suffixes which may change word class, e.g. rewrite, writer, writing, written, writable, etc. In Arabic, by contrast, the root functions as input for semantically related verbs which in turn function as input for other derivation processes, ' ك ككبHe wrote', ' كاتككبHe corresponded with', ' ك ّكبHe dictated', ' اك كبHe underwrote', ' اوك ا بHe asked to write', etc. Each of these semantically related verbs can be input for other derivation processes, 25 e.g. from اوك ا بwe can derive ' دسك ا ببthe one who asked to write', ' دسك ا بthe one who was asked to write', ' او ا اasking to write', etc. In terms of translation, most semantically related Arabic verbs would usually require morphologically unrelated verbs. By way of illustration, consider the following examples, along with their English renderings: )12( a. b. c. d. ٌ قتل عل .ي جنديين في المعركة .قاتل عل ٌي في معارك عدة ٌ تقاتل عل .ي م أناس عدة .استقتل عل ٌي للحصول على الوظيفة Ali killed two soldiers in the battle. Ali fought in several battles. Ali quarreled with several people. Ali made every effort to get the job. As can seen in (12), the four Arabic verbs that are derived from the same root require different renditions in English. This morphological difference may cause problems to translators, as can be illustrated in the authentic example below: ) وكاا كان جا الياصان اوص ياو صع الح العالاي الةا ي دةاو صو ا اص ت صه وول أخ ى ف ش ق آويا فإن جااا31( (The Arabic Newsweek, February 4, 2003) . دع و ف الع اق اا أن ش د واع الايلحي ف الانطق [As the Japanese economic success after the Second World War was a solid example copied by other countries in East Asia, so an average success in Iraq may strengthen the arm of reformers in the region] Apparently, the translator has confused the two semantically related Arabic words ' الايكلحيreformers' and ' اإلصكًاييliberals' (which share the same root) when rendering the word 'liberals' in the English ST. This confusion has skewed the coherence of the text, that is, the Arabic translation incoherently talks about 'social reformers' instead of 'liberal politicians' in a political context. For its turn, English prefixal and suffixal derivation may present some challenges to terminologists and translators. Notice, by way of illustration, how English morphology can readily account for fine semantic distinctions via suffixation, e.g. legitimacy vs. legitimization and secularism vs. secularization. While it is usually easy to find Arabic corresponding terms for the English nouns designating states, viz. ّ الشك عيfor 'legitimacy' and العلاا ي كfor 'secularism', it is more challenging to lexically account for nouns designating processes, viz. ش عنfor 'legitimization' and علانfor 'secularization'. In many cases, such nouns are paraphrased into Arabic, as can be illustrated in the following example: (14) The industrialization of Europe started in the late nineteenth century. . ) ص أت عالي ال حول اليناع ف أوروصا ف أواخ الق ن ال او عش31( [The process of the industrial shift started in Europe in the late nineteenth century] For lack of an Arabic term, as can be seen, the English process noun in (14) needs to be paraphrased into three Arabic words in (15). Below are some authentic examples where the translator has opted for two strategies (deletion and paraphrase) when encountering a 26 morphological gap, namely the English -able in this case: (Khalid Hosseini’s novel The Kite Runner, 2003, translated by Manar Fayyadh, ع ا الطائ الورصي, 2010) (16) (17) (18) (19) She did blood tests for every conceivable .صادت صإج ا فحوص وم لال م دون hormone. (p.200) We Afghans are prone to a considerable degree . ح األفغان ديّالون للا الغ of exaggeration (p. 153) Baba's cancer was advanced. Inoperable. (p. غيككك صاصكككل. وكك طان صاصكككا ككككان فككك االككك د كككأخ 168) .لًو ئيال But theft was the one unforgivable sin. (p. 172) لا ك الس ك ص م ك الخطيئ ك الواي ك ال ك و اا ك .غف ا اا As can be observed, the translator has unjustifiably opted for deletion of the -able words in the translations of (16) and (17). The translator has either deemed them unimportant (which is not true) or found them problematic, so she has decided to drop them. She could have rendered them as follows: .) قامت بإجراء فحص دم لكل هرمون يمكن تصوره20( [She did a blood test for every conceivable hormone] .) نحن األفغان ميّالون لدرجة عالية من المبالغة21( [We Afghans are prone to a high degree of exaggeration] In (18) and (19), however, the translator has succeeded in paraphrasing the -able words correctly by adopting the paraphrase strategy. Inflectional morphology may also present some translational problems. To give an example relating to gender, in English a shark has a masculine gender (a 'he'), while in Arabic, being a fish, a shark has a feminine gender (a 'she'), viz. سمكة القرش. Therefore, Ba‘albaki’s translation ( الشيخ والبحر1985) of Hemingway’s 1952 novella 'The Old Man and the Sea' has rightly changed the recurrent 'he' in reference to the shark to a recurrent feminine noun السمكةor a recurrent feminine pronoun clitic. In fact, there is no natural way to maintain the masculine gender in Arabic. However, there are cases in the translation where the coherence of gender cannot be preserved, as is illustrated in the following example: (22) He [the shark] took the bait like a male and he pulled like a male ... ... وهي تشد كأنها ذكر،) لقد تناولت الطعم كأنها ذكر23( [She took the bait as if it were a male, and she is pulling as if it were a male ...] As can be seen, the gender issue causes a coherence problem, viz. while the ST talks about a male 'he' behaving like a male in eating the bait and in pulling, the TT talks about a female fish 'she' behaving as if it were a male. In this way, the ST and the TT present two different world views. One might argue that it would be more coherent in the translation to refer 'a female fish' behaving like 'a female fish' rather than as if it were 'a male fish'. This might be more congruent with the wise decision to change the 'he' to 'she' in the Arabic translation. Gender, therefore, may present itself as a problematic issue between English and Arabic because there is no oneto-one correspondence in gender specification. Nouns like teacher, nurse and translator are 27 gender underspecified in English, whereas they are gender specified in Arabic, viz. معلمة/معلم 'male/female teacher', ممرضة/' ممرضmale/female nurse and مترجمة/' مترجمmale/female translator'. The translator may go a long way in his/her translation before discovering, for instance, that the referent of a referring expression like 'John’s teacher' is a 'she' rather that a 'he'. Number marking may also present itself as a problematic matter in translation. In the prepublished version of his translation of C. McCarthy’s novel The Road (2006), Farghal (2009) decided to replace the recurrent marked dual form in the Arabic translation with the plural form. Being mainly a story about a father and his little boy, the Road makes frequent narrative use of the pronoun 'they' in reference to them. The translator, in this case, has two options: either to use the Arabic formal correspondent throughout, i.e. the marked dual form or to replace the dual form with the unmarked plural form. Farghal’s decision was to employ the dual form only in a few cases where intimacy is communicated. Otherwise, the unmarked plural form is to be used for ease of articulation and naturalness, thus giving priority to the smoothness of the flow of discourse over the grammatically prescribed form. Again, the reviewer and the commissioner, without consulting the translator, decided to awkwardly preserve all the dual forms in the published version. The dual form numbers in thousands in the translation as it is not only verbs but also nouns, pronouns, adjectives, and adverbs must have it when reference is made to the father and his little son. Below is an excerpted sentence (24) from the translation (p. 19) where there are five dual forms, which can be compared with (25), where the marked dual form is replaced with the unmarked plural form: ) تركا العربة في أخدود مغطى بالمشم وشقا طريقهما إلى أعلى المنحدر عبر جذوع األشجار الواقفة المتفحمة24( إلى مكان. حيث جلسا في ظل صخرة وراقبا حبات المطر الرمادي تنهمر عبر الوادي،فيه صخور بارزة [(They) left-dual the cart in a groove covered with the linoleum and found-dual their way-dual to the top of the slope through the standing charred trunks of trees to a place with protruding rocks, where the sat-dual in the shade of a rock and watcheddual the rain drops pouring through the valley] (25) تركوا العربة في أخدود معطى بالمشم وشقوا طريقهم إلى أعلى المنحدر عبر جذوع األشجار الواقفة حيث جلسوا في ظل صخرة وراقبوا حبات المطر الرمادي تنهمر عبر المتفحمة إلى،مكان فيه صخور بارزة .الوادي [(They) left the cart in a groove covered with the linoleum and found their way to the top of the slope through the standing charred trunks of trees to a place with protruding rocks, where the sat in the shade of a rock and watched the rain drops pouring through the valley] Given the high frequency of the dual form in (24) and in the entire translation in question for that matter, the Arab reader would not feel at ease encountering the marked dual form so frequently in the narrative and, one can argue, would feel more comfortable with it being replaced with the unmarked plural form, whose referential value is readily recoverable from the 28 novel’s macro-context, i.e. being a story about a father and his little son. Here, once more, we have inflectional morphology interfering with decision making in translation. Apart from derivation and inflection, other word formation processes may present some translation problems. For example, whereas conversion is a highly productive word formation process in present-day English, it is completely missing in Arabic where changing the part of speech of a word must involve a formal change. In many cases, English verbs resulting from conversion need to be paraphrased when rendered into Arabic, as can be illustrated in the following examples: (26) Before water is bottled for human consumption, it is thoroughly checked in highly specialized laboratories. . يتم فحصه بدقة في مختبرات متخصصة،) قبل أن يعبأ الماء في قوارير لالستهالك البشري22( [Before water is filled in bottles for human consumption, it is tested closely in specialised laboratories] (28) The first step in researching a topic nowadays is to google it. .) أول خطوة في تقصي موضوع ما في هذه األيام هو أن تبحث عنه في شبكة جوجل االلكترونية29( [The first step in researching a topic these days is to search for it in the electronic Google net] Other English word formation processes such as compounding, blending and acronymy/abbreviation may also cause some translation problems when rendering them into Arabic because Arabic is much less receptive of them than English. English technical compounds where the first syllable of the first word is usually prefixed to the complete second word to form a compound, for example, may demand a different lexicalization process in Arabic. To explain, while Arabic manages to form a few compounds when rendering cases such as electromagnetic كهرومغناطيسي, Anglo-American أنجلو أمريكيand Afro-Asiatic أفرو آسيوي, it often resorts to paraphrase in rendering compounds such as biodiversity ' التنوع البيولوجيbiological diversity', geopolitics ' الجغرافيا السياسيةpolitical geography', ecosystem ' نظام بيئيecological system' and psychoanalysis ' التحليل النفسيpsychological analysis'. English technical compounds, therefore, can be broached using two strategies in Arabic: borrowing the compound (which may involve translation as well) or paraphrasing the compound’s content (which may involve borrowing as well). The choice between the two options often depends on level of technicality and acceptability (for more on the translation of English reduced forms, see Al-Hamly and Farghal, this volume). For its turn, Arabic has a few religious initialisms that must be unpacked into full English sentences in translation. The procedure involves employing verbs featuring the most salient and/or important sounds in a phrase/sentence such as هلّلfor the act of uttering the sentence ' ال إله إال هللاThere is no god but God', كبّرfor the phrase ' هللا أكبرGod the greatest', and بسملfor the phrase ' بسم هللا الرحمن الرحيمIn the name of God, the most gracious, the most merciful', among 29 a few others. Thus, an Arabic sentence like نهض من الفراش وهلّل عندما رأى ضوء النهارneeds to be rendered as 'He rose out of bed and testified to the oneness of God when he observed the light of day' or 'He rose out of bed and said "There's no god but God" when he observed the light of day'. 4. Syntactic Features Syntactic asymmetries between Arabic and English require special attention from translators. Most importantly, the translator needs to be aware of the mismatches at the sentence level which involve word order variation. English (which relatively has a fixed word order), for example, overwhelmingly employs the unmarked 'Subject Verb Object/Complement' word order. By contrast, Arabic (which is more flexible in word order) uses the unmarked 'Verb Subject Object/Complement) word order as well as the less unmarked 'Subject Verb Object/Complement' word order, which, at face value, corresponds to the unmarked English word order. The competent translator, however, needs to dismiss this superficial correspondence as inappropriate, as the Arabic word order corresponding to English S V O/C is the V S O/C rather than the S V O/C, which coincides with the English word order. Note how Munir Ba‘albaki (1985) and Nabil Raghib (2004) in (31) and (33) below respectively are aware of this structural mismatch in their translations of Hemingway’s The Old Man and the Sea: (30) The fish just moved away slowly and the old man could not raise him an inch. .ً) لقد ابتعدت السمكة في تؤدة وعجز الشيخ عن أن يرفعها إنشا ً واحدا31( [moved away the fish slowly and the old man could not raise her one inch] (32) The old man went out the door and the boy came after him. .) خرج العجوز وتبعه الولد33( [went out the old man and followed him the boy] In some cases where prominence is sought, however, a match between the two word orders obtains. For example, when translating English newspaper headlines, the S V O/C should be maintained in Arabic. Thus, an English newspaper headline such as 'Barak Obama arrives in Damascus' translates into ' باراك أوباما يصل إلى دمشقBarak Obama arrives in Damascus' rather than ' يصل باراك أوباما إلى دمشقarrives Barak Obama in Damascus'. The competent translator, however, would switch to the V S C Arabic word order in his/her first sentence detailing the news story, viz. ... ' وصل الرئيس األمريكي باراك أوباما إلى دمشقarrived the American president Barak Obama in Damascus ...'. This functional shift between the two word orders in Arabic is very significant in translation activity. It is a syntactic means to improvise prominence through word order variation. Grammatical resources employed to achieve major semantic functions like negation and emphasis may be similar in some cases but different in others. Let us first consider negation which can be syntactically accomplished by the use of negative particles like not in English and لن/لم/ الin Arabic depending on the category of Tense. This will usually cause no difficulty for translators, e.g. the sentence 'John will not try to get a PhD' is straightforwardly rendered as لن 30 يحاول جون أن يحصل على شهادة الدكتوراه. However, notional (implied) negation involving an adverb like 'too' will be more challenging to translators who need to render the meaning of negation rather than be trapped by the form of the sentence, e.g. the negation in the sentence 'John is too old to get a PhD' should be unpacked when rendering it into Arabic, viz. لن يكون بمقدور جون أن ' يحصل على شهادة الدكتوراه بسبب تقدمه بالعمرJohn will not be able to get a PhD because he has progressed in age' or ' تقدم العمر بجون ولن يستطي الحصول على شهادة الدكتوراهJohn's age has progressed and he will not be able to get a PhD', etc. This kind of negation in English may cause problems for student translators as well as professional translators. Note the erroneous renditions of (34) and (36) in (35) and (37), which are extracted from two different published Arabic translations: (34) I think you've been too busy to notice where I've been. .) أظن أنك كنت مشغوال جدا لتالحظ أين أنا35( [(I) think that you were very busy to notice where I am] (36) ... but his hands were shaking too hard to pin it on. . لكن يديه كانتا ترتجفان بقوة لتدبيس الباقة على الفستان... )32( [... but his hands-dual were-dual shaking-dual strongly to pin the bouquet on the dress] The renditions in (35) and (37) can hardly make any sense in Arabic because they confuse implied negation with emphasis. The interpretation of the negation marker 'too' as the emphatic marker 'so' does irreparable damage to the meaning. Working from Arabic into English, the translator may also encounter several syntactic hurdles. One interesting example is the emphatic cognate accusative where an act is emphasized by deriving a masdar (present participle) from the verb predicator instead of employing an adverbial, as can be illustrated below: ّ هز الصبي الغصن ّ )32( .هزا (38) * The boy shook the branch shaking. (39) a) The boy shook the branch indeed. b) The boy did shake the branch. In terms of translation, as can be noted, the cognate accusative constitutes a grammatical gap in English (note the ungrammaticality of 38) and, consequently, it needs to be rendered as an adverbial (39a) or a grammatical emphatic marker (39b), (for more on this, see Farghal 1991, 1993a, 1993b). To observe the loss that may result from overlooking the cognate accusative in translation, let us consider the following excerpt taken from Elyas’ (1982:105) translation of N. Mahfouz’s (1923) novel ‘ اللص والكالبThe Thief and the Dogs’, along with a suggested translation (41) that maintains the role of the cognate accusative, among other things: 31 (40) My father was able to understand you. You have avoided me until I thought you were trying to get rid of me. With my own free will I came back to the atmosphere of incense and to anxiety. That's what the homeless and the deserted do. (41) My father was able to understand you. So many times did you avoid me that I thought you were dumping me indeed! With my own free will I came back to the atmosphere of incense and to anxiety. That's what the homeless and the deserted do. Note how the translator’s disregard of the exclamation (a taxing construction in this case) and the cognate accusative in the original has compromised the emotiveness of the text. The second sentence in (40) is unduly under-emotive and relatively detached when compared with its duly highly emotive and involved counterpart in (41). Unfortunately, this kind of loss can go unnoticed for long, as the inadequate translation may read smoothly and relevantly, hence the urgent need for sensitizing translators to the fact that grammar is meaning-bearing, just like lexis. Another area where there is a syntactic asymmetry that needs special attention from translators is the definite article. Both languages use the definite article referentially with plural and noncount nouns. However, only Arabic may employ it generically with both categories of nouns, in which case English must use the zero article. This mismatch may pose problems, even to the most professional translators, as can be illustrated by the translations in (43) and (44) of the Quranic verse in (43) below: ) فأرسببلنا علببيهم الطوفااان والجاارا والقماال والضاادا وال ا م آيببات مفصببالت42( فاسببتكبروا وكببانوا قوم با ً مجببرمين )133:(األعراف (43) So We sent on them: the flood, the locusts, the lice, the frogs, and the blood (as a succession of manifest signs), yet they remained arrogant, and they were of those people who were Mujrimun (criminals, polytheists, sinners, etc.). (Al-Hilali and Khan 1993) (44) So We sent down on them the flood, the locusts, the vermins, the frogs, and the blood; these were clear miracles, but they were arrogant and guilty people. (Al-Hayek 1996) 32 As can be seen, the five bold-faced nouns (3 plural count nouns and 2 non-count nouns) in (42), which all involve generic reference in the Quranic verse, are rendered erroneously as nouns involving specific reference. This comes as an immediate consequence of the translators’ not being sensitive to a syntactic asymmetry at the level of definiteness. Thus, instead of correctly using the zero article with these nouns, they employ the referential definite article. Epistemic modality, which constitutes the ways speakers view the world around them in terms of (un)certainty (Halliday 1970 and Lyons 1977), also involves mismatches between English and Arabic. In fact, one cannot assume a one-to-one correspondence between English and Arabic modal verbs. A grammatical gap may sometimes cause a translator to use an inappropriate translation correspondent. For example, the English modal verb 'must' and 'should' are bi-valent, as they can be employed deontically to express strong obligation and epistemically to express strong conjecture, whereas their formal Arabic correspondents يجببand ينبغبيmay express strong obligation only. This problematic mismatch is illustrated in the translations in (47) and (48) of the bold-faced segments in (45) and (46): (45) … They [the fish] are moving out too fast and too far. But perhaps I [the old man] will pick up a stray and perhaps my big fish is around them. My big fish must be somewhere. (The Old Man and the Sea). (46) I wonder what he [the fish] made that lurch for, he thought. The wire must have slipped on the great hill of his back. (The Old man and the Sea) )42( .إن سمكتي الكبيرة يجب أن تكون في مكان ما [Verily my big fish has to be somewhere] ) ينبغي أن يكون الشص المعدني قد انزلق فوق ظهرها الشبيه بالجبل48(. (Ba‘albakī 1985) [The metal wire has to have slipped on her back (which is) like a mountain] Ba‘albaki’s (1985) translations in (42) and (48) erroneously express the fish’s obligation to be somewhere and the wire’s obligation to have slipped on the fish’s back respectively. In both cases, however, we have epistemic modality expressing a strong conjecture/possibility. To communicate the intended epistemic readings, the translator should have employed the modalized verb ال ب ّد, which can, in contrast to يجبand ينبغي, be used to convey both epistemic 33 and deontic modality in light of the context it occurs in. In this way, what is a bi-valent modal English verb (must) corresponds to two different modalized verbs in Arabic, i.e. ينبغي/ يجبvs. ال ب ّد, depending on whether the modality is deontic or epistemic respectively. In some cases, what is a predominantly structure-based pattern in the SL may turn out to be a mainly semantics-based pattern in the TL. A good example here is English basic passive structures which lend themselves to translating into many Arabic alternatives including basic passive structures, basic active structures, nominalization, passive participles, and active participles. Therefore, the general claim that an English basic passive structure needs to be translated into an Arabic basic active structure (Al-Najjar 1984; Mouakket 1986; Saraireh 1990; Farghal 1991; Khalil 1993; El-Yasin 1996) accounts for only one translation alternative among many (Farghal 1996; Khafaji 1996). Following are some illustrative examples, which were all excerpted from an article titled 'Soviets in Space' published in Scientific America (Vol. 260, No. 2, 1989) and its Arabic translation which appeared in the Kuwait-based Majallat Al-Oloom (Vol.6, No. 8, 1989): (47) Buran (the Russian word for snowstorm) was lifted into orbit by the world's largest rocket. (48.) ُرفع بوران (تعني بالروسية العاصفة الثلجية) إلى مداره بواسطة أكبر صاروخ دف في العالم [Buran (which means snowstorm in Russian) was lifted to its orbit by the biggest launching rocket in the world] (49) New-generation space stations would be needed to house assembly workers. (50. ) وستبرز الحاجة إلى جيل جديد من المحطات الفضائية لسكنى عمال التجمي [There will occur the need for a new generation of space stations for housing assembly workers] (51) The space-endurance record was systematically extended. (52) .ً وارتدع الرقم القياسي للبقاء في الفضاء ارتداعا منظما [The record number for staying in the space rose a systematic rising] (53) Salyut 7 was equipped with a redesigned docking adapter. (54) . مزو ة بوحدة مهيأة أعيد تصميمها2 وكانت ساليوت [Salyut 7 was supplied [passive participle in Arabic] with a docking unit (which) was redesigned] 34 As can be noted, the authentic translation examples above instantiate agentive passivization (48), nominalization (50), activization (52), and the passive participle (54) as workable alternatives to render English passives. This empirical fact led Khafaji (1996:37) to conclude "Hence Arabic, as has been demonstrated in this section, does not avoid passivity but only expresses it differently". Finally, let us examine the progressive aspect as a micro-syntactic feature in order to see how the two languages can handle it in translation. English mainly expresses the progressive aspect grammatically by verb to 'be' + the marker -ing (e.g. John is writing a book). In contrast, Arabic usually expresses the progressive aspect lexically, e.g. ' ينببري جبون علبى تبأليف كتباب اآلنJohn is busy with authoring a book now' or ' يقبوم جبون بتبأليف كتباب اآلنJohn is engaged with authoring a book now'. Therefore, translators need to be aware of this grammatical mismatch. To see how subtle this asymmetry is, witness how Ali (1934/2006) and Arberry (1955/1996) fall short of rendering the progressive aspect properly in the following Quranic verse, respectively: .) فسبحان هللا حين تمسون وحين تصبحون55( (56) So glory be to Allah when you enter the evening and when you enter the morning. (52) So glory be to God in your evening hour and in your morning hour. One should note that the combination of the time marker and the verb حبين تصببحون/حبين تمسبون gives a sense of progressiveness in the Quranic verse, which is missed out in the two translations, viz. Ali renders the combination as a punctual act, whereas Arberry renders it as a state. To capture the sense of the progressive aspect, the translator needs to choose a similar strategy where a time marker interacts with a verb to bring out this progressiveness, viz. 'So glory be to Allah as you progress/move into the evening and as you progress/move into the morning'. 5. Semantic Features The semantics of a language mainly consists of lexical as well as phraseological features. Together, they cover both meaning that is compositional in nature as well as meaning that is unitary in nature. The former follows the Open Principle (Sinclair 1991) and accounts for meaning compositionally by deriving it from individual lexical items which are strung together according to the grammar of a given language. For example, the meaning of the sentence 'The boy chased the cat' is compositionally derived from the meaning of the content words boy, chase and cat combined with the function words/markers. The latter, in contrast, follows the Idiom Principle (Sinclair 1991) and derives a unitary meaning from the entire multi-word phraseology. For example, the meaning of the bold-faced idiomatic expression in the sentence 'In her attempt to convince John, Mary is flogging a dead horse' cannot be derived from the literal meaning of the words in it. Rather, it has a conventional unitary meaning which comes to mind once encountered in communication. Mismatches between Arabic and English that need 35 careful decision making exist at both word level and phraseology level as this section will demonstrate. 5.1 Word Level At word level, the semantic blankets of languages are never complete; there are always gaps involving both lexical and referential gaps (Rabin 1958; Ivir 1977; Dagut 1981). To start with lexical gaps, they represent holes where, in a language pair, one language lacks some lexemes that stand for shared concepts while the other language has compressed lexically those concepts in single words. Despite the fact that both English and Arabic are highly lexicalized (e.g. in terms of nominalization and verbalization) when it comes to familiar concepts, some lexical gaps do exist between them. Therefore, when translating an SL lexeme corresponding to a lexical gap in the TL, the translator needs to unpack the sense of that lexeme if s/he is to render the sense correctly. Working from Arabic into English, for example, four of the names of the fingers of the human hand, viz. الخنصببر، البنصببر، الوسببطى، السببّابة، اإلبهببامusually undergo lexical unpacking when rendered into English, viz. thumb, the index finger, the middle finger, the ring finger and the little finger, respectively. In many cases, Arabic lexemes corresponding to lexical gaps in English undergo lexical approximation, e.g. ' عبمpaternal uncle' and ' خبالmaternal uncle' are usually rendered as uncle and ' عمّبةpaternal aunt' and ' خالبةmaternal aunt' as aunt. While this may work in many contexts where the side of kinship is not important, it may seriously fail in instances where this kind of thing is significant. In such cases, the lexical unpacking of the kinship term becomes necessary. To see how lexical gaps can present formidable problems to even highly professional translators, let us cite an example from fiction translation to observe how rendering an Arabic lexeme by approximation can be damaging to the coherence of the text. In his translation of ‘AbdulRahman Munif’s تقاسيم الليل والنهار:مدن الملح, 1992 (Cities of Salt: Variations on Night and Day, 1993), Peter Thereoux translates the Arabic proverb [ ثلثين الولد لخالهTow thirds of the boy for his maternal uncle] as 'Two thirds of a boy are his uncle’s'. The fictitious encounter involves the citation of this proverb by one of the characters to claim more influence for maternal kinship than paternal kinship on children. Unfortunately, the English translation obliterates this culturebound schema by neutralizing the distinction between the Arabic lexemes ' عمpaternal uncle' and ' خالmaternal uncle' in a context where the discrepancy constitutes the intended message. The TL reader will definitely fall prey to the incongruence brought about by a rendition that does not cohere with the surrounding co-text and context. Following are some target reader responses (American native speakers’ responses) to the English translation above in its context (reported in Farghal, 2004): (58) - Family is everything. - Apples don’t fall far from the tree. - A boy learns from his family around him. - People trust their uncles 36 - People follow their masters, etc. As can be observed, the above English native speakers’ responses obscure the intended message and consequently, on a closer examination, render the TLT seriously incoherent. This incoherence is an immediate consequence of replacing the culturally determined, specific role of maternal kinship with a universally determined, general role of family relatedness in the context of the formation of children’s future behavior. To capture the intended message in such cases, where lexical approximation alone does not work, the translator needs to be an insider in both cultures: the SLC and TLC, i.e. s/he needs to unpack the Arabic kinship term, viz. 'A boy is his maternal uncle’s by two thirds' or 'Like maternal uncle like boy', which remodels the English proverb 'Like father like son'. Only in this way will the text make sense (see Chapter Two for more details). Working from English into Arabic, there also exist some English lexemes that correspond to lexical gaps in Arabic. Depending on context, among these we find words like 'spouse' which translates into ' روجhusband' or ' روجكwife', and 'parent' which translates into الوالك 'father' or ' الوالكmother'. In some cases, the translator has to read a sizeable portion of a text (e.g. a novel) in order to decide 'which is which' in the treatment of a lexical gap. To cite a real example, the first author of this book has recently translated the novel entitled 'Maps' (1986/ خك ائط2013) by the celebrity Somali writer Nurrdeen Farah in which there is a recurrent reference to Askar’s (the protagonist’s) two uncles (Uncle Orrax and Uncle Hilal). Starting to translate the novel without having read far through the text, the translator chose the Arabic paternal option for rendering both of them, viz. العك أوراككسand العك مكًل, respectively. It was not until having gone past halfway in the translation that he discovered that the latter referred to a maternal rather than paternal uncle. This being the case, an order was made to the computer to replace all the occurrences of الع مًلby الخكال مكًل. Without having done that, the Arabic translation would have offered a distorted world of kinship relations. In the following example, the translator has opted for awkward paraphrase based on dictionary definition because the lexeme 'affidavit' is not lexicalized in Arabic: (15) In the words of a Lonrho affidavit dated 2 November 1988, the allegations …. 3588 وفا2 ) واسب النص الوارو ف إفاو ك اصي دشفوع صياي ص د اا دؤوس لو و ص ار خ06( (printed in Baker 1992: 38; emphasis hers) ... فإن اووعا ات [And according to the text found in a testimony accompanied by an oath presented by Lourho corporation dated 2 November 1988, the claims ...] While it is true that the term 'affidavit' is not lexicalized in Arabic, the wordy definition is not justified in the Arabic rendering. A more acceptable and economical rendition would involve modifying the Arabic hyperonym ' شكااوtestimony' by one word without falling prey to wordiness (60 above) as in (61) below: فإن3588 وفا2 ) واسب النص الوارو ف شااو ع لي ص د اا دؤوس لو و ص ار خ03( ... اووعا ات [And according to the text found in a judicial testimony ....] 37 In fact, most English lexemes corresponding to Arabic lexical gaps need to be unpacked naturally and economically, viz. 'alibi' is rendered as ' ال ف صالغياdefense by absence' and 'date' is rendered as ' دوع غ ادa love appointment'. Because lexical gaps relate to familiar, but unlexicalized concepts in the TL, the most important step is to locate the relevant hyperonym, and then to modify it by a lexical descriptor in order to communicate the unlexicalized sense component. For their part, referential gaps, which represent partially shared or completely unshared concepts, i.e. those concepts that exist in one language but they are only present partially or they are completely missing in the other, are more challenging in translation activity. To start with partial referential gaps, one can refer to the many religious concepts that are partially shared between Islam and Christianity, being the relevant religions when translating from Western Christian cultures into Arab Muslim culture. Among these terms we find 'charity' vs. ص ص/ ركا, 'pilgrimage' vs. عا/اج, and 'ablutions' vs. تيا/ وضو. As can be seen, for each of the English terms we have two Arabic terms that come under a hyperonym, e.g. the hyperonym 'giving to the poor' has one form in Christianity (charity), whereas it manifests itself in two functionally different forms in Islam ركا (which is compulsory) vs. ( ص صwhich is optional). In terms of translation, such partial referential gaps usually lend themselves to the strategy of approximation in casual mentions (e.g. the rendition of ركاas 'charity' in fiction translation) and to other strategies, including approximation, in technical/religious texts. Below are five excerpted translations of a Quranic verse featuring this partial referential gap: َّ ج ُدوهُ عِن َد َّ صلَ ٰو َة َوءَا ُتوا َّ ) َوأَقِيمُوا ٱل62( َٱّلل ِبمَا َتعْ َملُون ِ الز َكا َة َومَا ُت َق ِّدمُوا ِألَنفُسِ ُكم مِّنْ َخي ٍْر َت َ َّ َّٱّللِ إِن 110﴿ البقرة بَصِ ي ٌر، / The Cow, 110) (63) And be constant in prayer, and render the purifying dues; for, whatever good deed you send ahead for your own selves, you shall find it with God: behold, God sees all that you do. (110) (Asad, p. 32) (64) Establish worship, and pay the poor-due; and whatever of good you send before (you) for your souls, you will find it with Allah. Lo! Allah is Seer of what you do. (110) (Pickthall, p. 18) (65) And be steadfast in prayer and regular in charity: And whatever good ye send forth for your souls before you, ye shall find it with God. For God sees Well all that ye do. (110) (Ali, p. 48) (66) And perform the prayer, and pay the alms; whatever good you shall forward to your souls' account, you shall find it with God; assuredly God sees the things you do. (110) (Arberry, vol. 1, p. 42) (67) Keep up prayer and pay the welfare tax; you will find any good you have sent on ahead for your own souls' sake is already [stored up] with God. God is Observant of whatever you do. (110) (Irving, p. 9) As can be observed, the translation strategies adopted include approximation (charity/alms) and descriptive translation (the purifying dues/the poor-due/ the welfare tax). On the one hand, one should note that in the SL culture the concept of زكاةis very specific and is associated with 38 obligatory giving, so the approximating terms charity/alms, which are associated in the TL culture with voluntary giving, are too general. Moreover, in the SL culture voluntary giving is associated with another term, that is صدقة, thus 'charity' and 'alms' technically become more appropriate renditions for صدقةrather than زكاة. In this way, Ali and Arberry’s translations seriously diverge from what is meant by the Islamic concept and, without a footnote, the relevant features required for the full and coherent interpretation of the term are lost in translation. On the other hand, descriptive translation employs the headwords 'due(s)/tax', i.e. something that is required, to denote the obligatory sense of زكاة. This decision succeeds in conveying the main, general aspect of the term (i.e. its being obligatory); however, the translators differ in the choice of the modifying word to render the more specific meaning. Asad derives his rendition from the spiritual connotations of ;زكاةhe states in a footnote that its main function is to “purify a person’s capital and income from the taint of selfishness” (p.18), thus basing his translation on the connotative meaning of the term. By contrast, Pickthall derives his translation from the category of people who are eligible to receive it, so he renders it as ‘the poor-due’. In this way, both translators attempt to explicate the concept to TT readers within the text as well as in footnotes. For its part, Irving’s translation 'Welfare tax' may give rise to different implications. It pertains generally to the amount of money paid by all people, the rich and the poor alike, to the government for the advancement of society as a whole. Without a footnote, target readers are likely to interpret this term in a different way from that intended in the source text. For example, without specifying that زكاةis obligatory and levied on the well-to-do for the welfare of the poor, the readers might infer that it is required of the rich as well as the poor. This inference does not serve the intended message, which aims at compassion and social justice rather that placing an extra burden on the poor. Added to this are the pejorative associations which the term tax may arouse in tax payers. The above different renditions give us an idea about how challenging the treatment of referential gaps in translation can be in authoritative texts like the Holy Quran. Referential gaps in less authoritative texts may also involve a variety of translation strategies including transliteration, approximation, descriptive translation, definition, omission, etc. Following are examples extracted from Ramses Awad’s translation titled 'The Beginning and the End' (1985) of Najeeb Mahfouz’s novel ( بداية ونهاية1949), where different strategies are employed to render referential gaps: (68 :) لهذا قال أحدهم قبل البدء في اللعب . ال نريد غشا:فقال حسن . طبعا:فقال الشاب . فلنقرأ الداتحة.(وقرءوا الداتحة جرعا بصوت مسموع ولعل حسن حفظها حول هذه المائدةp. 40) 39 [Because of this one of them said before the beginning of playing: - We don't want cheating. Hasan said: - Of course. The youth said: - Let's read al-faatiḥah. And they read al-faatiḥah in an audible voice and perhaps Hasan learnt it around that table] (69) Thus before they started the deal, one of them said, "No cheating". "Of course not", answered Hassan. The young man said, "Let’s recite the opening exordium of the Koran". They recited Al Fatihat audibly; it was possible that Hassan had learned it at that gambling table. (p. 53) (p. 51) ... ت جل اصا ودعطفا أدا ا ده فق ال فت صال و ) وكان ف أفن06( [And Fareed Affandi was wearing a julbaab and an overcoat. As for his wife, she wrapped (herself) in a (bathroom) robe ... (73) ... Farid Effendi wearing an overcoat over his gown, and his wife a dressing gown. (p. 66) فدخلت خادمة فريد أفندي، وخفت نفيسة إليه ففتحته،) ثم بلغ المسام طرق على الباب فقط أحدهم الحديث22( :محمد حاملة سلة مغطاة بغطاء أبيض وضعتها على السفرة وهي تقول )46 (ص. ستي تسلم عليك يا ستي وتقول هذا فطير القرافة[Then reached their ears knocking on the door, so one of them stopped talking and Nafeesa went to it (the door) and opened it. So the servant of Fareed Affandi Mohammed entered carrying a basket covered with a white cover, placing it on the table and saying: - My mistress greets you with peace and says this is pastry of graveyard] (73) A knock on the door interrupted their conversation. Nefisa hurried to open it. The servant of Farid Effendi Mohammed entered carrying a basket with a white cover and placed it on the table. "My mistress sends you her regards, madam," she said, "and she sends you mourning pastry." (p. 59-60) 40 As can be observed above, the translator has used different strategies for dealing with referential gaps. In (69), the referential gap is defined in the first mention and transliterated in the second one. This is a successful strategy where a contextual and/or co-textual link is established between definition and transliteration. In (70), the first gap ( أفنa title of address indicating respect for and superiority of addressee or referent) is transliterated in (71) and elsewhere in the text. The title is constantly employed as an absolute social honorific when referring to Farid (Farid Effendi). One should note that this title may be used relationally (interactionally) in Arabic (for more on absolute and relational social honorifics, see Farghal and Shakir 1994), in which case it should be approximated to something like 'sir', 'man', 'guy', 'big fellow', etc. depending on the context it occurs in. The other two items ( جل اa loose garment covering the body from neck to feet) is successfully approximated to 'gown' in English, while ' ال وrobe', which is a borrowing in designating the kind of gown worn when taking a bath, is unjustifiably approximated to 'a dressing gown'. To bring out this cultural nuance, the translator could have maintained the same lexeme (robe), adding a modifier, viz. 'a bathroom robe', in order to capture the extreme informality of that encounter. Last, the gap [ فطي الق افpastry of graveyard], which is a kind of Egyptian pastry offered at the graveyard when visiting the dead, is rendered as 'mourning pastry', thus substituting a more general term for a specific one. In this case, the reader is introduced with the function of the pastry independently of the location, i.e. the pastry, according to the translation, may be served in any place, which is not the case. To render the gap more accurately in terms of culture transfer, the location at which the pastry is served needs to be pointed out, viz. 'graveyard pastry'. One should note that the function of the referent here is incorporated in the location. In some cases, several referential gaps belonging to different cultures become, through the passage of time, familiar internationalisms. Examples like 'Rock and Roll' روك آ رول, 'hamburger' ماد غand 'MacDonald' ( داك و لAmerican), 'pizza' صيزاand 'pasta' صاو ا (Italian), 'Allah' هللا, 'Imam' إدام, 'hummus' ااصand 'falafel' ( فًفلArabic) have become largely familiar worldwide. The translator is not expected to struggle with internationalisms in translation. Once recognized, they should be formally borrowed. A good clue for the translator’s judgment would be Wikipedia, where such items are usually illustrated and many of them are displayed in pictures. 5.2 Phraseological Features At the phraseological level, collocations and idiomatic expressions stand out as two important types of multi-word units that often necessitate special attention from translators. They are a major component of the lexicon and constitute an indispensable element of lexical competence (Alexander 1978; Yorio 1980; Nattinger 1980, 1988; Aisensadt 1981; Cowie 1981, 1988; Strassler 1983; Benson et al. 1987; Baker and MacCarthy 1987; Sinclair 1987, 1991; Farghal and Obeidat 1995; Farghal and Al-Hamly this volume, 2007, among others). In terms of translation, Farghal and Shakir (this volume) argue that collocations are more communicatively useful than idioms because they are more familiar in discourse and can only be hardly replaceable by individual lexical alternates. For example, the English collocation 'public support' and its Arabic counterpart ال ع الشعare only awkwardly paraphraseable in translation. By contrast, idioms are less common in discourse and are usually replaceable by lexical alternates, for 41 instance, the Arabic idiomatic expression [ ا على ورقink on paper] and its English counterpart 'dead letter' can be replaced (albeit at the expense of reducing the degree of the text’s emotiveness) by غي دف ّعلand 'unimplemented' in the two languages, respectively. Collocations, to start with, manifest the behavior of words when they combine or keep company with each other. Word company may be derived predictably from the primary meaning of a word, in which case semantic correspondence would often obtain between languages. For example, the English verb 'pay' can collocate freely with words relating to money, viz. pay wages ف األجور, pay debts ف ال ون, pay the ransom ف الف, pay the rent ف األج, etc. In all these collocations, the collocator 'pay' maintains its primary meaning, hence the ease of rendering them into Arabic. However, the verb 'pay' predictably collocate with a few other items that have nothing to do with money, viz. pay attention [ عي او اlend attention], pay a visit [ قوم صز ارperform a visit], pay a compliment [ ع ّ ع اوعجاexpress admiration], and pay respect [ ع ّ ع اوا امexpress respect]. In all these cases, the verb 'pay' has acquired collocational/secondary senses that largely differ from its primary sense, hence semantic correspondence rarely obtains between English and Arabic in collocations that sail away from primary sense. Collocations that feature secondary rather than primary senses may present the most problematic area for student translators (and even for practitioners) because of two reasons: firstly, they are mostly lexicalized differently between any two languages and secondly, they do not usually lend themselves to acceptable paraphrase in the TL (for more details, see Farghal and Shakir (this volume); Farghal and Obeidat 1995; Farghal and Al-Hamly 2007). Consequently, the only guarantee to deal with collocations appropriately is the translator’s possession of a good bank of them in the language pair. By way of illustration, following are some English collocations juxtaposed with their Arabic counterparts: (74) a. b. c. d. e. heavy rain heavy sleep heavy meal heavy fog heavy smoker [أمطار غزيرةpouring rains] [نوم عميقdeep sleep] [ وجبة دسمةfatty meal] [ضباب كثيفcondensed fog] [شبببببخص مفبببببرط بالتبببببدخينperson extreme in smoking] f. g. h. i. heavy sea heavy industry heavy traffic heavy bread [بحر مائجwavey sea] [صناعة ثقيلةheavy industry] [أزمة مروريةtraffic crisis] [ خبز من عجين غير مختمبرbread from unleavened dough] A close examination of the above English collocations points to three possibilities when rendering them into Arabic. The first possibility, which is the least likely, is to have 42 semantic/formal correspondence in lexicalization, as in (74g) where the primary sense of the collocator 'heavy' is maintained. The second, which is the most likely, involves corresponding collocations where the collocator 'heavy' is lexicalized differently in Arabic as in (74 a, b, c, d, f, h). The translator’s ability to call up TL collocations that differ in lexicalization but have the same communicative value is a foundation stone as regards naturalness of the translation product. Some translations may sound unnatural simply because the translator fails to access correct collocations in the TL by either imposing the first possibility or unjustifiably resorting to paraphrase, which is the third option. This option may be appropriately followed when the SL collocation does not correspond to a TL collocation (whether formally or functionally) as in (74 e, i). Thus, the paraphrase strategy in rendering a collocation is necessitated when the TL does not have a familiar lexicalized collocation. To observe actual problems that may arise from a mishandling of English collocations, following are some Arabic examples extracted from published translations: ( .) وقد عزمت المنظمة قرارها إلعطاء الجائزة للسيد أوباما25 [The organization intended its decision to give the award to Mr Obama] .) لقد كان هروبا في آخر لحظة26( [It was a last moment flight] ) بدأت مسئولة العالقات العامة التي عملت على صيانة الضرر22( الناجم عن أشخاص يطلقون العنان ألهوائهم مثل ... [The officer of public relations who worked on maintaining the damage resulting from persons unleashing their whims like ... started ...] . وكانت األسالك الكهربائية تلصق بأعضائهم الجنسية... )28( [ ... and the electric wires were attached to their sex organs] To explain, the English collocations 'make a decision' ً[ ي ّتخبذ قبراراtake a decision], 'have a narrow escape' ' ينجو بأعجوببةescape miraculously', 'repair the damage' [ إصبال الضبررrepair the damage], and 'sex organs' [ األعضبباء التناسببليةreproduction organs] in (75)-(78) respectively have been erroneously translated into Arabic. Except for (75), which is an erroneous paraphrase 'It was a last moment flight' of the English collocation 'have a narrow escape', the Arab reader would be struck by the unnatural Arabic collocations in the examples above. Idiomatic expressions, for their part, are frozen expressions whose unitary meaning cannot be worked out from the dictionary meaning of the individual words in them. Such idiomatic expressions usually render the text more emotive. In terms of translation, however, the tinge of emotiveness furnished by idiomatic expressions can be maintained only when they appropriately lend themselves to rendering into corresponding TL expressions (whether in form or function). Otherwise, their communicative import is rendered apart from the idiomatic phraseologies (for more details on strategies to translate idioms, see Newmark 1988 and Baker 1992). Following are some illustrative examples: (79) It started raining cats and dogs when Peter met his blind date at the park. 43 .) بدأت تمطر كأفواه القرب عندما قابل بيتر فتاته التي يجهل هويتها في المتنزه80( [It started raining like mouths of goatskins when Peter met his girl whose identity is unknown to him at the park] .) اضطر المضربون إلى رفع الراية البيضاء وقبول تسوية جائرة م الشركة81( [The strikers had to hold the white flag and accept an unfair settlement with the company] (82) The strikers had to throw in the towel and accept an unfair settlement with the company. (83) The Syrians are caught between the devil and the deep blue sea - their leaders cause great suffering , but an invasion would bring many other problems. سيجلب الغزو األجنبي، ففي الوقت الذي يسبب قادتهم الكثير من المعاناة،) لقد وقع السوريون بين فكي ك ّماشة84( .مشاكل عدة [The Syrians fell between a pair of pincers; while their leaders cause much suffering, a foreign invasion would bring many problems] . يع ّد ابن خلدون منظراً ال يشق له غبار في علم االجتماع،) من منظور تاريخي85( [From a historical perspective, Ibn Khaldun is considered a theorist whose dust cannot be penetrated in sociology] (86) From a historical perspective, Ibn Khaldun is considered a past master theorist in sociology. Except for the idiomatic expression blind date in (79), which is paraphrased because a corresponding idiomatic expression does not exist in Arabic, the other idiomatic expressions in (80)-(86) travel idiomatically (functionally) between the two languages. One should note that the major challenge for translators here is to recognize the idiomatic expression as well as understand its meaning before starting to search for a rendition. While the maintenance of the emotiveness in the SL text needs to remain a priority in rendering idiomatic expressions, there are cases when paraphrase may be the only available option. Witness how the two English idiomatic expressions in (87) and (89) may have to be paraphrased into Arabic in (88) and (90), respectively: (87) The gap between the haves and have-nots still shows up clearly at the parliamentary elections. .) ما زالت الفجوة بين األغنياء والدقراء تظهر بوضو في االنتخابات البرلمانية88( [The gap between the rich and the poor is still clearly visible in the 44 Parliamentary elections] (89) The officials went through the roof when a local newspaper published a report about corruption. ) استشاط المسئولون غضبا عندما نشرت إحدى الصحف المحلية90( .تقريراً عن الفساد [The officials became so angry when a local newspaper published a report about corruption] Sometimes, idiomatic expressions correspond formally between English and Arabic while maintaining the same communicative value or, alternatively, are employed with different imports. Note how the idiomatic expressions 'be all ears' كلّبي آذان صباغيةand 'mop/wipe the floor with somebody' يمسبببح األرض بشبببخص مببباcorrespond both in form and meaning, while 'Cinderella' سببندريالand 'wash one's hand of somebody/something'شببيء مببا/يغسببل يببده مببن شببخص convey different imports in the two languages. To explain, the former 'Cinderella' indicates bad, unfair treatment in English, but it signifies outstanding beauty in Arabic; the latter meaning disassociating with someone/something in English (i.e. to stop dealing with the entity in question), whereas it means stopping pinning hopes/relying on someone/something in Arabic. Translators, therefore, need to be wary of formal similarity between idiomatic expressions as they might turn out to be idiomatic false friends (for more details, see Taylor 1998; Al-Wahy 2009). Following are two authentic examples from the Arabic version of Newsweek where the two strategies of calling up a TL idiomatic expression and paraphrasing the idiomatic expression are employed, respectively: (91) The Fayeds have turned the pre-bid House Fraser strategy on its head. .) اإلخوة الفايد قد قلبوا إستراتيجية هاوس أوف فريزر على عرض االمتالك رأسا على عقب92( (93) Many Americans thought that Hillary Clinton would be the democratic nominee for president, but a dark horse, Barak Obama was instead. ) اعتقد الكثير من األمريكيين أن هيليري كلنتون ستكون مرشبح الحبرب البديمقراطي لمنصبب البرئيس ولكبن اوبامبا94( . هو من أصبح مرشحا ً لهم،وهو شخص لم يكن معروف Apart from the general quality of the translation, the English idiomatic expression 'turn something on its head' in (91) is correctly rendered into the Arabic idiomatic expression يقلبب [ الشبيء رأسبا ً علبى عقببturn the thing's head on its bottom] in (92). By contrast, for lack of a corresponding Arabic idiomatic expression, the English idiomatic expression 'a dark horse' in (93) has been reduced to sense in (94), viz. [ وهو شخص لبم يكبن معبروفwho is a person who was not wee-known]. 6. Conclusion Being fundamentally a linguistic exercise, the translation process needs to involve a close consideration of all linguistic aspects of the text, including phonological, morphological, 45 syntactic, and semantic features. The present chapter has selectively explored, with ample illustrative examples, linguistic features that translators need to be alerted to in their work. In particular, various strategies of handling linguistic parameters have been investigated in the hope of bringing them into the consciousness of practicing translators. The discussion also offers insights into further investigations of linguistic considerations in translation activity. Together with other considerations including textual, pragmatic, cultural, stylistic, etc. (see Farghal and AlManna, 2014; other articles in this volume), the translator's work will definitely become an informed act. The Translation of English Passives into Arabic Mohammed Farghal & Mohammed Al-Shorafat 46 Abstract The study aims to check the intuitions reported in studies on the translation of English passives into Arabic against empirical data that consist of translations of English passive utterances as they naturally occur in an English text. It examines the linguistic strategies and resources that translators from English into Arabic employ when encountering passive utterances. It is shown that translators use many strategies with this order of frequency: nominalization, adjectivalization, passivization, activization and pseudo-activization. It is also argued that the claim that Arabic does not tolerate agentive passives is inadequate, since translators employ a variety of formal markers in rendering English agentive passives. The study concludes that English passivization is predominantly structure-based, whereas Arabic passivization is predominantly semantics-based. 1. Background of Study 1.1 Types of Translation Equivalence In its essence, translation is a feat of transferring meaning as manifest functionally in a certain context from one language to another. This transfer involves ipso facto phonetic, lexical, structural, pragmatic and textual decisions (for more details, see Farghal 2012). Such decisions are meant to bring forth equivalence. Within the context of translating, equivalence is viewed as a dynamic parameter that constitutes a correlative of text type, author and audience. That is, any discussion of equivalence independently of these three variables is doomed because it is these contextual factors that direct the translator's options during the search for natural/appropriate equivalence. Being a correlative of contextual factors, translation equivalence may differ from one context to another. This has prompted translation theorists to distinguish between different types of translation equivalence: dynamic equivalence (Nida 1964), formal equivalence (Catford 1965), functional equivalence (de Waard and Nida 1986) and ideational equivalence (Farghal 1994). For instance, a creative metaphor in a work of art may call for formal equivalence, while a decorative metaphor in a newspaper editorial may call for functional equivalence or, more practically, ideational equivalence. It can be noted that while these kinds of equivalence address translation equivalence in its entirety by placing their primary emphasis on meaning as it is derived from the syntagmatic interrelations in discourse, they give minimal attention to the structural machinery (linguistic competence in Chomsky's terminology) that functionalizes such syntagmatic relations. For the purposes of this paper, we shall adopt Widdowson's (1971) trichotomy of equivalence: structural, semantic and pragmatic equivalence, not because of its comprehensiveness and validity in the context of translating, but rather because of its relevance to the structural machinery which is largely overlooked in other theories of equivalence. Firstly, structural equivalence makes direct reference to surface forms at the sentence level. For example, the English 47 sentences in (1) and (2) are viewed as structurally equivalent to the French sentences in (3) and (4), respectively (Widdowson 1971: 61): (1) The postman opened the door. (2) The door was opened by the postman. (3) Le facteur ouvrit la porte. (4) La porte fut ouverte par le facteur. Secondly, semantic equivalence captures the propositional content of sentences independently of their surface forms. Accordingly, other sentences expressing the same propositional content may stand as semantically equivalent to (1)-(4) above despite the differences they exhibit in surface forms as illustrated in (5)-(8) below (ibid., p. 63): (5) It was the postman who opened the door. (6) It was the door that was opened by the postman. (7) Ce fut le facteur qui ouvrit la porte. (8) Ce fut la porte qui fut ouverte par le facteur. Thirdly, pragmatic equivalence relates to the illocutionary effect of utterances and therefore cannot be established in relation to isolated sentences but rather by making references to what utterances count as in context. However, while leaving due room for contextual variation, we may consider (9) and (10) pragmatically equivalent to (1) and (2) above (ibid., p. 62): (9) Le facteur a ouvert la porte. (10) La porte a ete ouverte par le facteur. This being the case, structural and semantic equivalents, while being relevant to theoretical analyses, can be of only limited practical value in the process of translating. That is, structural equivalence may not capture the communicative functions between the source language (SL) and target language (TL) as demonstrated by (1) and (3), on the one hand and (2) and (4), on the other. Similarly, semantic equivalence is based exclusively on propositional content regardless of significant differences in topicalization (Fillmore 1968), or in focus (Chomsky 1968), or in evaluativeness (Farghal 1991, 2012). By contrast, pragmatic equivalence lies at the core of the translating process because it views equivalence in terms of communicative value as relating to utterances or kinds of message rather than sentences in isolation. In a few cases, however, semantic, or even structural equivalence, may coincide with pragmatic equivalence. 48 1.2 English Passives and Their Arabic Counterparts The translation of English passives into Arabic has attracted several studies in the existing literature in translation studies and contrastive linguistics. Al-Najjar (1984: 158-160) points out that English agentive passives can be restructured in Arabic in two ways: either as an agentive passive or as an active voice clause in the word order (Object - Verb in the active voice + object coreferential pronoun - Agent) as in (11) and (12) below: (11) He will study the plan which was presented by his advisor. (12) a. sa -yadr usu -l -xi ṭṭat a [llati quddimat min qibali mustašaar-hi] will-study-the-plan which was-presented by advisor-his .وي رس الخط ال ص ُ دت د ص ل الاس شار b. sa-yadrusu-l-xiṭṭata [llati qaddama-ha will-study-the-plan .ص داا الاس شار mustašaaru-hu] which presented-it advisor-his وي رس الخط ال However, judged by the intuition of speakers of Standard Arabic, Al-Najjar argues, the active voice structure is preferable and more well-formed. It should be noted that Al-Najjar confines his discussion to passivized relative clauses in English which can be readily translated into embedded topic-comment structures in Arabic (12b), rather than the clumsy agentive passive in (12a) above. In this way, he arbitrarily avoids the more serious problem regarding the translation of matrix passive sentences in English into Arabic. El-Yasin (1996) pushes Al-Najjar' s analysis a step further by arguing that English passives should be translated into Arabic topic-comment structures in order to capture both form and meaning, as illustrated in (13) below, which was originally used in Farghal (1991) for a different analysis (see below): (13) a. The president was assassinated by the leftists. b. 'ar-ra'iisu 'iɤtaala-hu-l-yasaariyyuuna the-president assassinated-him-the-leftists .ال ئيس اغ اله اليسار ون According to El-Yasin, the topic-comment or theme-rheme representation is maintained in translating (13a) into (13b), i.e., 'the president' and `ar-ra'isu occupy the topic position in both sentences, while the rest of either sentence occupies the comment position. While discussing evaluativeness, Farghal (1991: 144) deals briefly with passivization arguing that "Passivization is a matter of optional thematization where all arguments are usually mentioned in English, but it is a pragmatic choice in Arabic, where an argument, namely the agent is obligatorily dropped; otherwise, an active sentence is chosen" as can be illustrated in (14) below: 49 (14) a. 'iɤtaala-1-yasaariyyuuna-r-'aiisa assassinated-the-leftists-the-president .اغ ال اليسار ون ال ئيس `The leftists assassinated the president'. or `The president was assassinated by the leftists'. b. 'iɤtiila-r-ra' iis-u assassinated-the-president `The president was assassinated'. . ُاغ يل ال ئيس Thematization, Farghal argues, belongs to natural evaluativeness, which is brought about by obligatory Move α, a rule whose output is a focused constituent. As for topic-comment structures, they also feature natural evaluativeness as they are argued to be generated by PSrules in both English and Arabic as can be illustrated in (15) and (16) below: (15) As for the novel, Ali read it. ( 1 6 ) a . 'amma-r-riwaayatu fa-qara' a-ha ʻaliyy-un as for-the-novel . فق أما عل ?-read-it Ali أدا ال وا b. 'ar-riwayatu qara'a-haa ʻaliyy-un the-novel read-it Ali-NOM . ال وا ص أما عل This being the case, El-Yasin's (1996) analysis confuses thematization with base-generation by suggesting that Arabic topic-comment structures be natural equivalents for English passive sentences. Other studies (Mouakket 1986, Saraireh 1990, and Khalil 1993) reiterate the argument that English agentive passives should be translated into Arabic active sentences. For instance, Mouakket (1986: 140) writes: "In the Arabic passive sentence the agent must be deleted. If the agent is to be mentioned, an active sentence must be used". However, Saraireh (1990: 186) points out that "there are cases in which the agentive passive structure of an English sentence could be maintained in the Arabic translation" as can be illustrated in (17) and (18) below: (17) It is believed that the common cold .. . (18) yuʻtaqadu 'anna-z-zukaama-l-ʻaadiyy-a .. . 50 is believed that-the-cold-the-common . .. ... ع ق أن الزكام العاو Obviously, Saraireh is interested in the syntactic rather than semantic properties of English passives of this kind, as the grammatical subject of (17) is not agentive and functions merely as a pleonastic slot filler. For his part, Khalil (1993) bases his conclusions about Arab students' translations of individual English agentive passive sentences and Arab experts' intuitions about such translations on the dogmatic assumption that the only variety of Arabic that is relevant to translation is Classical Arabic, i.e., the Holy Quran and Classical Arabic literature, among other traditional sources. He concludes (p. 179) that "the Arab translator's tendency to render English agentive passive sentences into corresponding Arabic agentive passive sentences originated under the influence of translation out of European languages which allow passive sentences with expressed agents. Moreover, the frequent use of Arabic agentive passive sentences in the media and modern literature has created the impression that these constructions are acceptable in CA [Classical Arabic]". 2. The Present Study As is clear, the above-mentioned analyses of translating English passives into Arabic are mostly based on concocted, individual sentences. In the best of worlds, such analyses may instantiate structural and/or semantic equivalents which can be of only limited practical use to the actual process of translating, because they make minimal reference, if any, to the communicative value of passive utterances as naturally occurring segments of discourse/text. This being so, the present paper aims to check the alleged intuitions reported in such studies against empirical data that comprise translations of English utterances containing passive forms into Arabic as they naturally occur in an English text. Put differently, this study investigates the linguistic strategies and resources that translators from English into Arabic fall back on when encountering passive utterances. Further, the present study seeks to substantiate that grammatical equivalents established at the sentence level may be at variance with translation replacements established at the discourse level in the same sense that dictionary equivalents may be different from translation replacements (Tabakowska 1990). Thus, the pragmatic rather than the sole semantic import of an utterance becomes a focal point in the process of translating naturally-occurring discourse. Subsequently, it is hoped that translation replacements will gradually find their way toward acquiring the status of translation equivalents. 51 2.1 Material and Subjects The translation task consisted of an English text containing 13 passive forms: 3 agentive and 10 agentless passives (see appendix). As announced in the book from which the text was taken (A Modern Course in English Syntax by H. Wekker and L. Haegeman (1985)), the text was an adaptation from A. Hosain and S. Pasricha's Cooking the Indian Way, which may render it a bit unnatural, in terms of density of passive forms. This being the case, the more experienced subjects might have been able to guess what the aims of the experiment were. The subjects consisted of two groups. The first group comprised 11 Jordanian MA translation students enrolling in a course labeled 'General Translation from English into Arabic' in their first semester. They were asked to translate the English text into Arabic in class with a forty-minute time limitation. The second group consisted of 5 Jordanian translation professors who had been teaching translation courses at both the undergraduate and the graduate levels. They were asked to translate the English text into Arabic in their free time with practically no time limitation. The subjects of both groups were asked to hand in a final version, since tracing the emergence of a translation replacement was not one of the aims. The experiment was also subject to some practical constraints regarding the small number of subjects, especially professors (not only was the number of avail able professors not large enough, but their willingness to do the experiment was also an important factor) and the conditions under which the two groups performed the translation task (for instance, it was inconceivable to have the professors do the experiment with a time limitation imposed). 2.2 Hypotheses Prior to selecting the translation task and collecting the empirical data, the present study had set the following hypotheses: 1. When translating English passives into Arabic, translators would probably seek pragmatic equivalence where structural equivalence falters. 2. Student translators are likely to pay more attention than translation professors to structural equivalence when translating English passives into Arabic. 3. The empirical data are likely to instantiate formal markers of Arabic agentive passive equivalents other than the oft-cited, and perhaps criticized, formal marker min qibali. 2.3 Results The results of this study showed that the two groups of subjects employed translation replacements as manifest in five main strategies when translating English passives into Arabic (cf. Khafaji 1996). These are: passives into nominalization (verb + verbal noun), passives into adjectivals 52 passives into passi ves, passives into actives, and passives into pseudoactives. Table 1: Frequency of each strategy for both groups of subjects Strategy No. % Nominalization Passives Adjectives Actives Pseudo-actives Total No. of renderings 61 48 39 34 16 198 30.80 24.24 19.69 17.17 8.08 100% Table 2: Frequency of each strategy for translation students Strategy No. % Nominalization Passives Adjectives Actives Pseudo-actives Total No. of renderings 36 34 28 26 10 134 26.86 25.37 20.89 19.40 7.46 100% Table 3: Frequency of each strategy for translation professors Strategy Nominalization Passives Adjectives Actives Pseudo-actives Total No. of renderings No. 25 14 11 8 6 64 % 39.06 21.87 17.18 12.50 9.37 100% Table 4: Frequency of each formal marker in agentive passives Formal Marker bimin qibali li-ri1d biwaasiṭat Both groups No. % 15 31.25 13 27.08 10 20.83 2 4.16 Students No. % 11 33.33 10 30.30 5 15.15 2 6.06 Instructors No. 4 26.66 3 20.00 5 33.33 — 53 Zero (Active) Total No. 8 48 16.66 100% 5 33 15.15 100% 3 15 20.00 100% of renderings The four Tables above relate to the validity of the three hypotheses in 2.2. The Tables 1-3 indicate the frequency of each strategy for both groups, for the students and for the professors, respectively. Table 4, in turn, indicates the frequency of formal markers of Arabic agentive passives for the two groups combined and separately. 2.4 Analysis and Discussion As can be seen from Table 1, translating English passives into nominalization (verb + verbal noun) in Arabic is the most frequent strategy, viz. 30.80% of the subjects' renderings featured this strategy. The frequency of translating English passives into Arabic adjectivals and actives is similar, being 19.69% and 17.17%, respectively. The lowest frequency, as can be noted, is that of translating English passives into Arabic pseudo-actives. The results shown in Table 1 strongly verify our first hypothesis regarding the employment of strategies that furnish pragmatic equivalence independently of structural equivalence. Arabic linguistic strategies and resources appropriately catered for the semantics of passivization in English, thus moving from structure-based to semantics-based correspondence. The following discussion will shed light on each of these strategies, giving examples and explanations. 2.4.1 The Five Main Strategies 1. Nominalization Nominalization turned out to be the most useful strategy for translating English passives into Arabic, viz. 30.80% of the renderings involved Nominalization. Formally, this strategy features a pleonastic verb (a verb that is mostly devoid of semantic content) plus a verbal noun that is derived from the matrix verb in the English passive structure. In this respect, Arabic seems to possess two categories of pleonastic verbs: one category always goes with semantically based passives, while the other goes with actives. Semantically based passives are structures that are passive in meaning but not in form. By way of illustration, consider (19) and (20) below: ( 1 9 ) qama-l-mudiiru bi-ziyaarati-l-maṣnaʻi made-the-manager in-visiting-the-factory `The manager visited the factory'. . صام الا صز ار الاين ( 2 0 ) t aʻarr a ḍa-l -ma ṣnaʻu l i -l -huj uumi exposed-the-factory to-the-attack 54 `The factory was attacked'. .تع ض الاين للاجوم As is evident, qama is a pleonastic active verb, whereas taʻarraḍa is a pleonastic passive verb. These two verbs are used only to improvise nominalization in Arabic, and are void semantically. Interlingually, verbs of this sort may create structures featuring what is termed 'effected objects' in English such as 'make an attempt', 'pay a visit', 'give an interview', etc., bearing in mind the existence of two different categories of such verbs in Arabic. Empirically, our data demonstrate the successful employment of pleonastic passive verbs in nominalization for rendering English passives in Arabic, as illustrated by the English passives in (21) and their respective Arabic counterparts in (22) below: (21) a. Since the earliest times the sub-continent has been invaded by many tribes from the North. Only recently was the region divided into the independent countries of India and Pakistan. c. Nowadays Indian food is esteemed all over the world. b. (22) a. famunðu ʻuṣuurin mubakkiratin taʻarraḍat šibhu-1-qaarati since ages early exposed semi- the-continent 'ilaa ɤazwi qabaa'ila mutaʻaddidatin min-aš-šarnaal to invasion tribes many from-the-North .فانه عيور د ا تع ّر ضت ش ه القار إلى غزو ص ائل د ع و د الشاال b . wa-munðu ʻahdin qariibin tamma taqsiimu and-since time 'ilaa dawlatayni to tilka-l-manṭiqatu recent completed division that-the-area mustaqilatayni humaa-l-hindu wa-l-baakistaan two countries independent . الانطق إلى وول ي they -the-India and-the-Pakistan ودنه عا ص ب ت مّ تقسيم تل 55 c. wa-fi-l-waqti-1-ḥaḍir i yaḥða-ṭ-ṭaʻamu-1-hindiyyu bi-t-taqdiiri and-in-the-time-the-present enjoy-the-food-the-Indian with-the-esteem fi jamiʻi 'anḥaa'i-l-ʻaalami in all parts-the-world . وف الوصت الحاض حظى الطعام الان صال ق ف جاي أ حا العال As can be seen, the English passive verbs 'has been invaded', 'was divided' and 'is esteemed' have been translated into the Arabic nominalizations taʻarraḍat 'ilaa ɤazwi, tamma taqsiimu, and yaḥðaa bi-t-taqdiiri, respectively. It should be pointed out that the Arabic translations above sound natural. They therefore constitute very suitable pragmatic equivalents for the English passives in (21). Nominalization should no doubt be singled out as an important strategy when addressing the translation of English passives into Arabic. It should be mentioned that some Arabic translations (by 5 out of the 11 students only) of the English sentence involving the non-dynamic verb 'is esteemed' exhibited evaluative topiccomment structures (cf. Farghal 1992b and El-Yasin 1996). By way of illustration, consider the translation of (21c) above in (23) below: (23) wa-fi haaðihi-1-'ayyaam fa-'inna-ṭ-ṭaʻaama- l-hindiyy-a yaḥðaa and-in these-the-days ?-that-the-food-the-Indian-enjoy bi-t-taqd-ir-i fii jamiiʻi 'aqṭaari-l-ʻaalam with-the-esteem in all countries-the-world . وف مه األ ام فإن الطعام الان يحظى بالتقدير ف جاي أصطار العال Note that the Arabic sentence in (23) is a topic-comment structure (the topic is italicized and the clause that follows the topic is the comment) that has been rendered evaluative by the employment of the emphatic particle 'inna in introducing the topic. (For more on evaluativeness, see Farghal 1991) 2. Passives The translation of English passives into Arabic passives came second in frequency, viz. 24.24% of the renderings featured passivization. The English examples in (24) and their respective empirical translations in (25) below are illustrative: (24) a. Since the earliest times the sub-continent has been invaded by many tribes from the North. b. c. The art of the presentation of food was also developed in these countries. Silver or gold leaf is often used for decoration. (25) a. wa-munðu 'aqdami-l-ʻuṣuur fa-qad ɤuziyat šibhu-1- qaarati and-since earliest-the-ages ?-verily was invaded semi-the-continent 56 min qibali qabaa'ila kaθiiratin min-aš-šamaal by tribes many from-the-North .ودنه أص م العيور فق ُغريت ش ه القار د ص ل ص ائل كةي د الشاال b. wa-qad tuwwira fannu ʻard-i-ṭ-ṭaʻaami 'ayḍan fii haaðayniand-? was developed art presentation-the-food also in thesel-baladayn the-countries . وص ط ُور ف ّ ع ض الطعام أ حا ف مه ال ل c. wa-yustaʻmalu-ð-ðahab-u wal-fiḍḍattu and-is used-the-gold fi-t-tazyiin and-the-silver in-the-decoration . ويستعمل الهمب والفح ف ال ز ي 3. Adjectivals Adjectivals proved to be the third most common strategy for rendering English passives in Arabic, viz. 19.69% of the renderings used adjectivals as equivalents of English passive verbs. It should be noted that English passive verbs that are subject to adjectivalization in Arabic are stative or non-dynamic in nature, i.e., verbs that involve minimal action, if any. Consider the English examples in (26), along with their respective empirical translations in (27) below: (26) (27) a. The consumption of beef is forbidden to the Hindu, and the consumption of pork is not allowed for Muslims. b. Nowadays Indian food is esteemed all over the world. a . . . min dalika 'anna laḥma-1-baqari mamnuuʻun ladaa-l-hinduusi from that that meat-the-beef forbidden to-the-Hindu wa-tanawuli laḥmi-l-xinziiri muḥarramun ʻala-l-muslimiin and-eating meat-the-pork forbidden on-the-Muslims دح م على ل ى الان وس وتناول لح الخنز، د ذل أن لح ال ق دانو... . الاسلاي b. wa-li-ṭ-ṭaʻaamu-l-hindiyyi-l-'aana qiimatu-hu fii jamiiʻi and-for-the-food-the-Indian-the-now value-its 57 in all 'anḥaa'i-l-ʻaalam parts-the-world . اآلن صيا ه ف جاي أ حا العال وللطعام الان Syntactically, the Arabic translations above are verbless, that is, the English sentences in (26) have been rendered into Arabic equational sentences in (27), which enjoy a strong presence in the grammar of Arabic and are traditionally considered topic-comment structures (however, see Obeidat and Farghal 1994). This strategy proves useful when encountering non-dynamic (non-action) English passive verbs. It should be noted that verblessness is not a necessary condition on the Arabic sentence for inclusion in this category, because Arabic sentences may alternatively feature linking verbs as can be illustrated in (28) as a translation of (27b) above: (28) wa-fii haðihi-1-'ayyam 'aṣbaha-1-' akl-u-hindiyyu ðaa qiimatin and-in these-the-days became-the-food with value fii jamiiʻi 'anḥaa'i-l-ʻaalam in all parts-the-world . ذا صيا ف جاي أ حا العال وف مه األ ام أص ح األكل الان 4. Actives The frequency of translating English passives into Arabic actives came fourth, viz. 17.17% of the renderings exhibited actives. As we have already seen in Section 1.2., the existing literature overestimates the use of this strategy by translators from English into Arabic, particularly for English agentive passives. By way of illustration, consider the English examples in (29), along with their respective Arabic empirical translations in (30): (29) a. Later on it was occupied by the British. b. The art of the presentation of food was also developed in these countries. (30) a. θumma- ḥtalathaa-l-biritaniyytina fii waqtin laḥiq then occupied-it-the-British in time later . وصت واق طا يون ف ث اا لاا ال b. wa-qad tatawwara fannu taqdimi-t-taʻami fii haa ðihi-dand-? developed art presentation-the-food in these-the duwal countries . الطعام ف مه ال ول وص تطور ف ّ تق 5. Pseudo-Actives The low frequency of pseudo-actives, viz. only 8.08% of the renderings featured this strategy, may suggest that pseudo-actives are insignificant when it comes to translating English passives 58 into Arabic. This is not the case, however, as pseudo-actives constitute an important strategy in translating certain English passive verbs such as 'be influenced' and 'be divided', which are in fact the only instances in the translation task in this study that may call for this strategy. Interestingly, 16 of the 32 renderings — i.e. a full 50% — of these two cases employed pseudo-actives. Pseudo-actives are those Arabic verbs that are active in form but passive in meaning. By way of illustration, observe the English examples in (31) along with their respective empirical Arabic translations in (32): (31) (32) a. The eating habits of the Indo-Pakistan sub-continent are influenced by historical and geographical factors. b. Only recently was the region divided into the two independent countries of India and Pakistan. a. tata'aθθaru ʻaadaatu-1-' akli- fi šibh-i-l-qarrati-1-hindiyyati be influenced habits-the-food in semi-the-continent Indian bi-ʻawaamila taariixiyyatin wa-jaɤrafiyyah by-factors historical . صعوادل تار خي وجغ افي and-geographical ت أث عاوات األكل ف ش ه القار الان b. wa-lam tanqasimu-il-manṭiqatu 'ila baladayni and-not divided-the-area to two countries mustaqillayni huma-l-hindu wa- 1-bakistan-u 'iliaa independent they-the-India and-the-Pakistan except mu' axxaran recently . ، دس قلي ماا الان وصاكس ان إو دؤخ ا ول تنقس الانطق إلى صل While (32a) is probably the most natural equivalent for (30a), (32b) is acceptable as an equivalent for (31 b) only insofar as the anonymous agency in (31b) is overlooked. The dropping of the agent in (31b), however, should not be taken lightly or coincidentally; rather, as Bolinger (1980: 29) puts it, "The most useful — and dangerous — function of the passive (in English) is to enable the speaker to keep silent about who performs the action". To illustrate this point, consider (33) below in which agency, though anonymous, plays a significant role: (33) wa-lam tuqassam-il-manṭiqatu 'ilaa baladayni and-not was divided-the-area to two countries mustaqillayni huma-l-hindu wa-l-baakistan-u 'iliaa i nd epe nde nt t h e y -t he -In d i a n and-the-Pakistan except 59 mu' axxaran. recently. ،. دس قلي ماا الان وصاكس ان إو دؤخ ا ول ت ُقسم الانطق إلى صل 2.4.2 Translation Students vs. Translation Professors Having discussed the five main strategies for translating English passives into Arabic, let us turn to the significant performance differences between the two groups of subjects (Tables 2 and 3): translation students vs. translation professors. The second hypothesis in this study states that student translators are likely to pay more attention than translation professors to structural equivalence when translating English passives into Arabic. This hypothesis is particularly borne out by the significant difference in the employment of the strategy of nominalization: it accounted for 26.86% for students vs. 39.06% for professors. Bearing in mind the obvious usefulness of this strategy in translating English passives into Arabic, it can be deduced that translation professors give much more attention to translation replacements that achieve pragmatic equivalence than translation students. Further, the medium discrepancy of the strategy of translating English passives into Arabic passives between students and professors — 25.37% for students vs. 21.87% for professors — lends more support to our hypothesis by indicating that translation students attended to translation replacements that achieve structural equivalence more than translation professors. It can be argued here that professors are more experienced and probably more competent in translation; hence they give precedence to pragmatic (illocutionary) import over formal (structural) import for the purpose of furnishing more natural equivalents. Finally, the results showed that student translators resorted to the strategy of adjectivals more than translation professors: 20.89% vs. 17.18%, respectively. Likewise, student translators used Arabic actives as equivalents for English passives more than translation professors: 19.40% vs. 12.50%, respectively. Obviously, the greater part of the differential in these two strategies went to nominalization, which has been argued to be an important linguistic resource for translating English passives into Arabic, thus furnishing further support for our hypothesis. 2.4.3 Formal Markers of Arabic Agentive Passives The third hypothesis is also borne out as the empirical data have instantiated formal markers of Arabic agentive passive equivalents other than the oft-cited formal marker min qibali. As can be seen from Table 4, bi- 'by' was the most used formal marker in Arabic passive equivalents, viz. 31.25% of the renderings featured it. As the results showed, it was used more by students than by professors: 33.33% vs. 26.66%, respectively. min qibali 'by' came second scoring 27.08% of the renderings; again, it was used more by students than by professors: 30.30% vs. 20%. li/'ilaa 'to' came third in frequency, comprising 20.83%; it was used much more by professors than by students: 33.33% vs. 15.15%, respectively. As a fourth formal marker, biwaasiṭat 'by means of' was used only twice by students, comprising 4.16% of the renderings. Finally, a 60 zero marker was sometimes used, by translating English agentive passives into Arabic actives; this accounted for 16.66% of the renderings. Contrary to common claims in the literature, these results show beyond doubt that translators into Arabic fall back on many formal markers to indicate agentive passive equivalents. In most cases, these formal markers sound so natural in Arabic that one can trace no foreign influences. Therefore, the common directive to avoid agentive passives when translating English passives into Arabic should be reconsidered, if not rejected altogether. In the extreme case, min qibali should be tolerated in Arabic because, contrary to claims that ascribe it to the influence of translating from European languages into Arabic, it is used natively in Arabic in both the mass media and modern literature, among other genres of discourse. Consequently, the oft-cited instruction to translate only into Classical Arabic is, we believe, ill-conceived as translation is, after all, an act of communication that should be pertinent to language in use rather than to prescriptive and/or rather obsolete varieties of Arabic that are only of relevance for historical purposes. 3. Conclusion The present paper has departed from available treatments of passivization as a structural property in an attempt to reach for pragmatic parameters that govern the translator's decisions during the process of translating English passives into Arabic. It has been shown that translators fall back on various linguistic strategies and resources in their endeavor to search for Arabic natural equivalents for English passives. Thus, it has been argued that the communicative value of an utterance should be given precedence over the semantic and/or structural value for the purpose of furnishing pragmatic rather than frequently unfitting semantic or structural equivalents. Most importantly, the present study sends a clear message to translation practitioners and theorists alike that translation should essentially be viewed as an act of communication; hence it should be practiced independently of dogmatic and/or prescriptive tenets that frequently crop up on the linguistic scene of Arabic. For instance, the claim that Arabic does not tolerate agentive passives has been proved to be inadequate, as translators have a variety of formal markers at their disposal when phrasing out Arabic agentive passives. Further, this paper points clearly to the different nature of passivization in English and Arabic. While passivization in English proves to be predominantly structure-based, Arabic passivization turns out to be semantics-based, viz. more than 75% of the translations of English passives in this study employed linguistic strategies and resources irrespective of 61 the verb form. That is, the interpretation of an utterance in Arabic as passive may stem from nominalization, adjectivals, and pseudo-actives, in addition to passivization proper. Last but not least, this research offers implications for translator training programs. In particular, translator trainers should alert their students to the intriguing nature of passivization in Arabic. They should point out that Arabic has its own linguistic resources that can handle all cases of English passivization pragmatically as they occur naturally in discourse. Most importantly, students should be sensitized to different types of translation equivalence in general and pragmatic equivalence in particular. Appendix The Translation Task The eating habits of the Indo-Pakistan sub-continent are influenced by historical and geographical factors. Since the earliest times the sub-continent had been invaded by many tribes from the North. Later on it was occupied by the British. Only recently was the region divided into the two independent countries of India and Pakistan. The influence of all the different invasions can be found in the culture and the eating habits of the sub-continent. Indian and Pakistani food are very similar, but regional and religious influences can be observed. The consumption of beef is forbidden to the Hindu, and the consumption of pork is not allowed for Muslims. The art of the presentation of food was also developed in these countries. Silver or gold leaf is often used for decoration. The metal is beaten very fine: it can almost be blown away. Nowadays Indian food is esteemed all over the world. (Adapted from A. Hosain and S. Pasricha, Cooking the Indian Way) Translating English Modals by Zero Equivalents in Arabic: A Case Study of Macbeth Mohammed Farghal & Alban Beqri Abstract The present paper aims to show the discrepancy brought about when translating English modals by Arabic zero equivalents. It is a case study of Macbeth, whose English modals comprise the corpus of analysis. Within the 166 modals in the play, 39 (23.5%) have undergone deletion in Arabic translation. In the bulk of these cases, the failure to render English modals has seriously marred Arabic renderings, e.g. when shifting from epistemic possibility to epistemic certainty. This practice often results in the loss of significant nuances introduced by English modals. The 62 study concludes that, while English possesses a highly grammaticalized system of modality, Arabic employs a diversity of modal expressions which can effectively capture the shades of meaning encapsulated in English modals. The translator, therefore, should pay utmost care when translating modality from English into Arabic, and should view zero equivalents as adequate only in few cases where the import is not affected. 1. English and Arabic Modals The verb constitutes the most important element in a sentence because its main function is to relate the arguments in the sentence together in order to produce meaningful propositions, which are considered the backbone of discourse. While lexical verbs do this function quite straightforwardly, auxiliary verbs, including modals, are employed in verb groups to add nuances of meaning which relate to the entire state of affairs in the sentence. Thus, the task of rendering the meaning of sentences involving auxiliary verbs becomes more challenging to the translator. This challenge is most manifested when rendering English modals because these modals introduce a subjective parameter to the referential meaning whereby the speaker/writer includes his/her own attitude toward the proposition in terms certainty and/or necessity. Halliday (1970:335) maintains that modality is "a form of participation by the speaker in the speech event. Through modality the speaker associates with the thesis an indication of its status and validity in his own judgment, he intrudes and takes up a position”. In this way, the speaker/writer becomes an evaluator of the transmitted information in the process of communication. Essentially, most authors (Halliday 1970; Lyons 1977; Perkins 1983; Coates 1983; and Palmer 1986, among others) divide English modals into two basic categories: epistemic and deontic modals. On the one hand, epistemic modality involves the speaker's/writer's judgment of the proposition he is putting forward in terms of certainty, which ranges between factuality (e.g. The flight leaves at midnight) and possibility (e.g. The flight may leave at midnight). On the other hand, deontic modality views the proposition at hand in terms of necessity, which ranges between strong obligation (e.g. John must leave now) and permission (e.g. John may leave now). Thus, while epistemic modality looks at language as "information", deontic modality views language as "action" (Palmer 1986). In addition, the above authors usually address the issue of indeterminacy in English modals. This indeterminacy is so clear when dealing with decontextualized sentences but often disappears when locating utterances within their relevant contexts. For example, the individual sentence 'The road may be closed' may trigger an epistemic or deontic interpretation. The context of the utterance, however, would indicate which reading is intended. Moreover, the interpretation of English modals is largely affected by the grammatical category of time reference, which adds to the degree of their indeterminacy. For example, the modal may denotes permission when used deontically (as in 'He may turn the TV on'). However, the same modal will have an epistemic reading when the time reference is past (as in 'He may have turned the TV on'). The same is true for the modal must, which denotes obligation when used deontically (as in ' He must do his assignment'), whereas it takes an epistemic reading when the time reference is past (as in 'He must have done his assignment'). In contrast with the modals may and 63 must, the modal should maintains the same nuance of meaning when the time reference is changed to past. For example, the deontic interpretation is preserved in the sentences 'He should do his assignment' and 'He should have done his assignment'. Accounts of English modality overwhelmingly focus on grammatical rather than lexical modality. This is reasonably justified by the fact that English has a neat system of English modal verbs which manage to grammaticalize meaning in an efficient manner. However, this does not mean that English lacks lexical modal resources: it just means that the grammatical category of modals is more efficient and handier than the lexical counterpart. In fact, one may find a lexical counterpart for every modal auxiliary. Examples may include 'I can do it vs. I'm able to do it', 'I have to go vs. I'm obligated to go', 'John may come in now vs. John is allowed to come in now', etc. Therefore, the notion of modality in English can be said to be predominantly grammatical in terms of usage. Unlike English, Modality in Arabic involves a mixed bag of modal expressions rather that a neat category of modal auxiliary verbs. This bag includes modal verbs such as yajibu ' جبmust', yumkinu ' ااmay/might', and yanbaγii ' ن غshould', particles/prepositions such as qad ص 'may/might', la alla ' لعلmay/might' and alaa ' علىmust', prepositional phrases min-l-mumkini د ' الاااmay/might' and min-l-waajibi ' د الواجبshould', lexical verbs such as yuħtamalu ' ح الbe possible' and yastalzimu ' س لزمrequire'. This elaborate array and mixed bag of Arabic modal expressions may have been the reason why modality had not been a subject of grammatical analysis in the otherwise comprehensive medieval Arabic grammar (for more details, see ElHassan 1990). This being the case, most recent studies of Arabic modality (Zayed 1984; El-Hassan 1990; Farghal and Shunnaq 1999/2011; Abdel-Fattah 2005; Al-Qinai 2008; Wided 2010) take the neat system of English modals as a point of departure when approaching Arabic modality from a linguistic and/or translational perspective. The general conclusion of these studies is that Arabic lacks a highly grammaticalized system of modals although it possesses the lexico-grammatical means to capture all modal nuances in discourse. The authors usually engage in listing English modal verbs and their possible Arabic counterparts (e.g. See Abdel-Fattah 2005 and Al-Qinai 2008) in terms of epistemic and deontic modality. A more holistic approach (Zayed 1984) reduces the epistemic parameter in English and Arabic to may vs. must and rubamaa رصااvs. laa budda و ص, respectively, and the deontic parameter to may vs. must and yumkinu ااvs. yajibu جب, respectively. De Haan (1997: 50) rightly improves on zayed's typology by presenting each two items on separate continua, thus making it visible that the two types of modality exist in different degrees from weak to strong, as can be seen below: a. b. Epistemic modality Weak __________________________ Strong رصاا وص Deontic modality 64 اا Weak ___________________________Strong جب Clearly, this approach is based on modalistic generalizations without going into the details. For example, yajibu جبis taken to be exclusively deontic, whereas laa budda و صis viewed as exclusively epistemic. While this is true in the former case, it is not in the latter case, because laa budda can function both epistemically and deontically (compare laa budda ׳anna-hu majnuun-un ' و ص أ ه دجنونHe must be crazy' with laa budda ׳an ׳adrusa-l-yawma ' و ص أن أورس اليومI must study today' (for more details, see Farghal 2012: 102-104)). To conclude this section, and regardless of the way we approach modality, we must realize the fact that modal expressions inject our utterances with a personal perspective that incorporates context-dependent nuances of the notions of certainty and necessity. Therefore, the addition or deletion of a modal expression in an utterance will most likely result in marring the intended meaning in translation activity. Further, the shift from one modal in one category to another modal in a different category in translation usually distorts the nuanced import of an utterance (for more details, see Badran 2001). This does not mean that we cannot use more than one modal expression for the same purpose; on the contrary, it is the case that multiple modal expressions sharing the same nuance (whether epistemic or deontic) may replace each other both intralingually (within the same language) or interlingually (across two languages). For example, the English modal may in the sentence John may be at home may be rendered by many Arabic modal expressions, including yumkinu, qad, la ala, min-al-mumkini, etc. 2. Corpus and Procedure The present paper is a case study of the rendition of English modals in Macbeth into Arabic by zero equivalents. The choice of Macbeth has been motivated by the length of the play (being the shortest of Shakespeare's plays), the intensive use of modals, ease of reference (the translation being bilingual) and the relative recency of the publication. The study will confine itself to the English modals that have been lost in Arabic translation. In this way, it does not aim to examine the accuracy of rendering English modals by Arabic modal expressions in this play. Rather, it aims to show the discrepancy created when overlooking an English modal in an Arabic rendition. The Arabic translation was made by Abdurrazak Al-Khaffaji (2008) by Dar Al-Bihar, Beirut, Lebanon. It was published along with the English text where the Arabic translation is given on the opposite pages of the English original. 3. Results The play contains 166 English modal verbs. The number of Arabic zero equivalents for English modals is 39 (23.5%) in the corpus. The following Table shows the frequencies and distribution of English modals and Arabic zero equivalents: English Modal Shall must Should Would Frequency 35 28 25 20 % 21% 16.8% 15% 12% Zero Equivalent 7 7 6 7 65 % 17.95% 17.95% 15.38% 17.95% Can Will May Could Might Total 20 14 11 9 4 166 12% 8.4% 6.6% 5.4% 2.4% 6 2 0 2 2 39 15.38% 5.12% 0% 5.12% 5.12% 4. Discussion The above table shows that the modal verbs shall, must, and should are the most frequent in the play accounting for almost 53% of the total number. They are followed by would and can which account for 24% of the total. The rest of the occurrences are claimed by will, may, could and might, which account for almost 23%, with will being in the lead and might the least occurring. In terms of translation, shall, must and would (which register a high frequency of occurrence in the data) account for a little more than 50% of the number of Arabic zero equivalents. The following discussion will examine some examples of zero equivalents of English modal verbs to see the impact they leave on the Arabic renditions. The main concern will be on the study items (i.e. English modal verbs and their Arabic renditions) and other segments in the translation will be relevant only if they relate to the rendering of modal verbs, which are highlighted in boldface. 4.1 shall The modal verb shall displays a high degree of certainty that comes very close to factuality. That is why the translator can readily omit it and have the Present Simple form replace the English verb group without seriously affecting the translation in Arabic. This can explain the translator's option for the Arabic zero equivalents (7/35) for this epistemically strong English modal. It should be noted that the use of the Arabic Present Simple form shows a negligible epistemic extra strength than the English counterpart, as can be illustrated below: But I shall crave your pardon; that which are my thoughts cannot transpose; angels are still bright, though the brightest fell … (p. 219 / Malcolm to Macduff). ... فالاًئا دا تزال ودع رغ أن ألاعاا ص موى أفاار دا أ ت عليه ل تغي:،لان أو ايح عهرا. Thus, the English segment 'But I shall crave your pardon' is rendered as laakinni ٫astamiiħuka uðran ،' لان أو ايح عهراBut I crave your pardon'. This rendering sounds as natural as laakinni sa׳astamiiħuka uðran ،' لان وأو ايح عهراBut I shall crave your pardon', and the loss in epistemic modality is negligible. 4.2 must The modal must is the second most frequent in the play (16.8%) and is translated by Arabic zero equivalents seven times (7/28). The deletion of this modal in Arabic translation seriously impacts the meaning. In the following example, the deontic modality exhibited by the modal must is erroneously omitted in the Arabic rendition, as can be seen below (Back translations are squarebracketed): 66 If charnel-houses and our graves must send those that we bury back, our monuments shall be the maws of kites. (p.163) .لو أن وا يسنا وص ور ا تعي د وفنا ألص حت أض ا نا ف اواصل الح آت [If our charnel-houses and our graves bring back those that we buried, our monuments shall become in the maws of kites] In this example, Macbeth (addressing the ghost of Banquo) embeds deontic modality that involves strong obligation in a conditional clause. The translator opts for omitting the nuance added by the embedded deontic modality within the hypothetical conditional clause while keeping the clause itself. Thus, the sense of obligation is completely lost, which distorts the nuanced meaning. To capture the combination of hypothetical conditionality and obligation, the translator should employ either the deontic Arabic yajibu جبor laa budda و صin the hypothetical condition, as can be seen below: . إن كان يجب على واو سنا وص ور ا إعاو د وفنا ألص حت أض ا نا ف اواصل الح آت [If our charnel-houses and our graves had to bring back those we buried, …] . إن كان ال بد لنواو سنا وص ور ا إعاو د وفنا ألص حت أض ا نا ف اواصل الح آت [If our charnel-houses and our graves had to bring back those we buried, …] 4.3 should This modal claims 25 occurrences in the play (15%), 6 of which are translated by Arabic zero equivalents. Consider the following example which includes a subtle use of should: [Aside] Were I from Dunsinane away and clear, profit again should hardly draw me here. (p. 271) ][جا ا. فاا د داسب عي إلى منا ثا ي، ع و سينان وا ا،لو كنت صعي ا [[Aside] Were I from Duninane away and free, no profit brings me back here again] The above example involves a subtle interaction between the modal verb should and the negative adverb hardly in order to signal a hypothetical state of affairs which is less definitive than a negated proposition without should. The translator, however, opts for a definitive, negated proposition which can be back-translated into 'no profit brings me back here'. As can be seen, the modality in the translation amounts to factuality (the highest degree of epistemic modality). This contrasts with the less assertive degree of modality in the source text (ST). To capture this nuance by Macbeth, the clitic Arabic particle la along with the rendition of the negative adverb hardly can be employed as below: . لصعب على داسب أن عي إلى منا ثا ي، ع و سينان وا ا،لو كنت صعي ا [Were I from Duninane away and free, hardly should any profit bring me back here again] The hypothetical proposition in this translation parallels the degree of assertiveness in the source text. It would leave some room, albeit very little, for Macbeth to be tempted by profit, whereas this room is completely eliminated in Khaffaj's translation, i.e. profit would not tempt Macbeth to come back. Hence, the accompanying nuance is rendered adequately. 67 4.4 would The modal verb would occurs 20 times in the play (12%) and claims 7 zero equivalents out of the 39 instances (17.95%). Taken its frequency in the play into consideration, it emerges as the modal that undergoes deletion more than any other modal (7/39) if we exclude might, which only occurs 4 times in the play, of which two receive zero equivalents. The frequent deletion of would may be explained by the fact that it occupies a weak position on the scale of epistemic modality. However, it does nuance the import of a proposition by significantly lowering the degree of certainty. Therefore, the deletion of this modal will seriously distort the intended message, as can be witnessed in the Arabic rendition of Malcolm's statement below: Though all things foul would wear the brows of the grace, yet grace must still look so. (p. 221 / Malcolm to Macduff) .رغ أن األشيا ال يئ ت ل س صساات ال اا غي أن ال اا تظل فساا وو ت ل [Though the foul things wear the brows of grace, yet grace remains intact and does not change] Unjustifiably, the Arabic translation changes the weak degree of epistemic modality introduced by would into factuality introduced by the Simple Present form (tatalabbasu ' ت ل سwear') instead of the nuanced (qad tatalabbasu ' ص ت ل سmay wear'), thus marring the modality in the text. The modal would in this context allows some room for the non-realization of the proposition in unspecified circumstances, meaning that there might be a situation where 'some foul things do not wear the brows of grace'. A logical reason for such a thing to happen would be that such foul things have grown so powerful and influential that they fear nothing and no one. Consequently they do not need to wear the brows of grace because grace would become what they say or do. Below is a suggested translation where the modality of would, as well as the modality of must (which is not the study item here) is captured: .رغ أن األشيا ال يئ قد ت ل س صساات ال اا فال بد لل اا أن تحافظ على فساا [Though the foul things may wear the brows of grace, yet grace must remain preserve itself[ A simple comparison between Khafaji’s rendition and the one suggested here can clearly show the translation loss caused by rendering English modal verbs by zero equivalents. 4.5 can The modal verb can shows the same frequency as would in the play (20 occurences) and lags only by one instance in the number of zero equivalents (6/39). This modal is often ambiguous between a deontic ability/permission reading and an epistemic possibility reading in which the context plays a key role in preferring one reading to the other. Arabic also manifests the same kind of ambiguity in the use of the modal verbs yumkinu and, to a lesser extent, yastatii u 'can/may' and their kin modal expressions. In negative sentences, both English and Arabic employ can and yumkinu respectively to indicate prohibition (deontic modality) and impossibility (epistemic modality). Thus, the nuances added to the import of propositions by this modal should be maintained in translation. Witness how the nuance of impossibility introduced by the modal can in Macduff's utterance is erroneously disposed of in the Arabic translation below: 68 Not in the legions of horrid hell can come a devil more damn'd in evils to top Macbeth. ( p.223) .ليس ف جحافل الجحي ال مي د شيطان دلعون ف ش ور أكة د دا ث [Not in the legions of horrid hell exists a devil damned in evils more than Macbeth] * * While Macduff conveys the message that the state of affairs in his utterance is impossible to take place, the Arabic translation neutralizes the message by asserting that the state of affairs simply does not exist. In this way, the Arabic rendering dampens the argumentative tone in Macduff's utterance as a result of deleting the modal can. To capture the proper tone, the translator should relay the modality in the ST, as can be seen in the renderings below: . ا ى ف جحافل الجحي ال مي ال يمكن ظاور شيطان فوق دا ث ف ش ور . ا ى ف جحافل الجحي ال مي يستحيل ظاور شيطان فوق دا ث ف ش ور Both renderings can be back-translated as [Not even in the legions of horrid hell can appear a devil that surpasses Macbeth in his evils]. This back-translation reflects the modality-nuanced tone in the ST, which is completely missed in Khaffaj’s translation above. 4.6 will The modal verb will is comparable to shall in its epistemic strength. It occurs 14 times in the play and receives zero equivalents only in two cases, in which the Present Simple form or the Present Participle form is employed. Both forms can functionally replace this modal with a negligible difference in the degree of epistemic modality. Following is an illustrative example: What will you do? (p. 115 / Malcolm to Donalbain) داذا أ ت فاعل؟ [What you doing?] The use of the present participle form faa ilun ' فاعلdoing' nearly captures the use of will in the ST's utterance, which alternatively translates into maaðaa sa-taf alu ' داذا و فعلWhat will you do?' Both renderings capture the intended meaning of the utterance. In this way, will (just like shall) may be replaced with other forms (i.e. the present simple and the present participle) that exhibit a slightly higher degree of epistemic modality. 4.7 could and might The modals could and might have 9 and 4 occurrences respectively in the play and both undergo deletion two times. In the subjunctive mood, could usually carries a combination of ability and possibility reading, whereas might normally takes a possibility reading alone. Both modals significantly shade the propositions in which they are employed, each in its own way, and, therefore, should be rendered in Arabic translation. The following extract instantiates both modals, which receive zero equivalents in the Arabic rendering: 69 … if the assassination could trammel up the consequence, and catch with his surcease success; that this blow might be the be-all and end-all here … (p. 67 / monologue by Macbeth) ... إن كان اوغ يال ؤك الن يج و حي صهل النجا ولو أن مه الح ص م الال والقاضي على الال منا [… if the assassination was trammeling up the consequence, and reap with his surcease success; and if this blow is the be-all and the end-all here] It is unfortunate that the ability/possibility reading of could in the English conditional clause is lost in the Arabic rendering. Therefore, the hypothetical condition is rendered as a real condition in Arabic, thus allowing room for its realization in the future, which can be back-translated as 'If the assassination trammels up the consequence'. To capture the hypothetical condition, the modal could should rendered in Arabic as: ... إن كان يمكن لًغ يال أن ؤك الن يج. Notably, the Arabic modal encapsulates a combination of ability and possibility reading just like the English modal could. For its turn, might in the above extract carries a possibility reading which is completely lost in the Arabic translation; it has been changed into a factuality reading, thus moving from slight possibility to complete certainty. The nuance introduced by might can be captured by a variety of Arabic modal expressions, as can be seen below, among others: ولو أن مه الح ص قد تاون م الال والقاضي على الال منا ولو أ ه من الممكن أن تاون مه الح ص م الال والقاضي على الال منا .ولو أ ه يحتمل أن تاون مه الح ص م الال والقاضي على الال منا The availability of a host of Arabic modal expressions that can relay the slight possibility in the English segment leaves no reason for the translator to render it by zero equivalent whereby the degree of epistemic modality is seriously distorted. Interestingly, the modal may, the twin member of might, which occurs 7 times more than might (i.e. 11 occurrences) in the play does not receive any zero equivalents. It could be that the translator felt that the epistemic degree of may is higher than that of might, hence it must be preserved in Arabic. While it is true that the location of might is lower than that of may on the epistemic English scale, this does not justify the deletion of might in Arabic. As a matter of fact, Arabic does not distinguish between may and might in translation: they both indicate slight possibility. 5. Conclusion Despite the fact that Arabic modality is not as grammaticalized as English modality, Arabic proves to possess a myriad of modal expressions which enhance stylistic variation in translating from English into Arabic. This fact, however, is of minimal value if the translator fails to properly understand the modal system of English and the fine distinctions that each modal displays in context. The data analyzed in this study shows clearly how damaging the nuances of English modality can be when the translator fails to capture them in Arabic translation. In this way, the zero equivalent strategy, which is employed in 39 instances (23.5%) in the play, proves 70 seriously inadequate in most cases when put under close scrutiny. The only exceptions are the epistemic English modals will and shall, which may be replaced by the Arabic Present Simple or Present Participleforms without affecting modalistic realities. 71 Semantic/Syntactic Hurdles in Machine and Student Translation Mohammed Farghal & Adnan Gergeos Abstract The present article addresses itself to some subtle English syntactic features that present formidable problems to both Machine Translation (MT) and Student Translation (ST) into Arabic. First, it deals with the recognition of variation in parts of speech, showing that polysemous and/or homonymous syntactic categories may cause serious interpretation problems to both machine and student translators. Next, structural ambiguity at phrase and sentence level is shown to create interpretation difficulties in translation activity. Third, parenthetical structures in English sentences are argued to trigger reference and focus mishaps in MT and ST alike. Finally, the paper offers ample evidence from MT and ST indicating that word order variation affecting the basic English word order SVO may cause fatal damage to both machine and student translators. In particular, marked stylistic inversion and subject deletion prove problematic in this respect. 1. Introduction Despite the tremendous progress that has been made in the field of Machine Translation (MT), or what is commercially known now as Translation Software (TS), in the past 50 years or so, human aid by professional translators is still required to post-edit and revise poor quality outputs of MT texts that are not suitable for publication (Meijer 1993, Feder 2003). It is well known that computer-aided systems of translation cannot aspire to gain authenticity and universal acceptability without the help of professional human translators (Hutchins 2001); hence the initial skepticism and fear of MT replacing and discarding human translators is by far a mere illusion. With the-state-of-the-art technology and progress made in the domain of information and computer sciences, it has become a necessity to make use of the available tools to save time and cost by utilizing MT to produce draft assimilated compatible versions of the original texts in the Source Language (SL) that still need the expertise of the human translator to revise and make them ready for publication. This paper doesn't aim to analyze or evaluate the mechanisms and shortcomings of the existing systems of MT; neither does it try to find answers and solutions to the difficulties faced by such computer-aided systems of translation. Rather, it aims to survey and show the semantic/syntactic complexities and subtleties of natural language that may obstruct the perfection of advanced systems of MT and hinder the performance of student translators trying to find their way into the profession of translation. Therefore, wherever possible, a comparison is made between the 73 machine’s output and the student translator’s output in English into Arabic translation. The corpus in this study is drawn from MT outputs of some Internet online services that provide immediate translations of miscellaneous short sentences and texts (e.g. www.ajeeb.com, www.akuwait.net), and authentic translations made by senior Arab students in the College of Languages and Translation at King Saud University and the Department of English at Kuwait University. Despite the fact that computer-aided systems of translation may be able to implement tools to handle difficulties related to compounding, idiomatic expressions and metaphors by treating them as fixed units (Hutchins 1991), some syntactic features are still causing interpretation problems for computers and student translators alike (Lewis 1997). It is interesting to note that the capability of MT can be compared to the developing translational competence of student translators. It will be shown that subtle syntactic knowledge involving parts of speech, structurally ambiguous phrases and sentences, parenthetical structures and word order variation can pose serious problems to both MT and student translators. 2. Parts of Speech The successful identification of parts of speech (i.e. N, V, Adj, Adv, etc.) in English discourse is an important aspect of reading comprehension, which is a pre-requisite for professional translation activity. Conversion, which is a highly productive word formation process in English, doesn’t involve the introduction of affixes when changing the part of speech of a word, e.g. dirty (adj) and dirty (V), call (V) and call (N), etc. The recognition of the part of speech in English discourse, which is dependent on the accompanying co-text and context, proves impenetrable to MT and some student translators. Consider the example in (1) below, along with its MT in (2) and student translation (ST) in (3): (1) And H represents Higher Order needs, and covers self-esteem, expectations, ambitions and sense of humor. (2 ( 3 ) وH . لل عاص . ) وإتش اةل ااجات ال تيب األعلى واا ام الهات األغطي ال وصعات الطاواات ورو ال ّ عاص طلكككب اوا ياجكككات وال غطيككك وا ككك ام الكككهات وال وصعكككات والطاكككو واسككك فااكككه وإوراككككه،تاةكككل ارتفكككا As can be seen, both renditions fail to recognize the correct part of speech of the word covers and, as a result, give an inaccurate translation. The confusion comes from the fact that the word covers can potentially function as a plural noun or as a verb in the present tense form. However, the syntax of the sentence in (1) makes it clear that covers is used as a verb rather than a noun. Syntactically, there should not have been any sort of confusion since the conjunction and precedes the word covers, giving rise to the paralleled coordinated verbs ‘H represents… and covers…’. It should be noted that while the ST adopts an erroneous linear/literal rendition of the 74 phrase ‘covers self-esteem’, the MT takes the second item as the head noun, hence the impossibility of the MT rendition of the said phrase. This may point to a serious need to sensitize MT systems to the ’iḍaafah construction in Arabic whereby the impossible MT rendition above would be replaced with a possible, though erroneous, rendition, viz. أغطي إا ام الهات. The failure to identify the correct part of speech resulting from conversion as in (1) above may be further complicated by the existence of English homonyms, which are identical word forms that do not relate to each other in meaning. The following example in (4), taken from a newspaper political article, along with its identical MT and ST in (5), bears witness to this: (4) Much lies behind those words. (5.خلف مه الالاات ) أكاذ ب كةي On the one hand, the word lies may function as a verb and a noun (which mean ‘not to tell the truth’ and ‘the plural noun of lie’, respectively), just like covers in example (1) above. However, both interpretations are irrelevant in example (4). On the other hand, the relevant interpretation stems from a homonym of both forms, viz. lies, meaning ‘is found’, which comes from a completely different semantic field. Clearly, this further complicates the task of MT and ST, which were entangled within one of the irrelevant interpretations, being insensitive to the syntax of the sentence. The syntax of (4) calls for three consecutive steps on the part of the interpreter, whether it is a machine or a human analyzer. First, the sentence should be broken down into a Subject and a Predicate. Second, the analyzer should be aware of selectional restrictions rules which require a non-count noun after the quantifier ‘much’ in order to rule out the interpretation of ‘lies’ as a noun. Finally, the verb (head) in the predicate should be assigned the correct sense, i.e. ‘is found’. It should be noted that human professional translators have managed to automate such conscious syntactic knowledge in translation activity. However, much work is still needed in both MT and ST in this area. Further, problems may arise within the same part of speech rather than across parts of speech as in the examples above. For example, capitalization, which is an important layout feature in English, may neutralize the difference between a proper name (which is always written with a capital initial regardless of its position in a sentence) and a common noun (which can be written with a capital initial only at the beginning of a sentence). The example in (6), along with its MT and ST in (7) and (8), illustrates this point: (6) Bush fires raging around the Australian capital Canberra have killed three people and destroyed hundreds of suburban homes. (7 اول كا ي ا العاصا األو اليّ ص ص لت ثًث أشخاص و و ّد ت الائات د ال يوت ف . الحّااي 75 ) ا ائق صوش ال ّش (8 را ضحا اما ثًث أشخاص وصعت ف العاصا األو الي كا ي ا ال ) غحب ال ئيس صوش صشأن الايي ال .وود ت فياا دئات الانارل الواصع صحوااياا As can be noted, the coincidence between the common noun bush with a capital initial at the beginning of the sentence and the popular proper name Bush (the present President of the United States) creates serious confusion for both the machine and student translators. Both MT and ST in (7) and (8) above base their translations on an irrelevant interpretation of the noun Bush, being misguided by the capital letter in the noun. The MT output, accepting things at face value regardless of congruence with world knowledge, suffers solely from that fatal decision. The ST output, however, runs into more serious problems as a result of rejecting a literal rendition of the lexeme fires in combination with the erroneously selected proper name interpretation because of its clear oddness insofar as world knowledge is concerned. Consequently, the student translator creatively improvises a far-fetched, though sensical, metaphorical interpretation of the lexeme fires, something which is, apparently, inaccessible to the machine translator. Interestingly, when given the sentence with an initial small letter of the noun bush, the machine produced a fine output as shown in (9) below: (9 اول كا ي ا العاصا األو اليّ ص ص لت ثًث أشخاص و و ّد ت الائات د ال يوت ف . الحّااي ) ا ائق األا اش ال ّش 3. Structural Ambiguity Structural ambiguity manifests itself at both phrase and sentence levels in natural language. At the phrase level, noun compounds in English may contain head nouns modified by other nouns functioning as quasi-adjectives. Problems may arise from the lack of explicit markers of such relationships, because in many other languages like Arabic these relationships are usually expressed by case endings or prepositions. For example, in the English sequence ‘adjective-nounnoun’ in (10), the adjective can modify either the first noun (11) or the second noun (12). (10) New computer software (11) [ [New computer] [software] ] (i.e. software for a new computer) (12) [New [computer software]] (i.e. new software for a computer) In languages which are highly inflectional like Arabic, such phrases are usually not ambiguous because they are often clearly marked for number and gender as follows: (13) baraamija ḥaasuub-in jadiid-in 76 software-f-pl computer-m-sg-gen new-m-sg-gen "Software for a new computer" ḥaasub-in (14) baraamija software-f-pl jadiid-at-in computer-m-sg-gen new-f-pl-gen "New software for computers" As can be seen, the gender agreement on the Arabic adjective disambiguates the phrase above by indicating gender agreement with either the singular noun or the plural noun, hence the two readings. Another ambiguity-creating sequence in English is the phrase featuring ‘Adjective-nounconjunction-noun’, e.g. ‘tall boys and girls’. Although it is possible to resolve certain ambiguities in the SL by semantic features and roles as well as syntactic information, some cases still require the knowledge about the things and events being referred to. The context and our knowledge of the world around us are extremely important tools for the translator that may help clarify certain rather ambiguous situations. Predictably, basic world knowledge proves accessible to student translators but opaque to machine translators. The example in (15), along with its MT and ST in (16) and (17), illustrates this point: (15) Pregnant women and men (of course only women can be pregnant) (16)النسا وال جال الحادًت (17)النسا الحوادل وال جال However, lack of more sophisticated world knowledge and insensitivity to the context may cause some student translators to produce far-fetched, even ridiculous, translations, as can be illustrated in (18), along with its ST in (19) below: (18) The Occupied West Bank (19)ال ن الغ ص الاشغول In a passage talking about Palestinians and their occupied lands, the denotation of the phrase in (18) above should be crystal clear. Lacking appropriate interlingual socio-political world knowledge in the said context, however, these student translators fell victim to fatal translational mishaps. 77 At the sentence level, it is a well-known fact in modern linguistic theory that grammaticality is never determined by meaning, yet meaning is highly dependent on grammar. Translation, be it an art, a skill or a science, is mainly concerned with the process of transferring relevant meaning (Gutt 1996; Farghal 2004) from a SL into a Target Language (TL). Aside from the mastery of the two languages involved in the process of translation, prior syntactic knowledge constitutes an essential element. Although a given context may illustrate the required meaning, it may, however, fall short of explaining certain types of ambiguity which need a great deal of experience in translation activity. Structural ambiguity at the sentence level is not always captured by machine and student translators. In many cases, polysemous and/or homonymous words enter into ambiguous syntactic relations in sentences, as can be illustrated (20) below: (20) We can fish. Both student and machine translators managed to offer a translation congruent with the more common reading of the sentence, i.e. ‘We can go fishing’ rather than the less common interpretation, i.e. ‘We put fish in cans’. However, MT as well as many student translators could not cope with the relevant interpretation when the sentence was disambiguated, as can be illustrated in (21) below, along with the elicited translation in (22): (21) We can fish in a small factory. . ) اا أن يطاو ف دين صغي22( Despite the fact that the co-text in the sentence in (21) rules out the interpretation ‘We can go fishing’, the MT and many student translators were not able to grasp the difference between can as a modal verb, and can as a lexical verb; hence the wrong translation. Syntactically, the source of ambiguity in (20) can be schematized in (23) and (24) below (Georges & Barakat 2000): (23) We can fish S Modal V Lexical V can fish Lexical V NP (24) We S 78 These two different structural representations give rise to two different semantic interpretations as well. In (23) can (a modal auxiliary verb) and the lexical verb fish form the combination meaning ‘the ability to go fishing’, whereas in (24) can is a transitive lexical verb that subcategorizes a direct object fish, hence the meaning ‘putting fish in tins or cans’. Despite the fact that fish in both interpretations occupies the same structural position after can, yet it is part of the verbal slot in (23) and outside the verbal slot in (24), occupying the nominal object position. This clearly shows that words occupying similar structural positions do not necessarily have the same grammatical functions. This example indicates that it is in fact possible for homonyms to be disambiguated by syntactic analysis, i.e. without using any semantic information. Apart from structural ambiguity proper, syntactic operations may render sentences opaque to both machine and student translators. Consider the syntactic operation affecting the English collocation pay a visit in (25) below, which proved problematic for MT (26) and many student translators (27) and (28): (25) This came during a field visit paid by Sheikh Mohammed to the General directorate for patrols and the General Security Sector…… (26) ... مها جا أثنا ر ار دي ان د فوع د ص ل شيخ (27) ... جا مها أثنا ر ار د فوع ال االيف د ص ل (28)... ايل ذل خًل ر ار اقل صام ص ف ثانه الشيخ دحا However, when an active voice sentence (29) involving the same collocation was administered to the machine and student translators, the output was correct, as can be seen in (30) below: (29) He paid a visit to his neighbors. ( 30.) صام صز ار لجي ا ه Apparently, the syntactic operations of passivization and relative clause reduction in (25) are to blame for the erroneous renditions in (26)-(28) rather than the collocation pay a visit itself, as (29) and (30) clearly show. Once again, syntax interferes with the interpretation of semantic units that are otherwise straightforwardly understood. 4. Parenthetical Structures 79 The corpus shows that reference interpretation in sentences including parenthetical structures may be problematic to both MT and some student translators. The example in (31) gives rise to misinterpreting reference relating to a non-defining adjective phrase relating to the subject of the main clause, as is clear in the sample translation (32): (31) Al Saqqaf, publisher and editor-in-chief of the Yemen Times, which is seen as vanguard of press freedom in the country … ّ طليع ا ّ ّ ال (32... يحاف ف ال ل ع اليا تا از اله the النّاش و رئيس تح,)) الساكااف(السقاف As can be seen, the relative pronoun which is mistakenly interpreted as referring back to the antecedent Al-Saqqaf, which lies outside the parenthetical phrase, instead of the correct antecedent Yemen Times, which is found within the parenthetical phrase. Also, this reflects lack of semantic knowledge on the part of machine and some student translators that the relative pronoun which can only refer to inanimate referents such as Yemen Times. Further confusion may be induced by the erroneous use of the subject of the parenthetical phrase as the subject of the main clause in both MT and ST. Consider the MT in (34) and the STs in (35) and (36) of the sentence in (33) below: (33) The tea, first discovered in China 5000 years ago, has long been thought to be beneficial to health. ّ ص اُ ْع ُقب أن اون دفي لا ّ طو ل إلى ال, ون1666 ف اليّي دنه، أ ّوو, ) اك شف ال ّشا (34. ي ّح (35. ون1666 ع ق أ ه دفي لليح اك شف ف اليي ص ل (36. عام1666 ف اليي دنه ) الشا اله ع ق أ ه دفي لليح ألول د ) اك شف الشا اله Other things being equal, both the MT and the sample STs above twist the focus in the packaging of information by using the non-finite verb in the subordinate phrase, i.e. discovered as the verb of the main clause instead of the correct verb has been thought. These examples show clearly that both MT and ST still need further work in the area of subordinated structures. Before closing this section, below is an example in (37), along with its MT in (38) and a sample ST in (39), which shows that subordinate clauses may cause wrong gender agreement in MT, but not in ST: 80 (37) The baby's father, she said, would soon travel to Brazil to offer scientists proof that the baby's DNA was identical to that of a deceased sibling. (38 إلى ذل،ًإ ه الطّفل كان د ااث.إن. إلى ال ار ل لع ض العلاا ال ّ ليل أن و، و ساف ص ا, صالت,) أ الطّفل . راال ألخ ٍ (39 ) صالت إن أ الطفل ويساف إلى ال ار ل ليع ض على العلاا أن " و أن أ ه " الطفل د طاصقا د ص ه .ال اال Because of the intervening clause, there was a gender agreement or concord mismatch between the subject and the verb of the main clause in the MT in (38). Being transparent to native speakers of Arabic, however, gender agreement was not affected by the parenthetical clause in the ST in (39). 5. Word Order English is mostly a configurational language in which the position of the constituent mainly indicates the grammatical function; hence the major claim by Transformational Grammar that functional labels such as Subject, Predicate, Direct Object, etc. are effectively derivable from categorial labels such as NP, VP, PP, etc. (Chomsky 1965 and his subsequent works). Therefore, any change in the English basic word order, i.e. SVO is the result of transformations that belong to the transformational component which supplements the base component in generating English grammatical sentences. Among these transformations, inversion constitutes one of the more common syntactic operations in English. On the one hand, inversion is highly standardized and straightforward in Yes/No questions and Wh-questions; hence it poses no challenge to MT and ST. On the other hand, inversion applies less frequently in stylistically marked constructions including conditionals and negative expressions, and consequently creates serious challenges to MT and ST alike. Similarly, informal subject deletion in declaratives and initial gerundive and participial phrases may create serious interpretation problems in both MT and ST. First, Type II and Type III English conditional sentences may stylistically delete the conditional marker if and, as a result, Subject-Auxiliary Inversion (SAI) obligatorily applies. This syntactic operation confuses both MT and ST, as can be illustrated in (40) and (42), along with their MT in (41) and ST in (43), below: (40) Had it not been too late, I would have accompanied you. (41. كان اا أن أصاا،) كان ل ى مو ليس كان د أخ ج ا (42) Had the president known this, he wouldn't have given them the permission to do 81 so. (43.) ع ف ال ئيس مها ل ا د الااا أن عطيا اإلذن لعال مها As can be seen in (41) and (43), the if-deletion and the subsequent SAI confused the machine as well as some student translators, resulting in nonsensical renditions of the first part of the conditional sentences. Predictably, conditional sentences with the ordinary word order created less serious problems to the machine and student translators. By way of illustration, consider the MT rendering in (45) of the conditional sentence in (44) below: (44) If he had known this in advance, Kofi Anan wouldn't have agreed to send the fact finding team to Jenin camp. (45. كوف عنان ل ا د الااا أن فق أن ول ف ق تقي الحقائق لاعسا جني، ) إذا ص ع ف مها دق دا Although the MT translation in (45) suffers many grammatical and lexical errors, it can be readily edited into an acceptable Arabic translation, as the rendition makes it clear that the SL text is a conditional sentence. By contrast, it would be so difficult for the editor to figure out that the translations in (41) and (43) above are conditional sentences. As is clear, the cause is the inversion in (44). Second, sentence initial negative expressions obligatorily trigger SAI. This contrasts with the unmarked position of negative expressions which are attracted by the auxiliary verb in the SVO word order. The MT rendering in (46) of the sentence in (44) is only illustrative: (46) No longer able to sustain her story, she admitted owning two more buildings in working class districts of the capital, acquired over a lifetime of begging in the streets. 47) ف دقاطعات ط ق الع ّاال للعاصا ايلت .،د عا ال س ّول ف الشوار ) صاور ل ع أن ت ق على صي اا اع فت صاد ًك د نيي آخ أكة Other things being equal, the gender agreement confusion at the beginning of the Arabic translation is an immediate consequence of the sentence initial negative expression in (46). That 82 is, the negative expression no longer able to is not marked for gender in English the way the noun phrase her story is, later on in the fronted adverbial phrase no longer able to sustain her story. As for the erroneous inversion performed by MT in (47), it may have to do with the presence of the non-finite subordinate clause. This may explain why MT does not operate inversion in (41) above where a finite subordinate clause is employed. However, inversion proves fatal to MT in both cases. Third, sentence initial gerundive and participial phrases can cause problems to both MT and ST. The following participial phrase in (48), along with its ST in (49) and MT in (50), bears witness to that: (48) Given their supernatural powers, it is not surprising that jewels have deep religious significance in India. (49. ) على اع ار ص راتا الخارص للط يع فا غي الاهمل أ ا الاون جوام ذات أماي و ني ف الان (50. إ ّه غي دهمل أن ل ى الجوام دغزى و ن ّ عايق ف الان, ) على اع ار ص راتا الخارص للطّ يع Both the ST and MT above misinterpreted the cataphor their by failing to cater for correct gender agreement with the referent in the main clause, i.e. jewels. Even worse, the ST, having committed the said mistake, fell victim to another interpretation problem whereby gender agreement is erroneously created in the main clause, i.e. أ ا الاكونon the basis of the misinterpretation of their in the sentence initial participial phrase, thus offering a worse interpretation than the MT. Apparently, the ST was sensitive to congruence in gender agreement in the participial phrase and the main clause, whereas MT was not. This may be explained by the fact that humans can backtrack and subsequently interpret forthcoming segments relevantly, while machines can do things consecutively with little or no backtracking taking place in interpreting pronominals in discourse, hence the incongruence between صك راتاin the participial phrase and لك ى الجكوامin the main clause in (50). Further, it should be noted that anaphoric reference is more natural and common than cataphoric reference in Arabic discourse. As a result, it would be a good idea to instruct both MT and ST to change cataphoric reference to anaphoric reference when translating from English into Arabic. This discoursal mismatch may have caused some, if not most, of the confusion in the reference interpretation in (49) and (50) above. 83 Finally, informal subject deletion in English declaratives/interrogatives may bewilder both machine and some student translators. The example in (51), along with its MT in (52) and some sample STs in (53), (54) and (55), illustrates this: (51) Sounds complicated? Actually, it isn't as difficult as it may seem. (52.و (53.و مو ليس كيعب كق ) ص وت دعق (عقّ ت الوص ) ؟ ف الواص ) تعقي ات اليوت؟ ف الواص صأ اا ليست صيع كاا (54. ) ردور دعق ؟ ف الواص م ليست صع كاا تظا (55.و ) تعق اليوت؟ ط يع إ ه ليس د اليعب كاا Ignoring the second part in the text above, the deletion of the dummy subject it in the declarative Yes/NO question rendered it incomprehensible to the machine as well as the student translators. Looking at the poor quality of the translations in (52)-(55), the drastic change in the word order seems fatal. In the case of MT, the machine should be made aware of subjectless declaratives/interrogatives, in addition to the familiar imperatives. As for student translators, their trainers should bring to their attention such marginal structures that can be optimally natural in conversational English. Such informal features, it should be noted, are more relevant to interpreting than translating. 6. Conclusion This paper has addressed itself to some syntactic areas with which MT and ST may experience difficulties. It has been shown that polysemous and/or homonymous parts of speech, structural ambiguity at phrase and sentence level, reference interpretation in parenthetical structures and word order variation may cause serious translation mishaps in MT and ST alike. In many cases, these problems produce incomprehensible renditions where heavy post-editing is needed. An important finding of this study is the striking similarity between the syntactic errors committed by the machine and student translators. Apparently, both groups seriously lack sufficient syntactic knowledge relevant to syntactic operations which affect unmarked constructions in English, thus rendering them stylistically marked and, consequently, impenetrable to the interim translational competence in MT and ST. Whereas future technological development induced by human intervention is expected, in theory at least, to remedy the MT deficiencies in this regard, translator trainers should shoulder the responsibility of bringing such syntactic subtleties to the consciousness of their student translators. 84 Finally, this study may offer us some insights into the modes of human reasoning and mechanical (machine) reasoning in translation activity. On the one hand, subtle syntactic processes proved fatal to both MT and ST. On the other hand, student translators demonstrated an awareness of basic grammatical and world knowledge such as gender concord and primitive coherence which proved opaque to machine translators. Apparently, the human brain’s ability to backtrack opens some reasoning avenues that are not yet available to MT. Therefore, any genuine development in future MT must take this important mismatch into consideration. Translational Miscues in Modern Arabic Verse Mohammed Farghal & Rula Naji Abstract This paper addresses itself to the notion of ‘miscue’ from a translational perspective. In particular, it constitutes a case study dealing with miscues in translating Modern Arabic Poetry into English. It is argued that oral reading miscues during the decoding process in translation manifest 85 themselves in written traces which may be called ‘translational miscues’. In effect, the paper systematically demonstrates that translational miscues may occur at the phonological, morphological, syntactic, and semantic levels, giving ample examples from the translation of Arabic poetic discourse into English. Throughout, it is emphasized that poetic discourse may fall victim to translational miscues, which may twist, or even cripple, the cognitive structures communicated by poets. This being the case, the translator’s ability to establish an informed and intimate rapport with the poem s/he is translating remains the sole possible guarantee for genuine translatability. 1. Concept of Miscue Historically, the term ‘miscue’ was first employed by Goodman (1962) where he develops a system for comparing expected oral reading responses with observed oral reading responses. His system is concerned with meaning and how it changes when miscues occur both at the word and sentence levels. To show the importance of miscues, he argues that the written language becomes parallel to the oral language in a literate society. To him, reading (the process through which miscues occur) is a psycholinguistic process involving the use of language in its written form to get to meaning, which may be twisted by misreading a word or a sentence. Subsequent work has focused on the interaction between different cue systems in the reading process (i.e. phonetic, graphic, syntactic, and semantic systems). Chafe (1970) views the reconstruction of meaning as the function of all these systems together, arguing that attempts to separate them are merely a linguistic tradition. Similarly, Burke (1976) explains that it is impossible to deal with any of these systems without addressing the others since they are interrelated. For his part, Sims (1976) states that substitutions, insertions, omissions, or reversals, which cause the most frequent miscues, may occur at the sound, morpheme, word, phrase, and clause levels, arguing that (p.49) “In order to gain a fuller understanding of the reading process and of the miscue phenomena, one must examine miscues, not simply as changes involving individual words, but as phenomena which occur in a total language context” (cf. Montoro 1976). Also, there has been emphasis in the existing miscue research on the interaction between the reader and the text. Rousch (1976) argues that words in the reading process possess a semantic relationship in the mind of the reader in a way such that he will anticipate the forthcoming word from the previous one, with the possibility that the reader may come up with a substitution that has no semantic relationship with the expected response. He concludes that (p.55) “The preservation of meaning through restructuring plus substitutions that vary from zero to close semantic relationship with the original reflects a strong awareness of meaning without close attention to individual words.” Along the same lines, Page (1976) considers reading as a process in which the reader attempts to reconstruct the author’s meaning. According to him, meaning in the context of miscue analysis is treated as “analogy, or the relationship between the reader’s idea 86 and what the idea represents”, emphasizing the interrelation between meaning and knowledge during the comprehension process. Finally, Durkin (1978) considers any deviation from what is printed as a miscue because deviation alters the meaning of words and sentences. She divides miscues into four types: substituting one word for another, inserting a word, omitting a word, and inability to identify a word. 2. Translational Miscues Translating involves many types of problems, some of which relate to reading, which is, according to Encarta 97 Encyclopedia, “An activity characterized by the translation of symbols or letters into words and sentences that have meaning to the individual.” Therefore, the ultimate goal of reading is to be able to understand written material, to evaluate it, and to use it for one’s own needs. Translational miscues may be viewed as the written traces of what is assumed to have been oral miscues during the reading process. They may be phonological, morphological, syntactic, or semantic. It should be noted that the area of translational miscues is still virgin ground as the existing miscue literature is dedicated to examining miscues in terms of education rather than in terms of translation. This paper, therefore, is meant to bridge this gap by systematically presenting a translational account of miscues, which are posited to form a subcategory of translational mistakes (i.e. mistranslations) and a completely different category from translational errors (Pym 1992: 178-9). The question that poses itself now is: How do translational miscues form a special subcategory of translational mistakes? To answer this query, one must look for a procedural distinction. On the one hand, translational mistakes result from conscious decision-making while the translator is not aware of the fact that s/he is committing a mistake. Translational miscues, on the other hand, are due to an accidental and/or subconscious phenomenon that involves minimal decision-making while the translator is not aware of the fact that s/he is committing a miscue. In this way, a translational miscue must be based on a mistaken oral response during the decoding process in reading, while a translational mistake may result from a host of other factors such as deficiency in language competence and inability to grasp a proposition. Translational miscues may occur both at the word and the phrase/clause levels. At the word level, they may result from misreading one letter or more in a word, or from misassigning phoneticallyunrealized vowels within a word, or from calling up a sense of the word that is incongruent with the context in question. The situation at the phrase/clause level is similar as miscues may bring about serious distortions of the meaning by assigning erroneous grammatical relations which are usually signified by phonetically-unrealized case markers in Arabic or by invoking irrelevant composite senses of phrases/clauses. 87 The present study addresses itself to translational miscues in modern Arabic verse, namely in a collection of poems entitled Victims of a Map, which consists of a select group of poems by three renowned contemporary Arab poets, viz. M. Darwish, S. Al-Qasim, and A. Adonis as translated into English by Abdullah Al-Udhari (1984). While translational miscues may turn out to pertain to translating from Arabic into other languages in general as Arabic is highly rich in phoneticallyunrealized voweling and/or case-assignment and an elaborate network of lexical ambiguity (among other sense relations), the translating task will even be more complex and challenging when it comes to modern Arabic poetry. Ismail (1967: 174) has the following to say about this complexity: [our translation of quotations below] The orientation of the neo-poetic movement experience deals with language in a special and innovative way. Such a creative use of language in modern Arabic verse is expected to furnish rich soil for translational miscues. In this spirit, Pickering and Hoeper (1982: 570) write that “No word in great poetry can be moved or replaced without harming or changing the whole” and that (p. 568) “This careful use of language [in poetry] is the most significant difference between ordinary prose and poetry.” Moreover, modern poetry is viewed as a new experience. The modern poem, argues Adonis (1978), is a case rather than a mere group of words and meanings. Similarly, Lewis (1969: 113) contends that “The modern poet then is faced with a difficulty of communication as great in its way as the difficulty presented by his subject matter.” Thus, unlike non-poetic discourse, the importance of Manner may greatly exceed the importance of Matter in verse. To this effect, Boutlon (1977: 152) writes: Poetry is made of words and obviously, the choice of words is important in poetry; indeed, in a sense it is the whole art of writing poetry. 3. Phonological Miscues Sims (1976: 46) defines submorphemic (i.e. phonological) miscues as “those which involve changes of one or two phoneme sequences within a word.” The reader may substitute one vowel sound for another, e.g. care for car and gave for give, or one phoneme sequence for another, e.g. kisser for killer, or omit or add a sound, e.g. right for bright and see for seed, or make a reversal of the phonemes, e.g. skate for steak and kiss for sick. This type of miscues is very common in reading Arabic, especially in vowels where we have long and short vowels enjoying a wide distribution. Consequently, we find a large number of words featuring the same radicals but different vowels. For instance, the word ḥilm means ‘patience, 88 forbearance, discernment, and gentleness’, while the word ḥulm means, ‘a dream’. It should be noted that such vowel differences are mostly unrealized phonetically and if they are phonetically realized, they will be superscripted or subscripted depending on the intended lexical meaning. Vowel miscues may seriously distort the intended meaning in the translation of verse as can be illustrated by the following example taken from Adonis’s poem ‘ دزدورPsalm’, along with its English translation: . و قل ال ح د داا ه، ق ل أعزل كالغاص وكالغي و ُ ُّو وأدس اال صا ّر ن ظ دا و . أت و س عي اها الليل ث،ين د ص ديه اارا ، حيا و حلِّ ُل اليأس داايا- والظل د ن حيا فسح .وللشج ك نام لل ا ك ةا، األدل راصيا و صفا الناار ُ ايث يي الحج صحي .سحر ّ على ج ي عي ا عًد ال، وما مو عل تقاط األط اف اصشا He comes unarmed like a forest, like a destined cloud. Yesterday he carried a continent and changed the position of the sea. He paints the back of day and creates daylight out of his feet, borrows the night’s shoes and waits for what will not come. He lives where the stone becomes a lake, the shadow a city - he lives and fools despair, wiping out the vastness of hope, dancing for the soil so it can yawn, for the trees so they can sleep. And here he is speaking of crossroads, drawing the magic sign on the forehead of time. 89 In this poem, Adonis intratextually (for more on inter-, intra-, and contratextuality, see Martin 1985, Lemke 1985, and Hatim 1997a) creates a vivid image where his subject performs extraordinary actions, viz. “carrying a continent”, “painting the back of day”, “creating daylight out of his feet”, “fooling despair” and “wiping out the vastness of hope”, among others. This intratextual weaving of the image culminates in his subject’s “drawing the sign of daybreak on the forehead of time”, thus intertextually establishing an oft-cited association between a physical phenomenon (daybreak) and a spiritual phenomenon (hope), which, together, function as a declaration of freedom in the Palestinian context. Being unaware of this poetic discourse, the translator fell victim to a phonological miscue (substituting [i] for [a] whereby he misread the key word [ السحal-sa ḥar] ‘daybreak’ as ْ[ السِّحal-siḥr] ‘magic’, and subsequently missed the intended meaning altogether in his translation. Sometimes, the phonological miscue constitutes the replacement of a consonant with another, as can be illustrated by the following stanza from Adonis’s poem ‘ أغنيA Song’, along with its English translation: د ال خور واليّن ل، عطين ورص أُاالُاا أك اوا د أُ قِّطاا كالع وس وأجْ لوما أ ْص أ علياا وور د ُّ أمز فوصاا جذوعي د ال ّشوق والحل وأرْ بولُاا إلى أا اص دليئ كال فّاا ! خفيف وخحْ ا ك ُا ْا ب الخح Who will give me and sandalwood, a piece of paper Adorn it with dots like a bride’s make-up, polish it, Bless it with the Koran’s praise of the Madonna, Shake my roots of longing and dreams over it And send it to the loved ones, Full like an apple, Fine and green like Khadir’s colt. 90 to wrap incense In this stanza, the translator fell victim to a phonological miscue by replacing the pharyngeal consonant [c] with [r], thus translating juðuuc ‘branches’ as juðuur ‘roots’, and consequently disrupted the coherence of the poetic discourse, which is established through intertextualizing with the highly celebrated Quranic verse in which Allah asks Maryam [Virgin Mary] to shake the branch of the palm tree in order to get fresh dates, viz. جنيا، النخل تساصط علي رط ا،ومز إلي صجه (Maryam: 25). Not only did the translator obliterate this intertextual web, but he also failed to furnish coherence in the poetic discourse as, in the most creative poetic context, one cannot imagine the shaking of roots over something, for roots are not to be shaken and if ever they were, the possibility of shaking them over something else is eliminated due to their being located underground and/or deep inside. The same phonological miscue occurs in Al-Udhari’s translation of the following stanza from Darwish’s poem “ ساف كالناسWe Travel Like Other People”, where a metaphor that is developed intratextually is disrupted: ّ . كأن الس ْف ساف ُ كالناس لاننا و عو ُو إلى أ ش ُ ط ْ وفنّا أ با نا ف ظًم الغيوم وصي جذوع الشج.ق الغيوم ْ لب ْ ن دنا دئات السني لناال مها ال ايل:وصلنا لزوجاتنا . ْإلى واع ٍ د صً ٍو ود د الاس حيل We travel like other people, but we return to nowhere. As if travelling Is the way of the clouds. We have buried our loved ones in the darkness of the clouds, between the roots of the trees. And we said to our wives: go on giving birth to people like us for hundreds of years so we can complete this journey To the hour of a country, to a metre of the impossible. It should be noted that the poet in the above stanza wants to communicate the message that the action at hand is not an ordinary burial in that the loved ones are not buried underground (i.e. between the roots of trees) but rather in the darkness of clouds and between the branches of trees. This creative poetic metaphor is designed to show that formal attributes of human activities may not signify their authentic nature, e.g. ‘travelling’ as a human activity would be genuine only and only if ‘the traveler’ would return to a free homeland, which does not obtain in the Palestinian context, to which the poet is schematically making reference. In effect, it is the intratextual twinning of the out-of-the-ordinary poetic image and the socio-cultural realities of the 91 communicative context that brings meaningfulness to the above text. This being the case, the translational miscue in this poem amounts to offering a fleshless skeleton insofar as poetic discourse is concerned. Our last example of phonological miscues comes from Adonis’s poem اليح ا (دخ ارات د وديات )3582 ‘ ايار صي وتThe Desert (The Diary of Beirut Under Siege, 1982)’ and involves the substitution of the glottal stop [’] in al-’amal ‘hope’ for the [r] in al-raml ‘sand’, as can be illustrated below: ، شج نحن ليقول وواعا ، ف ح زمو ُن ّاس أوراصه ليقول وواعا رم ، ت تقول وواعا ط ق كالفواصل صي ال نفس والالاا ب ب ، جس ل س الرمل سقط ف تيا به ليقول وواعا ُ ورق عش ق الح ، والشع ا قول وواعا واألصج .، والقيي صالت وواعا Trees bow to say goodbye Flowers open, glow, lower their leaves to say goodbye, Roads like pauses between the breathing and the words say goodbye, A body wears hope, falls in a wilderness to say goodbye, The papers that love ink, The alphabet, the poets say goodbye, And the poem says goodbye. The stanza above demonstrates the poet’s deep pessimism by making everything bid farewell as if an imminent, overall disaster were to take place. In such an atmosphere, there is no room for ‘hope’, viz. ‘A body wears [wearing] hope’, which can only be an indication of a vision of life, despite the fact that the original phrase, viz. ‘A body wearing sand’ can only be an indication of ‘death’. This being the case, the phonological miscue inadvertently replaces a thought-world of death with a thought-world of hope. 92 4. Morphological Miscues The morphological miscues discussed in this section belong exclusively to inflectional morphology. It should be noted that inflectional morphemes may constitute an integral part of poetic discourse in that they may be thematically employed to communicate ideological moves. Consequently, the translator of verse must be aware of the nuances of meaning deployed by such grammatical elements. To get started, let us examine two translational miscues relating to singularity vs. plurality in poetic discourse. The following extract is taken from Darwish’s poem “ إذا كان ل أن أعي ال اIf I were to Start All Over Again”. ُ أخ ا ُر دا اخ ورو السياج:ت .ص تُ ّؤو وص و تُؤو إلى ص ط ه ّ على غي ب ظل، عشا إذا كان ل أن أُعي ال ا ف الدروب ال، أواف ثا ي ُ ّأُعل ق ظل على صخ تي ل ن الطيو ُر الش If I were to start all over again I’d choose what I had chosen: the roses on the fence. I’d travel again on the road which may or may not lead to Cordova. I’d hang my shadow on two rocks for the fugitive birds to build a nest on my shadow’s branch, As can be seen, the translator inadvertently rendered the plural ‘ ال ُّ روthe roads’ as the singular ‘ ال ّ رthe road’ and consequently caused serious damage to the poetic discourse. That is, by speaking about many roads, the poet intends to communicate the message that he has many options where some lead to Cordova and some do not. In contrast, by restricting this state of affairs to one road, the translator obliterates the proposition of choice and, as a consequence, distorts the conveyed meaning. Later on in the same poem, Al-Udhari had another translational miscue by rendering the singular ‘ وروتmy rose’ and ‘ خطوتmy step’ as the plural ‘my roses’ and ‘my steps’, respectively, as can be seen below: أعو ُو إذا كان ل أن أعوو إلى وردتي فساا وإلى خطوتي فساا .ولانن و أعو ُو إلى صُ ْ طُ ْه I will return if I have to return to my roses, to my steps, 93 But I will never go back to Cordova. It is unfortunate that the translator erased the symbolism embodied in the uniqueness of the signifieds which the poet deploys, viz. ‘ وروتmy rose’ and ‘خطوتmy step’ by translating them into ‘my roses’ and ‘my steps’, respectively. It should be clear that the poet is making reference to unique entities, that is, ‘my rose’ to symbolize ‘his occupied homeland (Palestine)’ and ‘my step’ to symbolize ‘his infancy’, but the translation, sadly, betrays the poetic discourse by dispensing with the encapsulated intimacy and uniqueness, thus reducing the symbolism to mere reference to ordinary belongings. Another area for morphological miscues is possessive markers in poetic discourse as can be illustrated by the following example from Darwish’s poem ٍ ف صعي ٍ ‘ دط اع ف خA Gentle Rain in a Distant Autumn): A gentle rain in a distant autumn ْ ف صعي ٍ دط اع ف خ And the birds are blue, are blue, ُ ررصا.. ُ والعيافي ُ ررصا And the earth is a feast. . ْ واألرضُ عي The birds have flown to a time which will not return. ْ طارت إلى رد ٍ و عو ْو ُ والعيافي You’d like to know my country? أن تع ف وطن ؟ وت واله صيننا؟ And what’s between us? وطن له ف القيو ْو- My country is the joy of being in chains, ْ A kiss sent in the post. ُ قُبلتي أْ رولت ف ال ْ وأ ا و أُر All I want ذصح ْن From the country which slaughtered me د صًو ال غي دن بل أد Is my mother’s handkerchief .. ْ ت ج ٍ وأو ا دو And reasons for a new death. The translational miscue in the above stanza impersonalizes a personal entity, viz. ‘ ص لmy kiss’ is rendered as ‘ ص لa kiss’ and consequently marginalizes the poet’s intimacy to his homeland, for the poet is identifying occupied Palestine with his own kiss rather than with a kiss understood generically. 94 Further, Arabic comparative and superlative forms may constitute a major area for translational miscues. The following poem ( ‘ العي الهمThe Golden Age’) by Adonis is an illustrative example: “Take him away, Officer ...” "... " ُج ّ ا ش ط- “Sir, I know the gallows ْأع ف أن ال با ْقيله Are waiting for me صا ظار But I’m only a poet worshipping my fire ار شاع أع "وي- غي أ And I love Golgotha.” ."وأاب الجُلجُله “Take him away, Officer! ّ اش ط Tell him: The Officer’s boot ّ صل له إن اها الش ط Is handsomer than your face.” " ُج- ."مو د وجا أجال آ ا عي الحها الهم Age of the golden boot You are the handsomest and most expensive. .أ ت أغلى أ ت أجمل The translator, being insensitive to the significance of comparative and superlative forms in poetic discourse, relayed the comparative ‘ أغلىmore expensive’ as the superlative ‘most expensive’ and the comparative ‘ أجالhandsomer’ as the superlative ‘handsomest’. Notably, the poet’s diction is based on a comparison of two items and should not be construed in an absolute sense or as relative to choice among a multiplicity of items. Therefore, the translator’s option for the superlative instead of the comparative forms amounts to a major distortion of the message, that is, the juxtaposition of power (i.e. the golden shoes representing the occupiers) with weakness (i.e. the poet’s folks under occupation) is completely erased in the translation. An equally interesting translational miscue relating to comparatives and superlatives comes from the translation of Al-Qasim’s poem ‘ خاتا النقاش د وجانEnd of a Discussion with a Jailer’ as can be shown below: الصغرى From the window of my small cell د ُكو ب ر زا ت ْ ُس ل،أصي ُ أشجارا I can see trees smiling at me, 95 Roofs filled with my people, األُما أمل، ووطواا Windows weeping and praying for me. ّو وافه ت ا وتيل From the window of my small cell د أجل الصغرى I can see your large cell. الكبرى د كو ر زا أصي ُ ر زا Most importantly, the poet’s use of the superlative forms اليغ ىand الا ىmust be construed in an absolute rather than relative sense if it were to communicate the schema the poet is working with. The translator, however, inadvertently relayed the superlative forms in question into gradable adjectives, viz. ‘small’ and ‘large’, respectively. As a result, the translator’s option distorts the poetic thought-world by interpreting these superlative forms, which are supposed to be taken in an absolute sense, i.e., ‘micro-cell’ and ‘macro-cell’ respectively, in a relative sense. The movement from ungradability to gradability, i.e., ‘micro-’ and ‘macro-’ vs. ‘small’ and ‘large’ cripples the poetic discourse encapsulated in the symbolism of the poem, that is, ‘the macro-cell’ represents ‘the entire occupied homeland (Palestine)’. In effect, the poetic schema takes a concrete entity (the macro-cell) as a point of departure for initiating a symbolic entity (occupied Palestine). Unfortunately, this area of cognitive correspondence between the two entities is lost altogether in the translation above. 5. Syntactic Miscues Translational miscues in the area of syntax manifest themselves at the phrase and clause levels. At the phrase level, miscues in poetic discourse relate mainly to the Arabic phrase called الاحاف ‘ والاحاف إليهConstruction of Addition’, which is sometimes interpreted inadvertently as adjectival rather than additive by translators. Following are illustrative examples from Darwish’s poem عن دا ‘ همب الشا ا إلى النومWhen the Martyrs Go to Sleep’. عن دا همب الشا ا إلى النوم أصحو وأا وُا ْ د هواة ال ّرثاء تُي حون على وط ٍ د وحا ٍ ود شج ٍ د و ا ٍ ودا: أصول لا أمنئا صالسًد د حادث المستحيل ود صيا الاهصح الفائح مل ُكلنا شا ا ْ؟.ت د الوص ب س صو ُ وأو لا، ق وص ا When the martyrs go to sleep I wake up to guard them against professional mourners. 96 I say to them: I hope you wake in a country with clouds and trees, mirage and water. I congratulate them on their safety from the incredible event, from the surplus-value of the slaughter. I steal time so they can snatch me from time. Are we all martyrs? The translator, who was miscued by the ‘Construction of Addition’, rendered ْ ‘ موا ال ثاlovers of mourning’ as ‘professional [amateur] mourners’. It should be noted that while both renditions in English may carry the intended pejorative nuances, only the rendition taking the Arabic phrase as a ‘Construction of Addition’ proper may refer to all categories of mourners (i.e. professional, amateur, or otherwise). As a matter of fact, the poet intends to harshly criticize those writers whose tacit profession is to compose insincere elegies upon the falling of Palestinian martyrs in the occupied Arab land. Similarly, the translator was miscued by the phrase ( ااوث الاس حيلa Construction of Addition) in the above extract by translating it as ‘the incredible event’ instead of ‘the event of the incredible’. This miscue twists the reference from ‘the incredible as an event (by way of metaphor)’ to ‘the event that took place though it was thought to be incredible’, thus mitigating the mood of melancholy and despair cast by the original phrase. At the clause level, translational miscues may involve tense and voice, among other grammatical categories. Following is an extract from Darwish’s poem “ أف دةل مها النشي ؟Is it in Such a Song?’, whose English translation features a syntactic miscue relating to tense: ق داتوا؟ ُ ستعش ع فت وكل اله.. عشقت.. وداذا تقو ُل الحيا ُ لاحاوو ورو ش؟ عشت و حا ُل شار ي ٍ ودف ا آخ صا، أف دةل مها النشي ُوو ُ الاا . ألن الحيا ايا.. لنغلق مها النشي علينا؟ ولاننا ووف حيا What does life say to Mahmud Darwish? You have lived, loved and those you loved are dead? Is it in such a song we cushion a dream, hold a victory sign and the key to the last door So as to shut this song from us? But we will live because life goes on. 97 It is clear that the replacement of the future verb form (‘ و عشقyou) will love’ with the past verb form ‘you loved’ in the translation has erased the paradox intended by the poet, that is, ‘all those you will love are dead’, which is meant to convey the totality of loss, inclusive of the future. Notably, the paradoxical inclusion of the future in the past is an integral part of the poetic discourse, but is, unfortunately, completely missing in the translation. The translator may sometimes fail to assign the correct voice, as can be illustrated by the translation of the following extract from Adonis’s poem اليح ا (دخ ارات د وديات ايار صي وت )3582 ‘The Desert (The Diary of Beirut under Siege, 1982)’: A star was drowned in blood, - ْ غرقت جا ف ال دا The blood a boy was talking about كان طفل ح ّ ث عناا And whispering to his friends: :و وشوش أصحاصه Only some holes known as stars ْ ل ع ف الساا Remain in the sky. ... ، ُو ّايت أ جاا ال دا ال الةقو ال غي ُ صع The syntactic miscue above involves the rendering of an active verb form ‘ غ صتdrowned [sank]’ as a passive verb form ‘was drowned [was sunk]’, thus initiating the counterintuitive poetic metaphor ‘A star was sunk in blood’ instead of the creative poetic metaphor ‘A star sank in blood’. As a matter of fact, the poet, on the tongue of a child, is establishing cognitive correspondence between the massive killing of civilians and the disappearance of stars, i.e. hope. It should be noted that the translator committed a translational mistake in addition to this translational miscue as he erroneously used the English verb ‘drown’ instead of the correct English verb ‘sink’ in his translation. 6. Semantic Miscues Semantic miscues occur both at the word and clause/sentence levels. In either case, a semantic miscue involves the translator’s assignment of a contextually incongruent but a lexically related sense or proposition. Contexts (both linguistic and physical, Yule, 1985) play a vital role in the assignment of meaning through an indispensable interrelation between linguistic meaning and world knowledge (Page, 1985). However, abstract paradigmatic lexical relations may sometimes intervene during the process of translation such that syntagmatic and/or semiotic lexical relations are disrupted - a situation which gives rise to a semantic miscue. Following is a semantic miscue at the word level in the translation of the first stanza of Darwish’s poem ‘ عارف الجي ار الا جولThe Wandering Guitar Player’: 98 He was a painter ، كان روادا But pictures ولا ّ اليور ، عاو Usually Don’t open doors و تف ُح األصوا Nor break them .. و تاس ما Nor turn the fish away from the face of the moon. . و ت و الحوت ع وج به القا The words ‘ اوتwhale’ and ‘ وااfish’ are hyponyms since both are sea creatures belonging to the class of fish, yet they stand for referents that carry different associations in Arabic, viz. the ‘whale’ is associated with monstrosity, while the ‘fish’ is invariably associated with positive connotations. The translator, being unaware of this important distinction, was miscued into rendering الحوتas ‘the fish’ instead of ‘the whale’, thus obscuring and distorting the poetic discourse, as pictures [paintings] cannot combat monstrosity (i.e. occupation). It should be noted that the ‘moon’ in this poem is a symbol of freedom for the Palestinians living under the yoke of occupation, while the ‘whale’ is a symbol standing for the occupiers. It seems the poet is communicating the message that it is not by painting and the like, but by getting up in arms that Palestinians can restore their occupied homeland. Sometimes, semantic miscues at the word level may relate to collocational meaning rather than lexical sense proper, as can be illustrated by the closing stanza of Adonis’s poem ‘ و الجThe New Noah’: We have an appointment with death, دوع ا دوت We have become familiar with our shores of despair, وشطآ نا أس ألبفنا رضينا صه We have grown to accept its frozen sea with iron water, حديد المياه، جلي ا،صح ا اح إلى ُدن اا And we sail through it to its end. ع We carry on moving and never listen to that God, اح وو يغ لهاك اإلله We long for a new god. . ووا تُقنا إلى ر ٍّ ج The translator, being insensitive to collocational meaning, was miscued into rendering ا الاياas ‘iron water’, which may not make sense in English at all, instead of the potential English collocation ‘numbing water’, which captures the intended meaning. As a result, this semantic 99 miscue disrupts the poetic coherence of the text, leaving the target reader confused as to what ‘iron water’ signifies. In some cases, semantic miscues may stem from dialectal differences in the lexicon, as can be demonstrated by the following stanza from Adonis’s poem اليح ا (دخ ارات د وديات ايار صي وت )3582 ‘The Desert (The Diary of Beirut under Siege, 1982)’: ووف ت ى You will see Say his name ُ صل اواه Say I painted his face ُ روات وجاه أو صل Stretch your hand to him حو Or walk like any man أو بو ْ كاا سي كل راجل Or smile ْ أو اص س Or say I was once sad .، أو صُلْ از ت د You will see ووف ت ى There is no homeland ... ... ليس مناك وط د As can be seen, the translator was miscued into relaying راج ٍلas ‘man’, which is a sense specific to Egyptian colloquial Arabic, instead of ‘pedestrian’, which is the sense in Standard Arabic. In fact, the poet is conveying the message that there are many ways to affiliate with a homeland, one of which is to ‘walk [to it] like any other pedestrian’. Consequently, the rendition of راج ٍلas ‘man’ distorts the poetic discourse as it is not only men who affiliate with the homeland, but also women and children. However, the search for a homeland, Adonis argues, is a futile enterprise because there is no homeland in the first place. Let us now turn to some examples featuring semantic miscues at clause/sentence level rather than at word level. Following is an example from Darwish’s poem ‘ دطار أثيناAthens’ Airport’, where the presence of an ambiguous clause miscued the translator: مالي ومالك؟:أوظف دال ؟ فقال الاةقف أ:صال الاوظف An employee said: “Where can I invest my money?” 100 An intellectual said: “Your money and mine?” Clearly, the translator misinterpreted the expression دال ودال ؟by translating it into “Your money and mine”, when it actually means “This is none of my business!” Not only was the translator insensitive to the poetic context where his generic referents expounded differing points of view toward Palestinian socio-cultural realities, but he was also unaware of the poetic utilization of two different varieties of Arabic: Standard Arabic vs. Colloquial Arabic. The translator, not being cognizant of this symbolic poetic shift in style, viz. the abrupt employment of colloquial instead of standard Arabic, committed a semantic miscue, and consequently missed the point altogether. It should be noted that the poet invests two discourses or thought-worlds, i.e., the one of the Employee, which is predominated by material interests, and the one of the Intellectual, which is supposed to be free of such interests. It is unfortunate that the translation fuses the two polar discourses into one by equating the Employee with the Intellectual in terms of interests. This being the case, the poet’s schema, which can be argued to be deeply rooted in the Palestinian socio-cultural reality in particular and the Arab one in general, is absolutely incongruent with that of the translator. However, it is needless to say that the poetic pun, which is doomed in translation, adds to the aesthetic value of the text in Arabic. ُّ ح Another semantic miscue at the sentence level occurs in the translation of Darwish’s poem ق لنا ‘ أن حب الخ فWe are Entitled to Love Autumn’, as can be illustrated below: ُّ ح :ق لنا أن حب اا ات مها الخ ف وأن سأله وح ونحن نمد ُد أجسادنا فيه فحما؟،أفي الحقل متسع لخريف جديد ا لي نا عش ُداال ُ لي نا ور، خ ف ُن ِّاسُ أوراصه ذم ا ق ال ي ُ لنشا دا الف جنو العيون لنسأل عاا،ْ و ا لي نا ل و ّو.ق صي الفيول . اانا الشع ُ وال سال.تسا ل آصاؤ ا اي طاروا على ص ّا ال دح We are entitled to love the end of this autumn and ask: Is there room for another autumn in the field to rest our bodies like coal? An autumn lowering its leaves like gold. I wish we were fig leaves, I wish we were an abandoned plant. To witness the change of the seasons. I wish we didn’t say goodbye to the south of the eye so as to ask what 101 Our fathers had asked when they flew on the tip of the spear. Poetry and God’s name will be merciful to us. By examining the translation closely, one can see how the translator was miscued into relaying the sentence و ح ا ُو أجساو ا فيه فحاا؟ فج ٍ أف الحقل د س لخas involving a purpose relationship rather than a circumstantial relationship. It should be noted that the autumn is employed as a symbol of death with leaves falling all over, saving no room even for a new death. Therefore, the sentence above may be relayed as ‘Is there any room for a new autumn with us resting our bodies in it like coal?’, where the question is rhetorical in nature. However, this rhetorical question is rendered erroneously in the translation above such that it is still begging an answer. Our last example of semantic miscues at the sentence level is drawn from the translation of Darwish’s poem ف صعي ٍ ‘ دط اع ف خA Gentle Rain in a Distant Autumn’, as can be shown below: A gentle rain in a distant autumn ْ ف صعي ٍ دط اع ف خ And the birds are blue, are blue, ُ ررصا.. ُ والعيافي ُ ررصا And the earth is a feast. . ْ واألرضُ عي Don’t say I wish I was a cloud over an airport. ال تقولي أنا غيمة في المطار All I want ْ فأ ا و أُر From my country which fell out of the window of a train ْ وقط ْرجاج القطار تد ب Is my mother’s handkerchief د صًو ال غي دن ل أد .ْ ت ج ٍ ْوأو ا دو And reasons for a new death. As can be noted, the rendition of the sentence ‘ و تقول أ ا غيا ف الاطارDon’t say I wish I was a cloud over an airport’ is semantically miscued, as there is no indication of wishing to be a ‘cloud’, which is a negative symbol in this poem, on the part of Darwish’s subject. This sentence, in fact, is merely a request by the poet of his subject not to say that ;she is only a cloud over the airport', thus unjustifiably belittling herself, for he will be asking for nothing but his mother’s handkerchief and reasons for a new death, which is not asking for too much. However, the translation does not communicate such a message; on the contrary, it causes a breakdown in the coherence of the poem by bringing in a wish that is completely incongruent with the subsequent modest demands. 102 7. Conclusion The present study constitutes a contribution to miscue analysis, which is originally a readingbased orientation, from a translational perspective. It has been shown that the discussion of translational miscues, which are argued to be the printed traces of oral miscues in reading, gives an insight into the nature of translational problems in general and miscued translational responses in particular. Thus, an elaborate examination of translational miscues in Modern Arabic Verse as exemplified by three celebrity Arab poets (M. Darwish, S. Al-Qasim, and A. Adonis) has been followed. Throughout, it has been pointed out that poetic discourse may be seriously twisted or even be crippled by the translator’s falling victim to phonological, morphological, syntactic, or semantic miscues. That is, if the translator is miscued during the process of translating, s/he may end up presenting a thought-world that markedly conflicts with the one communicated by the poet. To minimize miscues, therefore, the poetry translator needs to first establish a cognitive rapport with the text; as such a thing would be the only guarantee for capturing schemata as they manifest themselves in creative poetic discourse. 103 Audience Awareness and Role of Translator Mohammad Farghal & Abdullah Shakir Abstract The paper demonstrates that the translator needs to be aware of the culture-bound norms of expression in the TL if s/he is to offer a translation which is accepatable to target readers. By employing an Arabic political editorial translated into English by MA translation students, it shows that English native speaker informants are readily cognizant of deviations from TL textual norms. Such deviations appear to mar the translation product in terms of micro- and macro-features. Hence, student translators need to be alerted to features such as lexical appropriateness, tone of presentation, and content organization when engaging in translation activity. 1. Introduction Texts manifesting culture-bound features of expression tend to pose communicative problems to both the translator and the recipient of the target language (TL). The recipient approaches 104 the translated version of the source text (ST) with background knowledge and longestablished conventions of expression and processing most often dissimilar to those of the original ST recipient. When the ST is translated into another language, the TL recipient evaluates and sometimes modifies the in-coming message within a frame of reference in which his cognitive knowledge structures and convention-based nouns of address play a determining role. Thus, processing the content of the translated text tends to be contingent upon the recipient's ability to glean the message and appropriate it. Komissarov (1987) argues that this can best be achieved if the cognitive knowledge extracted from the TL text coincides with that of the SL text producer. This implies that processing the content of the TL text can be subjective in a number of ways, and that since the recipient is not a passive reader, the power of the text is not invariant and, therefore, the 'truth' of the content is not 'enshrined' (Rayor 1987). An SL text is normally addressed to an audience who, presumably, share with the text producer his/her cognitive knowledge structures, cultural values, and conventions of address. Contextual elements, which shape the message at both its micro - and macro -levels, are usually catered for in such conventions. The text, according to this view, complies with norms of rhetorical logic of address (Kaplan 1982) shaped by the culture in which it has developed. Among the contextual elements that determine the degree of processiblity, naturalness, and acceptability of address are the 'who' (recipient), the 'what' (the socio - linguistic occasion), and the 'how' (the style of address favored by the TL recipient). Consequently, it is not uncommon for a literal translation of an SL text to be rejected by the TL discourse community (Swales 1990) as being incongruent with their long established forms and strategies of address. One, in this case, can speak of communicative failure wherein the translator fails to realize translation equivalence (1E) in its multidimensional sense (cf. Saedi 1992). 2. The Present Study This study aims to investigate the role of the TL audience in determining the communicative acceptability of a translated version of an SL text. The investigation will be informed by the audience's interaction with and reaction to the predications manifested in the surface structure of the TL version of the text. Addressing the above notion will, inevitably lead us to look into other relevant notions, particularly the notion of the Equivalent Effect Principle, (Cauer 1896 and Koller 1978 cited in Newmark 1981) to see what or who determines the equivalence sought by the translator. The notion of audience and its role as a determinant of interlingual acceptability has been emphasized by translation theorists, e.g. Nida (1964), de Ward and Nida (1986), Jackobson (1957), Wilss (1982), Sa'deddin (1987). The essence of the notion is that translators (like successful writers) address, on behalf of the SL writer, TL recipients whose beliefs, traits, attitudes and modes of thinking need to be taken into account when translating. The 105 translator, according to this view, assumes the role of a mediator who first explores the SL text for elements which may handicap accessibility of the message and therefore may interfere with positive interaction on part of the TL audience. The consideration of the TL audience emanates from the translator’s awareness that the TL reader/recipient is not a passive target. Viewing the TL recipient as a passive target is, in fact, an oversimplification of both the translating and reading processes. Both the translator and the TL audience are engaged in a text -negotiating process (Hatim 1985), especially when the SL text incorporates culture - specific features. The TL audience processes the verbal input and transforms it into conceptual codes that can be integrated, modified, or totally rejected as being deviant from acceptable norms of address. Taking the TL text as the medium of interlingual communication, the translator gets engaged in a 'decentering' process ( K r o l l 1 9 8 4 ) w h e r e i n a n e f f o r t i s ma d e t o e l i mi n a t e t h e 'egocentricisms' of the SL text that may impede communication. In so doing, the translator attempts to make the text 'efficient' (effort -saving for the TL audience) and 'effective' (achieving sought results) (de Beaugrande and Dressler 1981). However, such a role is often constrained by considerations related to the degree of freedom that the translator can enjoy, and the bi- polarity of semantic as opposed to communicative or functional translation (Newmark 1981). Egocentricisms of SL texts crop up at two levels: the micro level (mainly lexis) and the macro level (content organization). At the micro level, egocentricism may unfold in lexical items blending into their sense components experiences unexchangeable in a given sociolinguistic event. To minimize such egocentricisms and divergent experiences, translators need to establish "... equivalence in the experiences of the participants (de Beaugrande and Dressler 1981: 216). To realize this, translators may expand, reduce, or modify textual components. In a similar argument, Geertz (1973) describes the translator as an 'ethnographer': he is the locus of contact between two cultures where each is equipped with its own linguistic, literary, and rhetorical techniques of articulating a message. Experiences are encoded in the memory of the SL writer and are, revealed in the surface structure of the text. Their decoding by the TL recipient hinges upon their availability to him. When the translator (informed by his analysis of both the SL text and the TL audience) perceives such experiences as opaque to TL audience, the translation process takes on the role of an approximation process whereby gaps are bridged. Rendering a replica of the SL text may, then, lead to pragmatic failure and communication breakdown. 2.1 Material This paper is based on translated versions of an Arabic recommendation letter (See Appendix 2). Thirteen MA translation students translated the letter into English. Their 106 translations were then given to eight native speakers of English doing advanced Arabic courses at Yarmouk University, Jordan. These courses include three translation courses from and into Arabic. The eight native speakers were asked to read the translations and record their comments, notes, and suggestions as regards the style in which the letter in its TL version was written. Their comments were collected, analyzed, and categorized under the following headings: 1. semantic load of lexical expressions 2. tone of presentation 3. content organization Under the 'semantic load of lexical expressions' and 'tone of presentation', the following comments were noted: a. unduly intimate expressions b. awkward in this context c. hyperbolous d. exaggerative e. inappropriate in this kind of text f. more of a personal letter than a recommendation letter g. more persuasive than descriptive or reporting h. looks like a recording of an argument between writer and reader Under content organization, the following general comments were received: 1. text lacks focus 2. confusing ordering of sentences 3. awkward structure 4. sentences too long 5. difficult to follow the thread of thoughts 2.2 Analysis Before dealing with the analysis, we wish to make it clear that the present text (the recommendation letter) was not selected as representative of Arabic style of writing recommendation letters. Rather, it is introduced here as a text incorporating stylistic features likely to pose translating problems at both the micro- and macrolevels. The argument and discussion, therefore, are generalizable only to texts manifesting comparable features. 2 .2.1 Lexical Expressions Lexical items play a key role in conveying attitudes, dispositions, emotions, and beliefs. As content bearers, they play a significant role in the present study. They project at the forefront of the discourse as spots of attraction and audience-enchanting elements. However, the force inherent in them in their SL version seems to have failed to elicit the intended response (i.e. appreciation) when transferred into TL equivalents. Some key lexical items were unanimously rejected as being awkward, contextually dissonant, and unduly hyperbolous. To illustrate, the following lexical expressions were literally translated in the thirteen TL texts and were marked as 107 awkward by the TL group: - tilmiiðii wa-bnii [my pupil and my son] - aṭ-ṭaalib-u-l-c aziiz [the dear student] - risaalah šaaqqah [arduous mission] - yanhaḍu bi-haa [undertake it (with utmost courage and forebearance)] - siwa-r-rijaal [except those men] - laa yušaqqu la-hu ɤubaar [his dust cannot be penetrated] The above lexical expressions lend themselves to two semantic categories: a. unduly intimate b. hyperbolous 'ibnii (my son) and al- c aziiz (the dear) fall into category (a), i.e. ‘unduly intimate expressions’. The two items are not, however, uncommon in teacher -student discourse. Teachers in Arab culture tend to look on their students as 'sons'. But in recommendation letters, such an intimate and personal tone tends to give way to a more distanced and neutralized descriptive tone wherein the recommended academic or professional achievements, traits, and potentials for future research are described. The 'I-You' or 'I-He' relationship (Moffette 1968) is not unusual in recommendation letters, but its occurrence is entailed by discoursal necessities, viz. mention of the 'participants' (Grimes 1976) in the text. Matters relating to the personal relationship between the writer of the letter and his/her student (e.g. he was/ is as dear as my son) are usually of no significance to the party to whom the letter is addressed. Category (b) of the lexical items in this text incorporates into their semantic make-up elements of exaggeration (when viewed within the contextual frame of the text and its text type) and a flavor of Classical Arabic where chivalric decorative style was a characterizing feature. Such a style was marked by proclivity for exaggeration, metaphoric expression, fondness of sonorous and penetrating lexical items, and a clear inclination for grandiloquence. Such stylistic features were used (and to some extent are still in use) to hold sway over the audience. Though pleasing in chivalric prose and poetry, such features would sound rather dissonant to most educated Arab readers when they occur in a text of a function similar to the one we are dealing with. They would sound as empty expressions void of convincing power. In order to indentify the senses each of the target lexical items invokes in the SL (Arabic in this case), we will provide below componential analysis of each. risaalah [mission/message] + religious association + arduous effort + endurance 108 + devotion to a cause + counter resistance yanhaḍu bi-haa [undertake it (with utmost courage and forbearance)] + acceptance of challenge + acceptance of responsibility + serious enterprise + voluntary undertaking ar-rijaal [the men] + strong masculinity + courage + sturdiness + allegiance to a cause + chivalric traits laa yušaqqu la-hu ɤubaar [his dust cannot be penetrated] + extreme (physical) effort + excellence + competitiveness + championship + extreme speed + chivalric traits A look at the analysis shows how the above lexical items are overloaded with senses usually uninvokable in this verbal event. A recommendation letter does not absorb all those exalting, luring, and hyperbolous senses. When these SL items were rendered literally in the TL, they were rejected as being dissonant, resonant and transgressive. Rejection of the literal renderings of the above items indicates that the translating process in the case of this text (and perhaps other comparable ones) is regarded as a process of mediation and subsequently communication rather than as a process of 'sensu stricto' and subsequently cultural accommodation. The translator needs to take into account that realizing the Equivalent Effect Principle (Nida 1964) in the TL can be unachievable, and that a process of approximation (Holes 1984) ought to be sought in this case. In such a process, two contextual elements play an informative role; these are: a. anticipating the TL audience's response to the senses conveyed b. the text type of the translated material. 109 Attention in such a process should be given not only to problems of language codification (e.g. lexicalizing of the SL vocabulary items), but also to problems emanating from cultural divergences and modes of thinking and address relevant to the SL and TL texts. Further, the translator should be aware of the fact that text type lies at the core of the translating process. Consequently, the translator's option for one type of equivalence rather than another is constrained by the type of the text, among other things. The text we have in this study is in principle meant to pass on information about a student by his professor in the form of a recommendation letter, hence its being a blend of exposition and instruction. Armed with this awareness, the translator needs to relay the message independently of form in order to approximate the effect, thus adopting Functional Equivalence (Kachru 1984 and De Waard & Nida 1986) or Textual Equivalence (Catford 1965), or at least Ideational Equivalence (Farghal 1994) rather than opting for Literal/Formal equivalence blindly. The inevitable tug of war between Form and Function in the translating process should always be informed by text type, that is, in a reader-centered text, priority should be given to function, whereas in an author-or text-centered text, priority ought to be given to form at the risk of sounding awkward and/or unnatural. 2.2.2 The macro-structure of the text Our central aim here is to identify those textual features responsible for promoting or impeding sense-continuity in the text as viewed by the TL audience. In other words, we are concerned with how content is mapped onto the text. The way it is mapped activates in the TL audience (and the SL audience) text-experiential strategies which function as a frame of reference for judging the communicative acceptability of the SL text (and the TL version). An audience processing strategies of a given text type are governed largely by those previously activated in processing texts of comparable function. Part of the experience of the TL audience is their experience of strategies (linguistic and non-linguistic) employed to produce such a text (see Kress 1982 for more details). Thus, one could argue, in the case of conflicting strategies of text production and text processing - both being nurtured over centuries of practice - interaction is expected to be crippled, or at best, impeded. A look at the comments provided by the group of native speakers of English (see section 2.1.) shows that 'lack of focus' in the text can be attributed to the 'confusing organization of content' and 'the awkward structure of the text'. In order to see how the SL text is organized and how that reflects upon the TL versions advanced by the student translators, we will have a quick survey of its macro-components (see Appendix 1). The text begins with an evaluative and affective sentence which expands to incorporate three clauses with referential items (the recurrent definite article 'al' and the pronoun ‘allaðii (which) which unduly assume shared knowledge with the reader who still doesn't know what almasraḥ (the theater) refers to, nor is he informed of how this masraḥ is 'extremely 110 lacking in theoretical studies based on practical experience'. The second sentence is obtrusive in its tone as indicated by the sentence initiator wa-ðaalika ma … (and that was what ...), which is an emphatic and concluding Arabic instrument. Its obtrusiveness derives from the function it performs in the discoursal environment it normally exists in. Often, it comes as a concluding and assertive statement in a tension-laden argument. The third sentence is also an evaluative and obtrusive statement, passing judgmental views on behalf of the reader, and leaving him with no chance to evaluate or appreciate the 'student's effort'. The fourth sentence reveals a tone similar to that in the previous sentences. In addition to its obtrusive and assertive tone, the sentence suffers two referential ambiguities signaled by the reference markers wahwa (and it is) in clause (a), and allatii (which) in clause (b). The reader is provided with little clue as to what these two elements refer to in the text. In response to the comments we received in section 2.1., the content of the SL text was reorganized and sentences were broken down into smaller and manageable units. (see Appendix 3). The new SL version, together with the old one, was shown to three professors of Arabic. The new version was judged as 'more compact', 'more acceptable', 'more coherent', and 'more representative' of an Arabic recommendation letter than the old version. The new version was then translated (see Appendix 4) by one of the present researchers and shown to the same group of native speakers of English. Their comments stressed that the TL text now 'sounds much more English ' than the versions they read previously. The main macro features of the new SL version relate to the reorganization of the content. The first sentence sets the scene, providing background information that identifies the relationship between the participants (teacher and student), together with the occasion that prompted the verbal event (a recommendation letter by a supervisor of an MA thesis). The second sentence, being contextualized by the first one, offers an assessment of the student's work and provides more context to the information introduced in sentence 1 as regards the student's subject of study. The third and fourth sentences substantiate the assessment set forth in sentence 2. The fifth sentence plays down the assertive and obtrusive tone of the one in the old version (sentence 3). The concluding and assertive tone has been mitigated by subordinating the predication of the statement through introducing it as shared and given information marked by the subordinator ca1a-r-raɤmi min (although), thus giving prominence to the student's 'academic achievement' and his research abilities. Sentences 6 and 7, being warranted by the preceding ones, fit very well into the hierarchic structure of the text by introducing the writer's conclusion and recommendations (for the modifications at the micro level, see the section below). 3. Conclusions and Implications 111 A close look at the modified version of the SL text shows that drastic changes have been made to its topographical structure and mode. The text was negotiated, but with an obvious 'bias' toward the TL audience. Two terms of reference informed the modification: the first was the views and comments of the group of native speakers of English, and the views and comments of the three professors of Arabic who read the original SL version. The second was the assumption that the translator is not a machine translator; rather, he/she assumes the role of reader and writer who is well aware of the TL audience culture, modes of thinking and conventions of expression. The translator, in light of the role he/she is entrusted with, can remap the SL text to fit it into the TL audience formal and content schemata (for more details, see Carrell 1982 and Farghal & Shakir 1992). Remapping the content of the SL text entailed changes at both the micro- and macrolevels. At the micro level, metaphoric, hyperbolous and grandiloquent expressions were played down, and the emotive charge was dwindled. Some constituents were even deleted e.g. 'ibinii [my son], al- c aziiz [the dear], and šurriftu [I was honored, as being awkward in this context of situation. At the macro-level, the text underwent a process of constituent transposing wherein sentences shifted positions and made up new paragraphs which emerged as a result of restructuring the content. A couple of points emerge from the discussion: 1. It is important for translator trainers to import to their students that translating is a process that is promoted not only by the translator's linguistic knowledge, but also by his knowledge of the TL culture, the audience modes of thinking, and the rhetorical and social function of the text. Experiential matching should be considered, and factors leading to acceptability or rejection of the translated version need to be monitored and taken into account 2. A pre-translating phase should be stressed, and student translators need to be engaged in an exploratory process wherein matters related to the function of the text (i.e. its text type) and the relationship between the writer's choice of lexical items in the SL text and their convincing power when rendered in the TL need to be carefully studied, making decisions informed by the translator's knowledge of the TL cultural conventions and modes of expression. In this spirit, practical translation courses need to be founded on a solid base of socio-linguistics, discourse analysis, and text linguistics. 112 Appendix 1 )(The SL Text شها ة وتزكية أود أن أذكر بكل الخير جهود تلميذي وابني -----------في مجال المسر وربطه بالعلم ،خدمة لفن المسر وخاصة في فرع التمثيل الذي ينقصه الكثير من التنظير المرتبط بالتجربة العملية في مجال اإلبداع .وذلك ما حاوله الطالب العزبز ،----------- إذ أحاط بحركة التمثيل في األردن وكذلك حركة اخراج وحركة النص (تاريخا ً وحرفة ونظرية ثم فنا ً) في محاولة منهجية لتوسي دائرة الفنون المسرحية عن طريق العلم .وهي رسالة شاقة ال ينهض بها سوى الرجال الذين تأسسوا تأسيسا ً ال يشق له غبار وهو أمر حققه -----------من خالل مثابرته وعقليته التحليليحة ودأبه وتحصيله في قسمي المسر في جامعة --------- -والمسر بآداب ------------في مرحلة التمهيدي ومرحلة إعداده للماجستير الذي شرفت حقيقة باإلشراف عليه واالستمتاعبإنجازه المنهجي التحليلي غالبا ً والوصفي أحيانا ً .وهو أمر ارى أنه خليق بعناية المسئولين في كلية التربية والفنون الجميلة بجامعة ،----------حثا ً وحضا ً على المواصلة األكاديمية التي ستجني ثمارها على يديه ويدي زمالئه في الحركة المسرحية األردنية خالل خمس سنوات. Appendix 2 )(A sample of the student translations of the SL text A Certificate of Merit With all the best, I would like to mention the efforts of my student and son in t h e f i e l d o f t h e a t e r a n d it s co n n ec t i o n w i t h s c i e nc e i n t h e interest of drama especially acting which lacks a great deal of theorization connected with practical expression in the area of creativity. That was what the dear student achieved as he mastered the movements of acting and scripting in Jordan (historically, professionally, artfully, and theoretically) in a methodological attempt to enlarge the circle of drama by means of science. This is in fact an arduous mission which only matchless and unparalleled men can a c co mp l is h . This in reality has been performed b y t hro u gh perseverance and analytical mentality, persistence and 113 great achievement in the two branches of theater in University -------------- and University -------------- b o t h i n his undergraduate studies and in his preparation of his M.A. degree. In fact it was a real honor for me to supervise him and enjoy his performance which was mostly analytical and sometimes descriptive. This is a matter which is worth the concern of the officials of the college of Education and Fine Arts in University by motivating and urging him to complete his studies whose benefits will be achieved in five years. Appendix 3 (The modified SL version) واطلعت على دناجه ف الاس، ال أع ما لنيل ورج الااجس ي ف دوضو----------- لق أش فت على روال الطالب وف مها السياق أوو أن أثن على جاوو ف.، ال حث اله س ن ف دعظاه إلى األولو ال حليل د ش د الوصف أايا ا . اقل الاس ف األرون ودحاووته ل وظيف ال قنيات العلاي خ د لاها الف لقد أحاط بحركة التمثيل في األردن كما أحاط بحركتي اإلخراج وكتابة النص المسرحي بأبعادهما التاريخية واالحترافية . ومن هنا حاول وبشكل منهجي توسي دائرة الفن المسرحي في األردن باإلفادة من التقنيات العلمية المتعلقة بهذا الفن.والنظرية إال أن هذا الطالب قد نجح في هذه المحاوالت بفضل مثابرته ومنهجه التحليلي،ورغم أن مثل هذه المحاولة تعد مهمة صعبة . ---------- و--------- اثناء دراسته في قسم المسر في جامعتي وخاصة في فرع التمثيل الذي يفتقر إلى الدراسات النظرية المبنية، وبناء على اطالعي على جهده في مجال المسر األردني جدير برعاية المسؤولين في كلية التربية والفنون الجميلة في---------- فإني أرى أن السيد،على التجارب العلمية واإلبداعية . كما أنه جدير بتشجيعهم له على مواصلة دراسته التي أتوق لها نتائج مثمرة في المستقبل القريب. ----------- جامعة Appendix 4 (A translation of the modified SL version) I have supervised Mr. X's thesis which he submitted for the M.A. degree in Theater and I have been aware of his approach which is basically analytical, though descriptive at times. In this context, I would like to highly commend his endeavors in the field of Theater to put theatrical technology to the service of theatrical arts in Jordan. Mr. X has demonstrated in-depth knowledge of the historical, professional, and theoretical aspects of the movement of acting, as well as the movements of directing and scripting in Jordan. Equipped with this knowledge, he attempted in a systematic approach to expand the sphere of theatrical art by making use of relevant technologies. Though a hard enterprise undertaken only by those who are well-qualified, his efforts proved to be a real success - thanks to his perseverance a n d analytical approach during his study at University -------------- and -----------University. On the basis of my knowledge of his efforts in the field of Theater, especially in Acting which is lacking in theoretical studies based on practical and creative experience, I believe that 114 Mr. X should receive due attention from the academics in the Faculty of Education and Fine Arts at University -------------. I also believe that he should be encouraged to pursue his studies which, I am sure, will yield fruitful results in the near future. Reader Responses in Quran Translation: The Case of Referential Gaps Mohammed Farghal & Mohammed Al-Masri Abstract This article deals with reader responses to select translations of Quranic verses that involve referential gaps. Based on the results obtained from two types of questionnaire (an open form and a closed form), the study shows that referential gaps may cause serious problems to native speakers of English when they interpret Quranic messages in translation. Further, open form questionnaires prove to be more indicative of readers' comprehension of Quranic translation than their closed counterparts because participants, as this study substantiates, may overestimate their understanding of Quranic messages when responding to closed form questionnaires. 1. Background There seems to be a consensus that translation leaves the door wide open to gain and loss in communication. The fact that we need to translate is in itself the proof that different language communities possess different cultural repertoires and, consequently, there is an inevitable possibility that cultural elements are changed in translation. However, the potential for a communicative gain or loss is not only present interlingually but also intralingually. We are unlikely ever to find two individuals having absolutely the same experience of any event or action. Thus, absolute communication remains a desideratum (cf. O’Shea, 1996:238). De Waard & Nida (1986:42) maintain that translation loss is due to the fact that “sources and receptors never have identical linguistic and cultural backgrounds.” Catford (1965:94) links the possibility 115 of untranslatability to the circumstance when “it is impossible to build functionally relevant features of the situation into the contextual meaning of the TL text”. He thus places paramount importance on the pursuit of a functional equivalent for a suitable rendition of the SL message. However, one should not overlook the possibility that the pursuit of a functional equivalent, independent of other factors, the most important of which is the faithfulness to the SL message, does not significantly add to the SL message but it only employs a target language equivalent of something which is found in the source language context. Nida (1994:148) points out that although it is unlikely that there may be “untranslatable languages”, it is not unlikely that in certain languages “there are idols of words as well as idols of wood”, and he adds that “the translating of religious texts can be a good testing ground for the limits of translatability”. We have to tolerate some inevitable loss, especially in religious texts, but this should not discourage translators from searching for strategies that make translation possible; at worst, they may resort to footnoting. This article discusses the degree of similarity in the responses of source language and target text readers. Since the former is taken for granted, the latter assumes maximal importance, as target language readers’ response will be important for deciding whether a translation is successful or not. That is to say, by studying the target language recipients’ responses, a given translation can be validated. Theoretically, a translation can be successful or valid when it is likely to bring about on the TL receiver effects similar, to a reasonable degree, if it has an effect on the recipient similar to that the source language text had on the source language recipients. Nida and Taber (1982:1) believe that “correctness must be determined by the extent to which the average reader for which a translation is intended will be likely to understand it correctly.” Most translations should target the average reader and the understanding of the texts should be basically tested by means of responses from average readers. This orientation implies that recipients are not ‘passive targets’, as some terminology would suggest; their role is integral to the whole process (cf. Waard de & Nida, 1986:33). It also argues that translators have to be sensitive to readers’ response, which, in turn, concerns their ability or readiness to translate. Translators, especially those who deal with culture-specific texts, must have enough intellectual capacity to overcome the chasms of languages and cultures. In order to enable people to comprehend the message, translators should be able to “draw aside the curtains of linguistic and cultural differences” (de Waard de & Nida (1986:14). Another important factor is the translator’s attitudes towards the message communicated. All translators would agree on the importance of rendering the SL message faithfully, but most may not be able to do this successfully. The translator’s own interpretation, for example, should not affect his rendition adversely. The fact that the original text may lead to different interpretations allows for what Farghal (1993:262) calls ‘extrinsic managing’, i.e. managing which serves the intentions of the translator rather than those of the original author (For more on this, see Shunnaq 1994; Farghal 2010, 2012). Admitting that “Difficulties arising out of differences of culture constitute the most serious problems for translators and have produced the most far-reaching misunderstandings among readers” (cf. Farghal and Borini, this volume), Nida and Reyburn 116 (1981:2) concede that the translator’s attitude towards the languages involved can, in fact, dominate over his knowledge of languages and their cultural interpretations because of his emotional identification with these languages. So, translators need to be sensitized to the crucial strategies by which the SL texts communicate. In the case of the Holy Quran, there is a principal trichotomy of loads – namely, (a) the didactic load, (b) the informative/expressive load, and (c) the evaluative load. Thus, a TL rendition should not inform, for example, when the SL text intends to preach or dictate. This feature imposes a choice between furnishing translations proper or interpretations, where needed, to provide the right information to the TL audience. In the Holy Quran, translators will frequently be obliged to resort to interpretations because of the nature of the text. As far as referential gaps are concerned, translations are vulnerable and they may jeopardize reasonable comprehension. The verses of the Holy Quran are woven in intricate and sharp ways. In addition, the presence of culture-specific terms and images augments the problems. Translation of religious discourse poses a large number of problems. In particular, religious discourse often has cultural and linguistic gaps that defy bridging. Religious language deals with supernatural events that lack finite or solid bases, i.e. they do not refer to any ‘extralinguistic reality’, which is a prerequisite for solid comprehension (Ivir, 1991:53). Consequently, religious language is subject to different interpretations; this implies that utterances translated today, for example, may develop new connotations and associations in the SL without influencing TL renditions. Religious language also reflects “transcendental experiences for which ordinary language seems to be so inadequate” (Waard de & Nida, 1986:21). In addition, religious discourse is timeless, which means that it may gradually come to have new connotations – a situation that gives rise to possible future paradoxes and inconsistencies. De Ward and Nida argue that religious language “generally recognizes the failure of words to communicate truth,” since there is always something “unutterable” in the words that can be “uttered” repeatedly. In other words, religious language often has connotations that can be rendered by specific equivalents at one time but, as time goes by, these equivalents may fail to render new connotations of the SL term brought about by, for example, a new exegesis. 2. Referential gaps Gaps in languages range from discoursal gaps to phonetic gaps. Since languages are different, there is not necessarily a one-to-one or even a one-to-many correspondence in semantic or syntactic relations between languages. The relations, as Nida (1994:147) puts it, “are always many-to-many,” a fact that opens up for ambiguities, obscurities, and fuzzy boundaries. Referential gaps are at the heart of such ambiguities. Rabin (1958: 127) defines them as “blank spaces in the field of reference, corresponding to referents outside the ken of the language community”. Dagut (1981: 63) asserts that “... we shall mean [by voids] that there is no single English [TL] designator which provides the required equivalent encapsulation of the situational features”. In this respect, Ivir (1991:50) explains that these gaps are created by different ‘extralinguistic realities’ in different cultures, for instance, when one culture lacks an element 117 which the other culture has. This leads to further gaps in the linguistic mapping of different as well as similar extralinguistic realities, a fact which makes for conceptual incompatibility between members of different linguistic communities. Long before Ivir, Catford (1965:94) distinguished between two types of “gaps” in respectively linguistic untranslatability and cultural untranslatability. Dagut (1981:64) regards the former as “voids resulting from intralinguistic factors” and the latter as “voids resulting from extralinguistic factors”. Along the same lines, Farghal (1995:198) explains that referential gaps are ‘experiential’, that is, they are “missing entities in a certain culture, as they enjoy no existence in the language community in question”. Dagut (1981 64) argues convincingly that certain languages are better equipped with the compressed power of designator than other languages, implying that some languages have well-defined terms for designating referents that correspond to more than one term in the TL, that is, a relation of one-to-many. In such instances, translators often resort to transliteration to overcome the gap, a strategy which, in effect, amounts to a confession that the gap is untranslatable. In addition, being completely alien to the TL recipients, transliteration offers no information since the transliterated term is not comprehensible by itself but only by means of the linguistic and the ideational context. Thus, the communicative value (Saedi, 1990:390) conveyed to the TL recipients is greatly impoverished. To sum up, competent translators need to realize these dimensions when handling texts with cultural or linguistic peculiarities. This endeavor is here examined by having TL recipients judge to what extent the TL text is comprehended. And, consequently, we hope that this article will shed light on some dimensions in translation in general and Quran translation in particular. 3. Methodology 3.1 Subjects and Corpus The participants in this study were 43 native speakers of English, 20 of whom responded to questionnaires in Jordan while on holiday and the rest responded to the questionnaires in the United States. They represented different age-groups, backgrounds, nationalities and professions and comprised both men and women. The translation corpus consisted of a total of 16 carefully selected translations of Quranic verses that were thought to cause some comprehension problems and were taken from the translations of Ali (1934), Arberry (1980), and Pickthall (1980). All these verses involve referential gaps (namely 22, partial and complete gaps). 3.2 Procedure The translations of 16 Quranic verses were set up in two questionnaires. The first questionnaire (Appendix I) checked the degree of comprehensibility on a scale of 5 levels, ranging from “straightforwardly understood” to “makes no sense at all.” The second questionnaire (Appendix II) arranged the 16 Quranic verses so that each verse required a different response, which allowed the participants to use their own words and to express their own attitudes and motives. The study thus has two dimensions: a receptive tool (the closed form) and a productive one (the open form). Twenty participants responded to the open form and the twenty others to the closed form. In order to check the validity of the answers to the closed form, a third group of 3 readers were asked to 118 respond to both questionnaires so that we could compare the same readers’ response to the closed form with that to the open form. 3.3 Data Analysis First, the corpus will be examined in light of the translation strategies adopted in the rendition of Quranic referential gaps into English (Ivir 1991). Then, the findings of the open form and the closed form will be presented and compared. Next, the match and mismatch results in the third group (3 subjects) will be discussed. Finally, the education variable will be examined to see whether it had any bearing on the results attained by means of the two different types of questionnaire. 4. Results and Discussion Table 1 illustrates that the linguistic corpus in the present study bears witness to many different strategies in the translation of Quranic referential gaps. Table 1. Distribution of translation strategies in the corpus Strategy Frequency Found in examples Literal translation 11 2, 3, 4, 11,15, 16, 21, 22, 26, 27, 28 Definition 6 5, 14, 17, 18, 19, 25 Substitution 5 1, 6, 9, 13, 23 Addition + Literal translation 2 8, 20 Substitution + Addition 1 12 Definition + Literal translation 1 29 When we turn to Table 2, which is an overview of comprehension, it becomes obvious that native speakers of English have problems in comprehending Quranic referential gaps in translation, at least according to this study. When we exclude the column "satisfactory sense" as not indicative, we find that 48.13% of the participants thought that they comprehended the verses in general, whereas about only 28.13% found the verses difficult to comprehend. Table 3, however, reveals that in the questionnaire with the open form, only 19.83% of the participants demonstrated a good understanding of the translations vs. 70.17% who were either unable to make sense of the translations in question or did not respond. This suggests that native speakers of English may overestimate their comprehension of Quranic translation. 119 The third group of readers furnishes evidence that the responses to the closed form presented in Table 4 are unreliable as a measure of genuine comprehension, since only 29.17% of their answers to the two forms match as opposed to a 70.83% mismatch. It therefore seems as if participants in studies like the present one tend to believe that they fully comprehend passages which are, in reality, only partially understood. Further, closed questionnaires always leave room for erroneous correspondences between source language and target language concepts. For example, many participants thought that the Islamic term zakat corresponded exactly to the Christian term alms, and they were not aware that in Islam, there are two terms for 'giving to the poor' namely: zakat and sadaqah, the former being compulsory and the latter optional. Such misconceptions can only be detected in an open questionnaire Table 2 . Results of the closed form questionnaire A Verse B No sense at all C D Very little Satisfactory sense sense No. E A lot of sense Straight- forwardly understood No. % No. % No. % No. % No. % 1. 0 0 1 5 6 30 1 5 12 60 2. 0 0 6 30 9 45 3 15 2 10 3. 10 50 6 30 2 10 1 5 1 5 4. 0 0 0 0 3 15 5 25 12 60 5. 5 25 5 25 7 35 0 0 3 15 6. 2 10 0 0 4 20 5 25 9 45 7. I 5 0 0 2 10 6 30 11 55 8. 3 15 4 20 4 20 3 15 6 30 9. 2 10 5 25 4 20 3 15 6 30 10. 0 0 7 35 5 25 4 20 4 20 11. 0 0 4 20 2 10 2 10 12 60 12. 1 5 5 25 4 20 4 20 6 30 13. 0 0 6 30 5 25 4 20 5 25 14. 0 0 1 5 8 40 I 5 10 50 15. 3 15 1 5 7 35 3 15 6 30 16. 6 30 6 30 4 20 1 5 3 15 120 Total 10.11 32 57 18.02 76 24.02 44 108 13.93 34.20 Table 3. Results of the Open Form (The Solid line indicates verse boundary). A B C D No response No sense at all Very little sense E Satisfactory sense No. A lot of sense F Straightforwardly understoo d The SL Term in Translation No. % No. % No. % No. % No. % No. % 1. 0 0 10 50 0 0 0 0 0 0 10 50 And He has enjoined... 2. 1 5 14 70 4 20 0 0 0 0 1 5 Purify ... Pickthall 3. 0 0 0 0 1 5 3 15 4 20 12 60 bow down... Pickthall 4. 0 0 1 5 0 0 3 15 4 20 12 60 prostrate... Pickthall 5. 2 10 12 60 2 10 2 10 1 5 1 5 Be as my mothers back 6. 1 5 12 60 4 20 1 5 1 5 1 5 unclean... Pickthall 7. 13 65 5 25 2 10 0 0 0 0 0 0 purify Pickthall 8. 3 15 9 45 3 15 0 0 2 10 3 15 reached their period Arberry 9. 5 25 11 55 2 10 0 0 0 0 2 10 defiled Arberry 10. 5 25 11 55 2 10 1 5 1 5 0 0 purify Arberry 11. 1 5 7 35 1 5 1 5 8 40 2 10 pilgrimage Pickthall 12. 2 10 12 60 3 15 2 10 1 5 0 0 visit to Mecca Pickthall 13. 5 25 9 45 3 15 2 10 1 5 0 0 gifts Pickthall 14. 6 30 3 15 3 15 1 5 3 15 4 20 private parts Arberry 15. 10 50 5 25 3 15 1 5 1 5 0 0 what their right hands own Arberry 16. 1 5 4 20 2 10 2 10 2 10 9 45 contact with women Ali 17. 2 10 7 35 3 15 6 30 2 10 0 0 take for yourselves clean sand 18. 1 5 8 40 11 55 0 0 0 0 0 0 vow ye vow 19. 4 20 5 25 3 15 3 15 3 15 2 10 no sense of the shame of sex 20. 2 10 7 35 6 30 1 5 1 5 3 15 ceremonial impurity 21. 5 25 4 20 3 15 0 0 1 5 7 35 have had contact.... Pickthall 22. 5 25 1 5 5 25 5 25 4 20 0 0 to clean, high ground and rub Pickthall 121 Arberry Arberry Ali Pickthall Ali Ali 23. 3 15 7 35 6 30 2 10 2 10 0 0 expenditure Arberry 24. 13 65 4 20 2 10 1 5 0 0 0 0 vow you vow Arberry 25. 3 15 0 0 3 15 0 0 1 5 13 65 monthly course Arberry 26. 3 15 1 5 6 30 3 5 4 20 3 15 have cleansed themselves Arberry 27. 17 85 2 10 1 5 0 0 0 0 0 0 God has commanded you... Arberry 28. 7 35 9 45 3 15 0 0 0 0 1 5 put away women.. Pickthall 29. 8 40 8 40 4 20 0 0 0 0 0 0 legal period... Pickthall Table 4. Results of the Third Group (Closed and Open Forms) Verse No. Match Mismatch Verse No. Match Mismatch 1. 0 3 9. 0 3 2. 1 2 10. 2 1 3. 2 1 11. 1 2 4. 1 2 12. 0 3 5. 1 2 13. 1 2 6. 0 3 14. 2 1 7. 1 2 15. 0 3 8. 2 1 16. 0 3 Total 8 16 Total 6 18 Match = 14 = 29.17% Mismatch = 34 = 70.83% In order to illustrate the readers' response, let us consider terms 4 and 6 on the open form "And if ye are unclean, purify yourselves" (Pickthall 1980: 135) and "If you are defiled, purify yourselves" (Arberry 1980:1:128) in order to see how the participants in the study responded respectively to the instructions "Give examples of people who are unclean" and "Give examples of 'people' who are in a state of being defiled. How and with what would they purify themselves? It should be noted that no native speaker of Arabic will miss the message of the Quranic verse in the original "wa'in kuntum junuban faṭṭahharuu" (al-maa'idah: 6) since the predicates junub (= having semen on oneself due to ejaculation by any imaginable 122 means) and taṭahhara (= to take a bath) both belong to the general, unmarked register in Arabic. From a translational perspective, Pickthall and Arberry both use substitution, that is, 'unclean' and 'defiled' for junuban, and both translations use a literal rendition, namely, 'purify', for faṭṭahharuu. Unfortunately, these decisions seriously distort the meaning of the terms of the original. Consequently there is a breakdown of communication in these translations. This can be clearly seen from the participants' interpretations of 'unclean' and 'defiled' and the way they thought one could purify oneself as indicated below: a) Sample interpretations of 'unclean' and 'defiled': 1. atheists, worshippers of idols, murderers, common criminals and thieves. 2. possibly sinners - those who have done wrong against society or others. 3. eaten the wrong food, unwashed, morally corrupt. 4. criminal people, atheists. 5. any sinner in a Biblical sense, anyone from a leper or a whore to an outcast or a low class person. 6. have touched certain criminals, dead people, etc. b) Sample interpretations of the ways in which one could purify oneself: 1. washing and obtaining forgiveness. 2. doing good and repenting their sins or wrongdoings. 3. asking forgiveness and seeking to lead a better life. 4. through prayer and worship. 5. prayer, godly acts, renouncing their sins. It is clear that the distortion of the messages in the translated terms is caused by the shift from a specific, material sense to a general, spiritual sense. This being the case, junuban, which is the natural consequence of a specific, material happening that is completely unmarked for religious appropriateness (which would qualify it in terms of the virtue vs. sin dichotomy), is construed in a general, spiritual sense that is denotatively related to sin. This twist in denotation amounts to a major difference in meaning between the original and the translation that impedes the understanding of the intended message. Similarly, most participants interpret faṭṭahharuu in a spiritual sense, which runs counter to the intended material sense of the original, namely, to remove the ejaculated semen by means of bathing. The findings of the closed form (in Table 5) show that holders of a Bachelor's degree ranked highest in the rating of comprehensibility of Quranic verses at 16.41% ("a lot of sense") and 41.41% ("straightforwardly understood") vs. 13% and 32% for the Master's and Ph.D. degree holders, respectively. High school students lagged negligibly behind at 18.75% and 25%, 123 respectively. There is, therefore, no positive correlation between the level of education and the rating of comprehensibility in this study. Table 5. Results of the Education Variable (The Open Form) Degree No sense at all Very little sense Satisfactory sense A lot of sense Straightfor ward ly understood Total No. No. of subjects No. % No. % No. % No. % No. % No. % 1. High school (1) 16 20 17 21.25 12 15 15 18.75 20 25 80 100 2. Bachelor’s D. (8) 4 3.13 20 15.62 30 23.43 21 16.41 53 41.41 12 8 100 3. Master’s D. (6) 8 12.5 13 20.31 14 21.88 9 14.06 20 31.25 64 100 4. Ph. D. (4) 6 12.5 5 10.42 15 31.25 6 12.5 16 33.33 48 100 The education variable can, however, also be examined in the open form questionnaire, which was, as mentioned, more reliable for uncovering participants' responses (Table 6). Although the figures are more indicative of the participants' comprehension than those of the open questionnaire, it is not enough to be statistically significant, especially when considering the generally poor performance of the participants in all groups. Table 6. Results of The Education Variable (The Open Form) Degree No response No sense at all Very little sense Satisfactory sense A lot of sense Straightforwardly understood Total No . No. of subjects N o. % No. % No. % No. % No. % No. % No. % 1. High school (1) 10 34.48 8 27.58 4 13.79 3 10.34 2 6.89 2 6.89 29 99.97 2. AA. (1) 3 10.34 11 37.93 0 0 2 6.89 6 20.69 7 24.14 29 99.99 3. Bachelor 53 22.84 64 27.58 43 18.53 17 7.33 18 7.74 37 15.95 23 99.99 124 (8) 2 4. Master’s (6) 47 27.01 53 30.46 26 14.94 14 8.05 19 10.92 15 8.62 17 4 100 5. PhD (4) 8 6.90 53 45.69 18 15.52 6 5.17 7 6.03 24 20.69 11 6 100 5. Conclusion This study has tried to explore the problem of referential gaps in the translation of unmatched cultural elements by drawing evidence from select translations of Quranic verses. We assumed that such gaps impeded cross-cultural communication; reader responses proved this to be the case. This fact calls for immediate, practical solutions. Even when it is kept in mind that perfect communication is impossible both intralingually and interlingually, it is thought-provoking that most of the translations used in the study failed to convey the source language message into the target language. Accordingly, they introduce false conceptions about the Holy Quran which is an extremely serious fact. For instance, one Western PhD holder participating in the study did not understand the following translation correctly "... If you have been in contact with women, and ye find no water, then take for yourselves clean sand and earth" (Ali 1934:242), commenting that "it may be helpful to notice that as a Westerner and a woman, I find the idea that I am considered unclean and a source of pollution offensive." It must be explained that the Quranic verse is not offensive to anybody, man or woman; it simply states that a man has to clean himself with earth if he had been in contact with women and finds no water before performing prayers. It can be concluded that readers' response should be considered a key variable in translation. In this study we have used passages from the Holy Quran. We can therefore claim that our findings apply to Arabic religious discourse in general and the Holy Quran in particular. The claim that the authoritativeness of the religious text is the determining factor when it comes to decisionmaking in translation should be reconsidered in light of the reader response variable because, in the final analysis, a translation is an act of communication. As such it cannot operate in a vacuum in which messages are blocked, regardless of how formal, poetic, elevated, and refined the style of the original. Translators should endeavor to convey the communicative value of the referential gap by any means, and not use problematic paraphrase and literal translation, such as those used by Arberry who rendered 'private parts' for "genitals" and 'what their right hands own' for "their woman servants" (Arberry vol.1 1980: 128). Consequently, when semantic translation (Newmark 1988) and semantic equivalence (Widdowson 1971) fall short of comprehensibility, translators should opt for `communicative translation' (Newmark 1988), or 'functional equivalence' (de Waard and Nida 1986), or 'ideational equivalence' (Farghal 1994) by employing one or more of the translation strategies at their disposal when they encounter referential gaps. 125 Last but not least, it should be borne in mind that theory may sometimes be far from practice. We wish to emphasize that this study is not meant to depreciate the efforts of the renowned translators whose work we have used. However, it serves to bring into focus reader response as an important variable in the translation of religious discourse and even on a small scale, as shown in this study. The mission of translating the meanings of the Holy Quran will invariably be in for more criticism than praise. All the same, we have to pay homage to those outstanding translators of the Holy Quran whose efforts and contributions are highly acknowledged throughout the world. Appendix I Dear Friend/Colleague, This questionnaire is intended to be the basis for a study on translation problems. It deals basically with the translation of certain elements from the Arab-Islamic culture into English. The study takes evidence from the Holy Quran for completely academic purposes. The questionnaire involves a number of selected translations of certain Quranic verses. You are kindly requested to tick () the box that best fits your understanding of the verse that corresponds to it. While trying to understand the content of each verse, focus on the meaning of the words in bold type. Notice that you should deal with each utterance as a completely independent entity. Be sure that your responses to this questionnaire will not be circulated and that they will only be used for purely academic purposes. Please be informed that you may not consult references in the course of responding to this questionnaire. Personal information: Name (optional):_______________________________________ Nationality:___________________________________________ Age:_________________________________________________ Sex: Male Female Bachelor’s Degree Education: High School or below Master’s Degree PhD or more General field of Study: _______________________________ 126 The ‘verse’ translated Makes no sense at all Makes very little sense 1- “And He has enjoined me to pray, and to give the alms, so long as I live”. 2- “Purify [Abraham and Ishmael] My house for those who go around and those who meditate therein and those who bow down and prostrate themselves (in worship)”. 3- “And those who say, regarding their wives, “Be as my mother’s back”, and then retract what they have said, they shall set free a slave.” 4- “And if ye [before prayers] are unclean, purify yourselves. 5- “O Prophet ! [Mohammad] when you divorce women, divorce them when they have reached their period. Count the period, and fear God your Lord.” 6- “If you are defiled, purify yourselves.” 7- “Perform the Pilgrimage and the visit (to Mecca) for Allah. And if ye are prevented, then send such gifts as can be obtained with ease.” 8- “Prosperous are the believers, who in their prayers are humble, and from idle talk turn away, and at almsgiving are active, and guard their private parts save from their wives and what their right hands own.” 127 Makes satisfactory sense Makes a lot of sense Staightforwardly understood 9- “... If you have been in contact with women. and ye find no water then take for yourselves clean sand or earth..” 10- “Whatever alms ye spend or vow ye vow, lo ! Allah knoweth it all. 11- “And say to the believing women that they should lower their gaze and guard their modesty,.... and should not display their beauty except to their husbands, their fathers,.. or small children who have no sense of the shame of sex”. 12- “If you are in a state of ceremonial impurity, bathe your whole body”. 13- “O ye who believe! when ye rise up for prayer, wash your faces, and your hands... . And if ye are sick or on a journey, or one of you cometh from the closet, or ye have had contact with women, and ye find no water, then go to clean, high ground and rub your faces and your hands with some of it”. 14- “And whatever expenditure you expend, and whatever vow you vow, surely God knows it”. 15- “They will question thee concerning the monthly course. Say: It is hurt; so go apart from women during the monthly course, and do not approach them till they are clean. When they have cleansed themselves, then come unto them as God has commanded you”. 128 16-“O Prophet! [Mohammad] when ye (men) put away women, put them away for their (legal) period and reckon the period. Appendix II Dear Friend / Colleague, This questionnaire is intended to be the basis for a study on translation problems. It deals basically with the translation of certain elements from the Arab-Islamic culture into English. The study takes evidence from the Holy Quran for completely academic purposes. The test involves a number of selected translations of certain Quranic verses. You are kindly requested to write in simple English what you understand from the “verses”, giving special attention to the word(s) written in bold type, when the utterance seems too long. When it involves one or two short sentences, try to paraphrase it in your own words. Along with some utterances, you will find certain questions. Try to make sure that you entirely depend on the utterance when you kindly answer them. Notice that you should deal with each utterance as a completely independent entity. Be sure that your responses to this questionnaire will not be circulated and that they will only be used for purely academic purposes. Please be informed that you may not consult references in the course of doing this task. Personal information: Name (optional):_______________________________________ Nationality:___________________________________________ Age:_________________________________________________ Sex: Male Female Education: High School or below Master’s Degree Bachelor’s Degree Ph.D. or more General field of Study:______________________________________. 1. “And He has enjoined me to pray, and to give the alms, so long as I live”. 129 Arberry, vol.1 (1980: 333) * Is ‘almsgiving’ optional or compulsory? 2. “Purify [Abraham and Ishmael] My house for those who go around and those who meditate therein and those who bow down and prostrate themselves (in worship).” Pickthall (1980: 23). * How would ‘purification’ take place? What would one use to purify Allah’s house? * Differentiate between meditate, bow down, and prostrate. 3. “And those who say, regarding their wives, ‘Be as my mother’s back’, and then retract what they have said, they shall set free a slave.” Arberry, vol.2 (1980: 263). * Does this make ‘their wives divorced? Explain. 4. “And if ye [before prayers] are unclean, purify yourselves.” Pickthall (1980: 135). * Give examples of people who are unclean. 5. “O Prophet! [Mohammad] when you divorce women, divorce them when they have reached their period. Count the period, and fear God your Lord.” Arberry, vol.2 (1980: 284). 6. “If you are defiled, purify yourselves.” Arberry, vol.1(1980: 128). * Give examples of ‘people’ who are in a state of being defiled. How and with what would they purify themselves? 7. “Perform the Pilgrimage and the visit (to Mecca) for Allah. And if ye are prevented, then send such gifts as can be obtained with ease.” Pickthall (1980: 37). * Differentiate between ‘Pilgrimage’ and visit (to Mecca). * Give examples of ‘gifts’ and how/where they can be sent. 8. “Prosperous are the believers, who in their prayers are humble, and from idle talk turn away, and at almsgiving are active, and guard their private parts save from their wives and what their right hands own.” Arberry, vol.2 (1980: 37). 130 9. “...If you have been in contact with women, and ye find no water, then take for yourselves clean sand or earth...” Ali (1934: 242). * Does the phrase ‘contact with women’ involve sexual moves? * How would one take for himself/herself ‘clean sand or earth’? 10. “Whatever alms ye spend or vow ye vow, lo! Allah knoweth it all. Pickthall (1980: 56). * Give examples of ‘vows you vow’. 11. “And say to the believing women that they should lower their gaze and guard their modesty, ... and should not display their beauty except to their husbands, their fathers, ... or small children who have no sense of the shame of sex”. Ali (1934: 905). 12. “If you are in a state of ceremonial impurity, bathe your whole body.” Ali (1934: 242). * Give examples of one who is ceremonially impure. 13. “O ye who believe! When ye rise up for prayer, wash your faces, and your hands...... . And if ye are sick or on a journey, or one of you cometh from the closet, or ye have had contact with women, and ye find no water, then go to clean, high ground and rub your faces and your hands with some of it.” Pickthall (1980: 135-6). 14. “And whatever expenditure you expend, and whatever vow you vow, surely God knows it”. Arberry, vol.1 (1980: 68). * What may ‘expenditure’ involve? 15. “They will question thee concerning the monthly course. Say: ‘It is hurt; so go apart from women during the monthly course, and do not approach them till they are clean. When they have cleansed themselves, then come unto them as God has commanded you”. Arberry, vol.1 (1980: 59). * What exactly is the monthly course? * How would women cleanse themselves? 16. “O Prophet! [Mohammad] when ye (men) put away women, put them away for their (legal) period and reckon the period.” Pickthall (1980: 747) 131 132 Coherence Shifts in Quran Translation Mohammed Farghal & Noura Alblushi Abstract This paper aims to shed light on reader-focused and text-focused coherence shifts in Quran translation. The data consists of selective examples involving potential problems related to such shifts, which are excerpted from five well-known English translations of the Holy Quran. The results show that Quran translators sometimes fall victim to such shifts. On the one hand, the wide cultural distance between source text and target text may cause reader-focused shifts when opting for literal translation in handling partial and complete referential gaps. On the other hand, the translator’s inadequate language competence in Arabic and the absence of consulting Quranic exegeses sometimes trigger text-focused shifts, thus offering unintended readings. The study concludes that coherence shifts constitutes a serious problem in Quran translation, which calls for remedial work in future endeavors. 1. Introduction Blum-Kulka (2004: 291) defines coherence as “a covert potential meaning relationship among parts of a text, made overt by the reader or listener through processes of interpretation”. She equates coherence with the text`s interpretability, thus considering general changes or loss in meaning of the Source Text (ST) through translation as affecting its coherence. More interestingly, she divides coherence shifts into two categories; the first results from the Target Text (TT) reader’s failure to make sense of the ST because of different world views, which leads to reader-focused coherence shifts, while the second involves text-focused shifts of coherence which result from mistranslations. When examining coherence shifts, Blum-Kulka argues that distinguishing the two types of shifts is important because it helps in having “a better understanding of what translation can and cannot do, or, in other words, to better understand the true limits of translatability” (p. 297). Translatability can be defined as “the capacity of some kind of meaning to be transferred from one language to another without undergoing radical change" (Alpert, 2001: 273). Being the Arabic Word of God verbatim, the Holy Quran constitutes the most important source of authority for Muslims. Hence, Quran translation into foreign tongues has always been a controversial issue among Arab/Muslim scholars since Medieval times (for more details, see Mustafa, 2001). However, there is a consensus nowadays that existing Quran translations should be taken as interpretations of the sacred text rather than exact translations. In particular, this issue was resolved when religious bodies, such as Al-Azhar (Egypt), gave permission for translating the Quran provided that Quran translators explicitly state that their translations are not replacements 134 of the ST and that they are merely rephrasings based on the agreed upon meaning of the original. Despite this, one can readily observe that most Quran translations are source-oriented and hardly, if at all, take the needs of the target audience to access the intended meanings into consideration (for more details, see Mustafa, 2001). The present study will draw selectively on examples excerpted from 5 translations of the Holy Quran as follows: (1) The Message of the Qur’an by Muhammad Asad, 1980/2003. (2) The Koran Interpreted by A. J. Arberry, 1930/1996. (3) The Qur'an; Text, Translation and Commentary by Abdullah Yusuf Ali, 1934/2005. (4) The Qur’an the Noble Reading by T. B. Irving, 1985/1993. (5) The Glorious Quran by Muhammad M. Pickthall, 1930/2006. 2. Reader-focused Coherence Shifts Blum-Kulka states that reader-focused coherence shifts occur “as a result of a text being read by culturally different audiences” (2004: 305). She argues that this kind of shift is unavoidable because, in most cases, a TT is likely to be new to the TL readers. Because of this cultural distance, TL culture might not share the same cultural assumptions, beliefs and value system recognized in the ST. This creates a void in translation, and in this case, the translator`s task is to fill any cultural voids that may impede the TT readers` interpretability of the TT; failure to do so may result in ineffective translation. In the case of the Quran, in a text that is deeply rooted in the source culture, the translator will often be confronted with culture-bound expressions that are difficult to convey to the TT readers (The TT reader is assumed to be the average Western reader). The Quran translator Mohammad Asad (2003: vii) notes that the linguistic gap between Arabic and English is a source of coherence shifts to a TT reader because “the coherence of the Quranic world view and its relevance to the human condition escape him altogether and assume the guise of what, in Europe, is frequently described as ‘incoherent ramblings’”. Therefore, examining potential reader-focused coherence shifts is a way towards rendering a more coherent Quran translation. Blum-Kulka (2004: 298) also notes that the normative system dominating the translation process can contribute to creating reader-focused coherence shifts in translation. In Quran translation, this normative system, being overwhelmingly source-oriented, hardly accommodates to the needs of TT readers. However, in their endeavors to supplement translations, most Quran translators resort 135 to long introductions to suras and a great amount of explanatory notes to bridge the cultural gap where potential reader-focused coherence shifts might arise. The following discussion will reveal that most reader-focused coherence shifts are found in translations that opt out of using parenthetical material and/or footnotes (for more details on strategies to deal with culture-bound items, see Newmark 1988 and Larson 1998), while those that employ them reduce the number of reader-focused coherence shifts. The area of reference is the primary source of this kind of shifts; unawareness of a referential item is likely to impede the TT reader`s comprehension and ability to interpret the TT coherently. These culture-bound expressions in the translation of the Quran constitute either partial or complete cultural/referential gaps that need to be filled in the process of translation, in order to prevent the occurrence of reader-focused shifts. A partial referential gap in translation, as the name indicates, relates to a referential item that exists in both languages and cultures with different implications. A familiar Islamic concept which constitutes a partial referential gap is the concept of zakat الزكا. This refers generally to a certain amount of money gathered from the well-to-do and paid to the poor; it is also obligatory and constitutes the third of the five pillars of Islam. The principles and rules encompassed in this term make it a challenging concept to render fully in translation. In the following verse, people are urged to pay zakat and are assured that, by doing so, God will reward them. Consider the verse and the translations: يي وأصبي ُاوا ٱليل ٰو و اتُوا الزكاة ودا تُق ِّ ُدوا بأل فُ بس ُا ِّد ْ خ ْي ٍ ت بج ُو ُ بعن ٱّللب إبن ٱّلل صباا تعْالُون ص ب ١١١﴿ ) ال ق And be constant in prayer, and render the purifying dues; for, whatever good deed you send ahead for your own selves, you shall find it with God: behold, God sees all that you do. (110) (Asad, p. 32) Establish worship, and pay the poor-due; and whatever of good you send before (you) for your souls, you will find it with Allah. Lo! Allah is Seer of what you do. (110) (Pickthall, p. 18) Both translators employ the word due(s), i.e. something required, to denote the obligatory sense of zakat in the TT. This modification succeeds in conveying the main aspect of the term; however, the translators differ in the choice of the generic word chosen to render it. Asad derives his rendition from the spiritual connotations of zakat: he notes in his footnote that its main function is to “purify a person`s capital and income from the taint of selfishness” (p.18), thus basing his translation on the connotative meaning of the term. By contrast, Pickthall derives his translation from the category of people who are eligible to receive it, so he renders it as ‘the poordue’. In this way, both translators attempt to approximate the concept to TT readers by defining it 136 in their footnotes as an obligatory tax and provide sufficient amount of technical information about it, in order to prevent a reader-focused shift. For his part, Ali employs “charity” as an equivalent for zakat: And be steadfast in prayer and regular in charity: And whatever good ye send forth for your souls before you, ye shall find it with God. For God sees Well all that ye do. (110) (Ali, p. 48) It should be noted that in the SL culture the concept of zakat is very specific and is associated with obligatory giving, so the term ‘charity’, which is associated in the TL culture with voluntary giving is too general. Moreover, in the SL culture voluntary giving is associated with another term, that is sadaqah, thus ‘charity’ becomes a more appropriate rendition for sadaqah but not zakat. In this way, Ali`s translation seriously diverges from what is meant by the Islamic concept of zakat, and without a footnote, the relevant features required for the full and coherent interpretation of the term are lost in his translation. Similar changes can be found in Arberry and Irving`s renditions of zakat. Consider their translations: And perform the prayer, and pay the alms; whatever good you shall forward to your souls' account, you shall find it with God; assuredly God sees the things you do. (110) (Arberry, Vol. 1, p. 42) Keep up prayer and pay the welfare tax; you will find any good you have sent on ahead for your own souls' sake is already [stored up] with God. God is Observant of whatever you do. (110) (Irving, p. 9) Arberry employs ‘alms’ as an equivalent for zakat. Collins Cobuild Advanced Learner`s Dictionary defines ‘alms’ as “gift of money, clothes, or food to poor people” (2003). This lexical item is too general and does not have the same denotative meaning of the Islamic term. In addition, the nature of these alms, defined above as being ‘gifts’, i.e. given with free will, diverges from the obligatory nature of zakat. Likewise, different implications might arise from Irving`s translation. He uses Modern American English in his translation of the Quran and renders zakat as ‘welfare tax’. Target readers may derive different implications from this rendition since it carries different associations from those of zakat in the Islamic context. Welfare tax pertains generally to the amount of money paid by all people, rich and poor alike, to the government for the advancement of society as a whole. Without a footnote or any other means of explication, these readers are likely to interpret this term in a different way from that intended in the source text. For example, without specifying that zakat is obligatory for the well-to-do and not the poor, the readers might infer that it is obligatory for the 137 rich and poor alike. This inference does not serve the message of zakat which aims at compassion and social justice rather that burdening the poor. Both Arberry and Irving`s translations may lead to reader-focused coherence shifts. Let us now examine the translation of a complete referential gap uddah to see how it can lead to reader-focused coherence shifts. The term uddah is a Quranic euphemism referring to the legally prescribed waiting period before a divorced woman can remarry; the legal period is prescribed to rule out pregnancy. Consider the following verse along with Asad`s translation: ۟ الطًق) ٰ َٰٓأ ُّاا ٱلن ُّى إبذا طل ْق ُ ُ ٱلنِّسآَٰ فطلِّقُومُ لعدتهن وأاْ ي١﴿ ُوا ٱلعدة ب O Prophet! When you intend to divorce women, divorce them with a view to the waiting period appointed for them, and reckon the period [carefully]. (1) (Asad, p. 994) Asad approximates this concept to TT readers by explicating its meaning of the first occurrence of the term through using a descriptive phrase, thus rendering it as ‘the waiting period appointed for them’. He also supports his translation with sufficient information in a footnote to clarify the meaning of this Islamic concept. Having defined uddah in the first instance, he then renders the second instance generically as ‘the period’ to ensure the consistency of his translation. Ali and Pickthall also explicate this meaning of this term in their translations by rendering it as the “prescribed period’ and ‘(legal) period’, respectively. However, a significant reader-focused coherence shift can be found in Irving`s translation. His translation lacks clarity and consistency as he translates the first instance as ‘[legal] number’ while the second as ‘amount of months’ as can observed below: O Prophet, whenever you [and other Muslims] divorce women, send them away according to their [legal] number and count up the amount [of months carefully]. (1) (Irving, p. 332) This translation is vague for two reasons. Firstly, neither renditions conveys the full sense of uddah, so both lack equivalence. Secondly, the translation lacks consistency because Irving translates the two occurrences of uddah in this verse differently. Therefore, it is very unlikely that TT readers will get the intended message, the result being a reader-focused coherence shift. For his part, Arberry employs the word ‘period’ as an equivalent for the term uddah, but his translation can be misleading, too, as can be observed below: 138 O Prophet, when you divorce women, divorce them when they have reached their period. Count the period. (1) (Arberry, Vol. 2, p. 284) Farghal and Al-Masri (this volume) note that using ‘period’ as an equivalent for uddah may lead target readers to confuse it with the menstrual cycle in women because “when the text was given to a group of 20 American native speakers, about 75% of them provided interpretations relating to the monthly period rather than the intended legal sense” (p.151). Without explicating the meaning, this translation will certainly cause a reader-focused coherence shift. To sum up, when a text involves a partial or a complete cultural gap, on the one hand, target readers will most likely interpret the message according to their own culture and experience of the world. The translator`s task is, therefore, to bridge the gap, as failure to do so may result in a reader-focused coherence shift. Notably, cultural gaps are unavoidable because language systems and cultures do not share the same reference networks. When dealing with such gaps in Quran translation, it is vital for translators to address them by employing appropriate strategies, including borrowing and paraphrase combined with descriptive modification and, in several cases, footnotes. On the other hand, when literal translation is employed, it usually fails to deliver the complete meaning of the cultural item, thus affecting the meaning of the Quran and its coherence in translation. Literal translation often fails because Arabic and English are two languages that are radically different, and the non-existence of equivalent concepts usually leads to incongruity in the meanings offered by English translations of Quranic culture-bound expressions. 3. Text-focused Coherence Shifts Whereas reader-focused coherence shifts are mainly incurred to fill any cultural void that may impede TT readers` interpretability of the translation, text-focused coherence shifts result primarily from the translation process and affect the ST’s meaning potential in translation. These shifts occur “as a result of particular choices made by a specific translator, choices that indicate a lack of awareness on the translator`s part of the SL text`s meaning potential” (Blum-Kulka, 2004: 301). Consequently, the TT will offer an interpretation that is not intended by the ST because of a mistranslation committed by the translator. Some text-focused coherence shifts may be caused by the translator`s unconscious interventions in the ST`s meaning potential, as can be seen in the translation of the following verse which involves an Islamic teaching: ،إن ٱليفا و ْٱلا ْ و بد شعآَٰئب ب ٱّللب ۖ فا ْ اج ْٱل يْت أ بو ٱ ْع ا فً جُنا عل ْي به أن يطوف بهما ۚ ود تطو ١٥١﴿ ا فإبن ٱّلل شا بك علبي، ًۭ خ ْي ) ال ق 139 [Hence,] behold, As-Safa and Al-Marwah are among the symbols set up by God; and thus, no wrong does he who, having come to the Temple on pilgrimage or on a pious visit, strides to and fro between these two: for, if one does more good than he is bound to do - behold, God is responsive to gratitude, all-knowing. (158) (Asad, p. 42) Safa and Marwa are some of God's waymarks. Anyone who goes on Pilgrimage to the House or visits [it] will not be blamed if he runs along between them. With anyone who volunteers some good, God is Appreciative, Aware. (158) (Irving, p. 13) Al-Safa and Al-Marwah are two hills between which Muslims are required to travel back and forth seven times during the annual pilgrimage and the Lesser pilgrimage. Therefore, طوف صب بااا (Literally ‘go round them’) signifies running, or pacing, between these two hills. This is rendered accurately by Asad and Irving in the above translations because they opt out of literal meaning in favor of intended meaning. However, text-focused coherence shifts occur in the translations below: Behold! Safa and Marwa are among the Symbols of Allah. So if those who visit the House in the Season or at other times, should compass them round, it is no sin in them. And if any one obeyeth his own impulse to good,- be sure that God is He Who recogniseth and knoweth. (158) (Ali, pp. 62-63) Safa and Marwa are among the waymarks of God; so whosoever makes the Pilgrimage to the House, or the Visitation, it is no fault in him to circumambulate them; and whoso volunteers good, God is All-grateful, Allknowing. (Arberry, Vol. 1, p. 48) Lo! (the hills) Al-Safa and Al-Marwah are among the indications of Allah. It is therefore no sin for him who is on pilgrimage to the House (of Allah) or visiteth it, to go around them (as the pagan custom is). And he who does good of his own accord (for him), lo! Allah is Responsive, Aware. (158) (Pickthall, 158, p. 24) 140 In these translations, Ali, Arberry and Pickthall render طوف صب باااrespectively as ‘compass them round’, ‘circumambulate them’ and ‘go around them’. These renditions create a reality that is completely different from the intended one, viz. going round two hills vs. traveling back and forth between two hills. This text-focused coherence shift committed by these translators distorts the meaning in the TT and is likely to conjure up completely different pictures in the minds of TT readers as to how this ritual is performed. Furthermore, Pickthall`s bracketed addition of the phrase (as the pagan custom is) detracts from the meaning and is completely irrelevant. To avoid similar text-focused shifts, translators have to invest contextual elements (e.g. It would be too difficult for pilgrims to go round two hills) or, if the context does not help, resort to exegeses. In addition to inaccurate decisions by translators, Blum-Kulka (2004) also indicates that the most serious text-focused coherence shifts occur as a result of the translator`s failure “to realize the functions a particular linguistic system, or particular form, plays in conveying indirect meanings in a given text” (p. 301). To demonstrate this point, consider the following verse which employs the comparative structure to explain a specific Islamic teaching: ۟ ُٰ َٰٓأ ُّاا ٱل به ادن ٰ ًۭدا د ْع، ) أ ا١١١﴿ وا ُك بب عل ْي ُا ُ ٱليِّيا ُم كاا ُك بب على ٱل به بد ص ْلب ُا ْ لعل ُا ْ ت قُون ت ٍ ٍۢ ُوو ،ا أوْ عل ٰى وف ٍۢ ٍ ف بع ًۭ ِّد ْ أ ٍام أُخ وعلى ٱل به ُ بطيقُو ۥهُ فب ْ ًۭ طعا ُم بد ْس باي ٍۢ ٍ فا تطو،فا كان بدن ُا د ب ح ۟ صو ُم ) ال ق١١١﴿ وا خي ٌۭر ل ُكم إبن ُكن ُ ْ تعْل ُاون ُ ا فاُو خ ْي ًۭ ل ۥه ُ وأن ت، ًۭ خ ْي O ye who believe! Fasting is prescribed for you, even as it was prescribed for those before you, that ye may achieve piety; (183) (Fast) a certain number of days; and (for) him who is sick among you, or on a journey, (the same) number of other days; and for those who can afford it with hardship there is a ransom: the feeding of a man in need - but who does good of his own accord, it is better for him: and that you fast is better for you if you did but know. (184) (Pickthall, p. 27) The above verses order Muslims to fast during the month of Ramadan and permits those who are unable to fast or are ill to make up for the days they have missed by either feeding the needy or fasting the same number of days later. Although Muslims are given two options, the comparative ۟ ‘ وأن تيُو ُدand if you fast it is better for you’ to demonstrate that structure is employed in ْ وا خ ْي ًۭ ل ُا making up for the days missed in Ramadan by fasting on other days is better than the other option. This specific instruction is rendered accurately in the above translation by Pickthall, whereas Asad`s failure to preserve the comparative relation that holds in the ST distorts the ST’s meaning in his translation below: O you who have attained to faith! Fasting is ordained for you as it was ordained for those before you, so that you might remain conscious of God: (183)[fasting] during a certain number of days. But whoever of you is ill, or on a journey, [shall fast instead for the same] number of other days; and [in such cases] it is incumbent upon those who can afford it to make sacrifice by feeding a needy person. And whoever does more good than he is bound to do 141 does good unto himself thereby; for to fast is to do good unto yourselves if you but knew it. (184) (Asad, p. 49) English does not lack the tools to construct a comparative structure, but the shift was caused due to the translator`s failure to realize the role of the comparative relation in delivering the proposition in the ST. As a result, Asad’s translation fails to deliver what the verse conveys. Similarly, the translator`s failure to realize the function of a preposition may result in a textfocused coherence shift. The following examples demonstrate how a literal translation of the preposition ‘ فin’ changes the intended meaning of the ST and leads to different implications. ُ إب ْذ صال ُدووى أل ْملب به إب ِّ آ س ) فلاا٧( س لعل ُا ْ تيْ طلُون ٍ ا وآتبي ُا ِّد ْناا صبخ ٍ أوْ آتبي ُا صب بشاا ٍ ص،ْت ار ) النال١( ُورك د في النار ود ْ اوْ لاا و ُو ْحان هللاب ر ِّ ْالعال باي جا ما ُو بو أن ص ب Lo! [while lost in the desert] Moses said to his family: “Behold, I perceive a fire [far away]; I may bring you from there some tiding [as to which way we are to pursue], or bring you [at least] a burning brand so that you might warm yourselves. (7) But when he came close to it, a call was sounded: “Blessed are all who are within [reach of] this fire, and all who are near it! And limitless in His glory is God, the Sustainer of all the worlds” . (8) (Asad, p. 643) Behold! Moses said to his family: "I perceive a fire; soon will I bring you from there some information, or I will bring you a burning brand to light our fuel, that ye may warm yourselves. (7) But when he came to the (Fire), a voice was heard: "Blessed are those in the Fire and those around: and Glory to God, the Lord of the Worlds. (8) (Ali, p. 979) Exegeses explain that the preposition فmust not be understood literally, so the phrase ُورك د فب ص ب ار ‘ الن بLit. Blessed be those in the fire’ means those who are within the area illuminated by the fire or within reach of it, an import that is rendered accurately by Asad. A literal translation like the one offered by Ali, as well as Pickthall, Arberry and Irving (not quoted here to avoid repetition), may lead to the wrong inference, possibly to the interpretation that those doomed to Hell are blessed by God. This mistranslation diverges from the intended meaning of the ST and is likely to result in a text-focused coherence shift. Sometimes, the phonological similarity between two function words may cause a text-focused coherence shift. The following example involves two heteronyms: the demonstrative adverb 142 өamma ‘there’ and the temporal conjunction өumma ‘then’, which are identical in the absence of voweling. Whereas Aberry is sensitive to the orthographical cue signaled by the first vowel, thus rendering them appropriately, Irving is miscued by the apparent phonological similarity, thus producing a text-focused coherence shift, as can be observed in the two translations below. ﴾ وأ ْرل ْفنا ثم٣١﴿ ق كالطوْ بو ْالع بظي ب ٍ ْ فأوْ ايْنا إبلى ُدووى أ بن اضْ ب صِّعياك ْال حْ فا فلق فاان ُكلُّ فب ﴾٣٣﴿ ﴾ ثُم أ ْغ ْصنا ْاآلخ ب٣٥﴿ ﴾ وأ جيْنا ُدووى ود دعهُ أجْ ا بعي٣١﴿ (الشع ا ) ْاآلخ ب Then We revealed to Moses, 'Strike with thy staff the sea'; and it clave, and each part was as a mighty mount. (63) And there We brought the others on, (64) and We delivered Moses and those with him all together; (65) then We drowned the others. (66) (Arberry, Vol. 2, p. 67) So We inspired Moses as follows: "Strike the sea with your staff!" It opened up and each section was like a huge cliff. (63) We brought the others up next, (64) and saved Moses and all those who were with him. (65) Then We let the rest drown. (66) (Irving, p. 201) The mistranslation committed by Irving changes the relation holding between verses 63 and 64 from an indication of a place into a temporal relation. Referral to exegeses could have prevented the shift caused by heteronyms in the above example, but the translation of the Quran could involve other linguistic phenomena which, combined with the translator’s lack of awareness of the ST`s meaning potential, can be a potential source of text-focused coherence shifts. The following example involves lexical homonymy, a linguistic phenomenon where lexemes share the same spelling and pronunciation but have unrelated meanings. Arberry`s translation of ا، ًۭ ‘ ش بايwitness’ in the verse below fails to deliver the accurate meaning because he confuses it with the Arabic word for ‘martyr’ , which is also spelled and pronounced the same way. The translations given by Irving and Pickthall deliver the accurate meaning, but the one offered by Arberry is erroneous: ۟ ت أو ٱ ف ب ۟ ۟ ۟ ٰ ) وإبن بدن ُا ْ لا ليُ طِّئ فإ ب ْن أ٧١( عا، ًۭ ُوا ج باي ص ْ ُا ادنُوا ُخ ُهوا با ْهر ُك ْ فٱ فب ُوا ثُ ا ٍ ب ) النسا٧٧( يي ًۭ صال ص ْ أ ْع ٱّللُ على إب ْذ ل ْ أ ُك دعاُ ْ شه ٌۭيدا ُّد ب ٰ َٰٓأ ُّاا ٱل به You who believe, take your precautions and march off in detachments, or march off all together. (71) Among you there are some who procrastinate. If any disaster strikes you, they say: “ God has favored me, for I was not a witness along with them. (72) (Irving, p. 45) 143 O you who believe! Take your precautions, then advance in groups, or advance all together. (71) Lo! among you there is he who loiters; and if disaster overtook you, he would say: Allah has been gracious unto me since I was not present with them. (72) (Pickthall, p. 78) O believers, take your precautions; then move forward in companies, or move forward all together. (71) Some of you there are that are dilatory; then, if an affliction visits you, he says, 'God has blessed me, in that I was not a martyr with them.' (72) (Arberry, Vol. 1, p. 110) The word ا، ًۭ ش بايin the above example is an exaggerated form derived from شام, i.e. ‘witness’ or ‘be present’, but it shares a relation of homonymy with the Arabic word for ‘martyr’ due to the transformation undergone during the derivation process. It is rendered accurately by Irving and Pickthall, but Arberry mistakes it for ا، ًۭ ش باي, the Arabic word for ‘martyr’. This results in a mismatch of the denotative meaning between the ST and the TT. Arberry`s erroneous rendition would also lead TT readers to derive implications other than those intended in the ST. In the above verse, the believers are urged to be careful during battles by advancing in consecutive groups or advance all together in a single group. The verses then warn believers of the hypocrites who join them in battles but loiter behind in an attempt to avoid fighting. The verses then narrate the hypocrites thanking God for staying behind and not being present at the time of the fight. However, Arberry`s translation presents a different understanding; his rendition of ا، ًۭ ش بايas ‘martyr’ gives the impression that they were present in the battle ground, which implies an opposite understanding from that present in the ST. This translation leads to a text-focused coherence shift. Usually, any ambiguity surrounding homonymy can be resolved through context, but homonyms can be problematic if they occur in a context that tolerates ambiguity, a case in point is the above example, where the verse that involves a battle might have contributed to confusing ا، ًۭ ش بايwith the Arabic word for ‘martyr’. The same confusion causes a text-focused coherence shift in the following example which also involves homonymy. In the verse below, ا، ًۭ ْ بديrefers to any town or place as Ali accurately renders it. However, Arberry (below) and Asad (pp. 20 – 21) mistakenly understand it as a reference to Egypt, as the lexical item is homonymous. Witness the translations below: 144 ُ ُْخ بجْ لنا بداا تُ ٍۢن ب ُ وإب ْذ صُ ْل ُ ْ ٰ ُاوو ٰى ل يْ ب عل ٰى طع ٍۢ ٍام ٰو با ٍۢ ٍ فٱ ْو ت ْٱألرْ ضُ بد ٍۢ ص ْقلباا وصبةآَٰئباا لنا رص، ۟ ُوفُو بداا وع بواا وصيلباا صال أت ْس ْ ب لُون ٱل بهى مُو أ ْو ٰى صٱل بهى مُو خ ْي ۚ ٱ ْم ط ۗ ْ ُ وا مص ٌۭرا فإبن ل ُا دا وأ ْل ب ب ۟ ُ ب ِّد ٱّللب ۗ ٰذلب صأ اُ ْ كا ْ ض بص ت ٱّللب و ْق ُلُون الن بيِّي ُ و ٍ ٍۢ ت عل ْي با ُ ٱل ِّهل ُ و ْٱلاسْان ُ وصآَٰ ُ و صبغح وا ْافُ ُون صبـا ٰ ب ب ٰ ۟ ۟ ْ ِّ صبغ ْي ب ٱلح ) ) (ال ق٣١( ق ۗ ذلب صباا عيوا وكا ُوا ْع ُون And remember ye said: "O Moses! we cannot endure one kind of food (always); so beseech thy Lord for us to produce for us of what the earth groweth, its pot-herbs, and cucumbers, its garlic, lentils, and onions." He said: "Will ye exchange the better for the worse? Go ye down to any town, and ye shall find what ye want!" They were covered with humiliation and misery; they drew on themselves the wrath of God. This because they went on rejecting the Signs of God and slaying His Apostles without just cause. This because they rebelled and went on transgressing. (61) (Ali, pp. 32-33) And when you said, 'Moses, we will not endure one sort of food; pray to thy Lord for us, that He may bring forth for us of that the earth produces - green herbs, cucumbers, corn, lentils, onions.' He said, 'Would you have in exchange what is meaner for what is better? Get you down to Egypt; you shall have there that you demanded.' And abasement and poverty were pitched upon them, and they were laden with the burden of God's anger; that, because they had disbelieved the signs of God and slain the Prophets unrightfully; that, because they disobeyed, and were transgressors. (61) (Arberry, Vol. 1, p. 36) The text narrates an incident involving the Children of Israel during their journey after fleeing Egypt. In verse 57, we are told that God sent down unto the Children of Israel food consisting of ‘ الا والسلوىmanna and quails’ to sustain themselves after they fled Egypt. Nevertheless, in the above text they complain to Moses about having to eat the same food everyday and ask him to pray for God to send unto them other kinds of food. They are reprimanded for not showing gratitude and Moses refuses to ask for this kind of food through prayer since it is considered lowly and is available in any city or town. Therefore, as a punishment they were told to seek any city or town to find the food they wanted. They were not told to head to Egypt (the place they were fleeing from) to seek this food. Hence, the translators' misinterpreting ا، ًۭ ْ بديas a reference to ‘Egypt’ changes the meaning of the ST in the TT. Moreover, this shift could have been avoided if the translators were aware of the morphological clues; the fact that ا، ًۭ ْ‘ بديa town/district’ here is inflected indicates that it is the singular of ‘ أديارtowns/districts’. Hence, it needs to be translated as ‘any town or place’. If the reference denoted ‘Egypt’, by contrast, it would be the usually uninflected noun دي. 145 Finally, let us consider the example below where the source of confusion is polysemy (the presence of related senses of the same word). The translator's unawareness of the polysemous nature of the word ْٱل ُافارin the verse below causes a text-focused coherence shift. This word refers to ‘tillers’ or ‘husbandmen’ in the verse, but it was misinterpreted by some translators as a reference to ‘unbelievers’, which also has the same spelling and pronunciation in Arabic. Witness the translations below: ث أ ْعجب ٍ ب ولا ًْۭو و بر ن ًۭ وتفا ُخ ٍۢ صيْن ُا ْ وتااثُ ًۭ فبى ْٱأل ْد ٰو بل و ْٱألوْ ٰل ب ۖ كاة بل غ ْي ًۭ ٱ ْعل ُا َٰٓو ۟ا أ اا ْٱلحي ٰو ُ ٱل ُّ ْيا ل بع ٰ ۚ ًۭاا ۖ وفبى ْٱآل بخ ب عها ًۭ ش ب ًۭ ود ْغفب ًۭ ِّد ٱّللب و برضْ ٰو ًۭن، ا ثُ ُاونُ اُط،ّ ًۭ ٱل ُكفار اتُهۥُ ثُ باي ُج ف ٰىهُ ُديْ ف ٰ ُور ) ) الح ودا ْٱلحي ٰو ُ ٱل ُّ ْيآَٰ إبو د ُ ْٱل ُغ ب.)٧١( Know that the life of the world is only play, and idle talk, and pageantry, and boasting among you, and rivalry in respect of wealth and children; as the likeness of vegetation after rain, whereof the growth is pleasing to the husbandman, but afterward it dries up and you see it turning yellow, then it becomes straw. And in the Hereafter there is grievous punishment, and (also) forgiveness from Allah and His good pleasure, whereas the life of the world is but matter of illusion. (20) (Pickthall, p. 545) Know that the present life is but a sport and a diversion, an adornment and a cause for boasting among you, and a rivalry in wealth and children. It is as a rain whose vegetation pleases the unbelievers; then it withers, and thou seest it turning yellow, then it becomes broken orts. And in the world to come there is a terrible chastisement, and forgiveness from God and good pleasure; and the present life is but the joy of delusion. (20) (Arberry, Vol. 2, p. 260) Know that worldly life is merely a sport and a pastime [involving] worldly show and competition among yourselves, as well as rivalry in wealth and children. It may be compared to showers where the plant life amazes the incredulous: then it withers away and you see it turning yellow; soon it will be just stubble. In the Hereafter there will be both severe torment and forgiveness as well as approval on the part of God. Worldly life means only the enjoyment of illusion. (20) (Irving, pp. 315-316) ْ the plural of كاف, is used The above verse is the only instance in the Quran where the word ٱل ُافار, in its original sense ‘tiller of the soil’ (Ibn Katheer, 2006, Vol. 3, p. 282; Asad, p. 954). Through semantic change, this word has come to be known primarily as reference to ‘unbelievers’, and in 146 both senses the connotation of covering or concealing something exists. In this way, the above verse employs ْٱل ُافارallegorically in its original sense to compare ‘unbelievers’ to ‘tillers’ or ‘husbandmen’, i.e. (those who cover or conceal things). This imagery is preserved in Pickthall's translation, which renders the relevant sense. However, Arberry and Irving`s choices depart from the intended ST meaning and results in a text-focused coherence shift. 4. Conclusion The paper establishes that coherence shifts in Quran translation result in mismatches that seriously affect Quranic meaning potentials. Some mismatches may arise from the cultural distance between ST and TT audience, thus resulting in reader-focused coherence shifts, while others may stem from inaccurate decisions on the translator`s part, thus leading to text-focused coherence shifts. The discussion shows how reader-focused shifts pose a challenge for translators, and we have concluded that paraphrase and footnotes are the most adequate strategies to bridge both partial and complete cultural/ referential gaps in translation. This exegetical strategy helps Quran translators avoid jeopardizing Quranic meanings. Text-focused shifts help identify another problem underlying the transfer of the ST meaning into the TT. This kind of shifts is not directly linked to the translatability of the Quran; it is linked to the translator`s interpretation of the ST and his choices in translation. Most importantly, the discussion of text-focused shifts reveals that serious skewing of meaning does exist in some professional translations of the Quran. Most of these shifts could have been averted if sound linguistic competence in both Arabic and English is secured on the translator’s part, in addition to subsequent consultation of major Quran exegeses, in order to derive and provide the accurate underlying meaning of Quranic text. Pragmalinguistic Failure: Arabic Politeness Formulas in Translation Mohammed Farghal & Ahmed Borini Abstract 147 This paper addresses itself to the translatability of Arabic politeness formulas into English from a pragmalinguistic perspective by examining 15 politeness formulas from Najeeb Mahfouz’s (1959) novel Awlad Haritna. It sheds light on the important role of illocutions, the Cooperative Principle (CP), the Politeness Principle (PP), and the Irony Principle (IP) in understanding the pragmatics of politeness and subsequently translating it appropriately into the TL. The paper argues that deficiency in pragmalinguistic competence usually results in communication breakdown or, at best, distortion of the original message. This claim has been supported by the poor performance of 20 MA translation students on a translation task of the 15 politeness formulas in this study, along with some serious problems with several of these formulas in Stewart’s (1981) translation of Mahfouz’s novel. Finally, the model translations of Arabic politeness formulas have been checked against the intuitions of a group of six native speakers of English. 1. Introduction The present study examines the translatability of politeness formulas from a pragmalinguistic perspective, where things like performatives and illocutions (Austin 1962, Searle 1981, etc.), CP and its maxims (Grice 1975, Levinson 1983, etc.) and the PP and IP of Leech (1983) play an important role. Pragmatic knowledge includes the ability to know the relationship between propositional content (i.e. semantic meaning) and illocutionary force (i.e. pragmatic function) of any politeness formula. Sometimes, the relation between the two is very clear and easy to determine as in the case of English Good luck, which corresponds to Arabic ḥaððan saciidan and Arabic šukran, which corresponds to English Thank you. In other cases, however, it is not possible to relate the propositional content to its pragmatic function. For instance, one may need to learn the convention and conditions of use of a politeness formula like yislam raasak, which literally means “May your head be saved”, but it has an illocutionary force of ‘offering condolences’. Davies (1987:82-83) assumes that “formulas may seem restricted to the kind of speaker who may use them, the kind of addressee to whom they may be used, the medium through which they may be expressed and various aspects of the setting in which they are used.” In particular, a difficulty may arize when the same formula is conventionally used to perform more than one illocutionary force in different situations. Farghal (1995b) contributes a study on the pragmatics of ’inšaallah [If God permitted] in Jordanian Arabic. He demonstrates that this expression has “drifted extensively from its semantic import by acquiring a wide spectrum of illocutions, thus becoming a pragmatically multipurpose expression”. By way of illustration, this formula may function as a threat as can be illustrated in (1) below: 148 1. ’inšaallah btilmis-it-tilvizyoon! touch+you-the-television “I dare you to touch the TV set!” It may also be uttered by an interlocutor to implicate unsatisfaction as in (2) below: 2. xer ’inšaallah! good “What’s the matter with you?” One of the serious obstacles that a translator may face is the ironic usage of some politeness formulas on certain occasions. More specifically, a speaker may use the following politeness formula to insult a person not participating in the interaction as in (3) below: 3. ’allah yirḥam ’abuuh ! God have mercy his father “May God have mercy on his father!” In example (3), the politeness formula has been ironically used to perform a ‘conflictive’ act of insulting. This is what Arndt and Janney (1985:285) call “politeness from an interpersonal view”, i.e., being impolite in polite ways. In other cases, it is not easy to determine the performative verb of the illocutionary act assigned by uttering certain formulas, especially in expressing various kinds of ‘wishing’. To demonstrate this, consider the following examples: 4. cafwan forgiveness (as a reply to thanking) “You’re welcome”. 5. ’aṭaala ’allaahu cumrak prolong God your age “May God prolong your life!” 6. tawakkal cala-llaah 149 depend on-God “Put trust in God.” Applying pragmatic analyses to problems of translation is based on Levinson’s (1983:326) assumption that “despite the probable universality of processes like implicature, there are likely to be significant differences not only in the structure of languages but in their use”. Even in the case of underlying universals, as in the construction of politeness formulas in Brown and Levinson (1987), there are certain peculiarities of every culture, which constitute one of the most serious challenges to translators; hence, there will always be considerable room for pragmalinguistic misunderstanding. Pragmalinguistics is the study of the linguistic ends of pragmatics. It deals with the appropriacy of encoding the pragmatic force of an utterance. Thomas (1983) introduces pragmatic failure as a kind of communication breakdown by second language learners due to the lack of pragmalinguistic competence. It occurs when the pragmatic force mapped onto a linguistic token or structure is systematically different from that normally assigned to it by native speakers. Politeness formulas tend sometimes to show some kind of divergence between the two languages in question. For instance, one very noticeable difference between English and Arabic formulas is the frequency of religious references in Arabic, which the corresponding English polite formulas may lack (Bentahila and Davies 1989:100). By way of illustration, a formula that may functionally correspond to English Goodbye in Arabic is ’allah macak (May God be with you) in a variety of contexts. A problem may also arise when the translator is faced with interjections such as English Oops, which may correspond to Arabic lah in some contexts. Ameka (1992:101) points out that “interjections - those little words, or non-words which can constitute utterances by themselves are another word class found in all languages”. Some interjections can be used as politeness formulas in a highly language-specific manner as can be illustrated in (7) below: 7. dastuur (said by a person entering a house to draw attention to his arrival and/or intrusion). This Arabic attention getter interjection is uttered in this context to attract the attention of people inside a closure, e.g. a house very politely, while in the same context in English, one may 150 not use an interjection for a similar situation. Instead, an English speaker may use formulas such as Hello! Or Anybody here? 2. Research Methodology 2.1 Research Design This study examines 15 Arabic politeness formulas identified by the researchers as involving potential pragmalinguistic difficulty to translators of Arabic texts into English. These formulas are taken from Najeeb Mahfouz’s (1959) novel Awlad Haritna, which was translated by Philip Stewart (1981) into “Children of Gebelawi.” The study was carried out by means of two questionnaires. The first questionnaire, which consisted of 15 underlined Arabic politeness formulas in their original contexts, was given to 20 MA translation students in Jordan. The participants were asked to translate only the underlined formulas and to take enough time to do so (see appendix 1). The second questionnaire, which aimed to consult native speakers of English with regard to suggested equivalents, included the source English contexts (from Stewart’s translation), along with literal translations of the Arabic politeness formulas. In addition, this questionnaire offered six suggested alternatives in each case: three selected from the participants’ responses, the translation used by Stewart, one rendition to demonstrate non-equivalence, and finally a blank alternative for any other appropriate translation (see appendix 2). 2.2 Participants For the purpose of this study, the first 20 MA translation students met by the researchers while distributing the questionnaire were selected. All of them were native speakers of Arabic, holding a BA in English language and literature and then working toward their MA in Translation. Another group of informants consisting of six native speakers of English who had been working for the British Council in Amman for at least two years was selected. These educated people were somewhat familiar with the Arab culture through their daily contact with Jordanians. They were instructed by their director to check the acceptability of the translations relying on both context and their intuitions. 3. Results The translations of 15 Arabic politeness formulas by Stewart and the participants in this research have been analyzed and categorized in light of some theoretical considerations and suggestions of English native speaker consultants. The following table (Table 1) summarizes the results by 151 giving the percentage of inappropriate renditions of each politeness formula and the model translations of the politeness formulas. TABLE 1 No Arabic Famula Percerntage Model Translation 1 yislam famak 40% Well said! 2 ’al’amr minkum wa ’ilaykum 45% It’s all up to you! 3 waḥḥidu-l-laah 45% For God’s sake, be tolerant! 4 tacabak raaḥah 25% Don’t mention it! 5 ’inšaallah 35% I hope so. 6 jabal yaṭlub-il-qurb minak 20% I want/would like to marry your daughter. 7 laa tu’aaxiðna ya bunay 30% Excuse me! 8 faqul calaykum-is-salaam 35% You won’t get much peace! 9 mubarakah ṣadaqah 40% Bless this great friendship! c alayka haaðih-iṣ- 10 c adam mu’axaðih 20% before that thing 11 rabbana yiḥfað maqaamak 25% I’m afraid to say. 12 yaa saatir 20% Hello, anybody here? 13 ’iḥim 75% (cough) 14 bilhana wi ššifa 60% I hope you’ll enjoy it. 15 taslam yadayki-l-jamiilah 40% Bless your beautiful hands. 4. Analysis and Discussion This section addresses itself to the major pragmalinguistic problems in translating the Arabic politeness formulas in the questionnaire by distinguishing different levels of analysis in an attempt to account for the main sources of pragmalinguistic failures in the participants’ renditions and Stewart’s translation. These levels refer to pragmalinguistic failures stemming from illocutionary force, indirectness, irony, euphemism, and interjections. 152 4.1 Illocutionary Force The analysis of the data shows that one source of pragalinguistic failure in translation is that the illocutionary force of most Arabic politeness formulas is not always visible in the surface structure as can be illustrated in (8) below : 8. Atris said, following on from the story about Gebelawi, “There was some good in the world; even Adham was never hungry, not even for a day.” The old woman appeared at the door and said to Atris!“yislam famak.” [May your mouth be saved!] Many student translators could not appropriately assign the intended force to this formula. While the speaker (S) intended to perform an act of complimenting, which is not obvious in the surface structure, 40% of the respondents did not express this force in their renditions, as can be seen in (9) below: 9. May God keep your mouth safe! In this case, the student translator did not conform to the pragmalinguistic norms of English. In fact, such an utterance would likely be perceived as sarcastic or hostile rather than polite. Therefore, on the basis of theoretical considerations and the intuitions of native speakers of English, we may suggest translating this formula into (10): 10. Well said! As can be noted, (10) observes formlaicity and preserves the intended illocutionary force in translation. Thus, it seems pragmatically adequate, as it does not convey the force directly, i.e. “Thank you!”, because this would constitute an under-translation. In some cases, the student translators managed to assign the intended illocutionary force, but they could not encode it appropriately in English as can be demonstrated in (11): 11. Huda turned to Gebel: “Have you anything to say, Gebel?” He looked at the ground and said: “’al’amr minkum wa ’ilaykum ...” [The matter is from you and to you ...] 153 Unfortunately, 45% of the subjects and Stewart mistranslated the above formula by ignoring the intended polite act into: 12. It’s your problem. To maintain the intended force of the formula in (11), the translator may opt for: 13. It’s all up to you. One should note that maintaining the propositional content of many Arabic politeness formulas in English would, in many cases, sacrifice politeness. That is, literal translation may fail to convey the illocutionary force of the polite formula and subsequently distort the message as can be seen in (14): 14. Farhat shouted to the crowds, “Come and hear what people are saying, and see the latest game that’s being played with the honor of Gebel’s people.” Abda shouted wretchedly : “waḥḥidu-l-lah!” [Say Allah is One] 45% of the participants translated the formula in (14) literally by maintaining its semantic import, as can be illustrated in (15): 15. a) Be a monotheist! b) Say that God is one! The TL reader may interpret (15) as utterances said by a preacher who requests his audience to reassert their believing in God; while, according to the context in (14), the intended illocutionary force in the Arabic formula is a polite piece of advice ‘to calm down’. In other words, such renditions manifest an ambivalence of function that makes the TL reader unable to grasp the conveyed meaning of the utterance. Sometimes, translators may fail to pragmalinguistically grasp the implicit performative verb of a politeness formula, which makes it difficult to assign its illocutionary force. By way of illustration, 25% of the student translators in this study failed to understand the implicit performative of the formula in (16): 16. A sweet voice roused him: “Coffee Mr. Qassem.” He turned and saw Badria holding out the cup to him. 154 He took it and said, “Why the trouble, don’t bother yourself for me.” Badria : “ tacabak raḥah.” [Your tiring me is a relief] In this politeness formula, Badria expresses her pleasure to be at the service of the superior hearer. The act performed comes as a reply to thanking. However, some renditions fail pragmalinguistically, as can be illustrated in (17): 17. a) Anything for you, sir! b) Your demands don’t bother me! Pragmatically, Badria intends to communicate something like “Don’t mention it, sir!”, which also goes against Stewart’s translation that overstates the implicit performative (18): 18. It’s pleasure to take trouble for you. Finally, it needs to be emphasized that a difficulty may arise when the same formula is pragmatically used to perform more than one illocutionary force in different situations. For example, Farghal (1995b) shows that the frequent formula ’inšaallah [If God permitted] has drifted extensively from its semantic import by acquiring a wide spectrum of illocutions, thus becoming a pragmatically multipurpose expression. The example in (19) bears witness to this: 19. Balkiti: “You are very cagy, but you’ll soon get used to me and tell me all your secrets.” Gebel: “’inšaallah.” [If God permitted] The analysis shows that 35% of the respondents translated the above formula maintaining its conventional semantic import thus: 20. God willing. Pragmatically, however, this tends to go against the producer’s intentionality and the reader’s acceptability. The speaker (Gebel) intends to perform a polite act of expressing possibility of and hope for complying with Balkiti’s wishes. Hence, it is more appropriate to render this formula as what most of the informant English native speakers of went for in (21): 21. I hope so! 4.2 Indirectness 155 Closely related to the notion of politeness is the notion of indirectness, which is a multifunctional linguistic phenomenon. It has been assumed that politeness is one of the main motivations for people to adopt the strategy of indirectness (Brown and Levinson 1978; Thomas 1993; Farghal and AlManna 2014). An important justification for this view stems from the fact that indirect polite utterances might have more than one plausible interpretation, which enables the speaker to avoid the responsibility of having committed a certain act (Sifiano 1993:70). Viewed in this way, indirect polite utterances may be problematic in translation, which may lead to pragmalinguistic failure. In their renditions, many student translators failed to make a clear distinction between direct and indirect speech acts. Therefore, 20% of them could not properly relay the indirect marriage proposal expressed by Gebel when uttering the politeness formula in (22): 22. Gebel said with an impetuousness (while he was thinking of Sayyeda, Balkiti’s daughter): “Gebel yaṭlub-il-qurb minnak.” [Lit. Gebel wants to be related to you] Balkiti: “You’re the man I’d gladly give my daughter to.” Most inappropriate renditions of the formula in (22) are similar to the overly direct expression in (23): 23. Gebel asked to marry your daughter. Although the student translators attempted to relay the implicated meaning of the utterance into English, this undoubtedly sacrificed the polite message conveyed by the strategy of indirectness. Stewart, by contrast, goes too indirect by translating the formula in (22) into: 24. Gebel wants to be your son-in-law. Interestingly, this may suggest that there are limitations for indirectness in both the SL and TL in accordance with the pragmatic norms of Arabic and English. More subtly, to translate an indirect Arabic formula into English either very directly or too indirectly could relatively distort the message and subsequently delude the recipient regarding politeness. However, we agree with Leech’s (1983:231) argument that some politeness formulas in one language can be more polite than their equivalents in the other only in the sense of 156 relativizing them to pragmalinguistic strategies such as indirectness. Therefore, we respect the intuitions of the English native speakers who went for translating the formula in (22) into: 25. a) I would like to marry your daughter. b) I want to marry your daughter. Further, some Arabic formulas expressing polite requests may look more direct than they really are. By way of illustration, consider (26): 26. Hanoura said, “You are nothing but worthless hashish addicts. If Khonfus came now, you’d fall at his feet.” Then to Rifa’a: “laa tu’aaxiðna ya bunay [Don’t blame us my son]. A hashish smoker has no inhibitions.” An examination of the students’ renditions of the formula in (26) shows that 30% of them tended to translate it more directly and less politely as can be illustrated in (27): 27. a) Be lenient with us! b) Don’t blame us! Not only do the renditions in (27) ignore to convey the appropriate illocutionary force of the Arabic formula, but they are also too direct to express a polite request. In light of the context of this formula, the speaker intended to express some kind of excuse, i.e. something like (28): 28. a) Excuse me! b) Pardon me! 4.3 Irony By way of conversational implicature (Grice 1975), some Arabic politeness formulas are ironically used to perform a conflictive act of insulting. In other words, we can be impolite in polite ways, which is called by Arndt and Janney (1985) “politeness from an interpersonal view”. According to Leech (1983: 142), the IP operates in human interaction alongside the CP and the PP. That is, the speaker may exploit the PP in order to uphold the CP though he/she seems to infringe it superficially. 157 In terms of translation, maintaining the IP in the TL can be considered one of the most serious linguistic obstacles the translator may encounter in relaying Arabic politeness formulas into English, as can be illustrated in (29): 29. Ali: “Rifa’a has friends. They attacked Batikha, and he disappeared from view.” Yasmine frowned and said: “Batikha is not Bayyumi. If you defy Bayyumi, faqul c alaykum-is-salam.” [Say peace be upon you] In their renditions of the formula in (29), 35% of the subjects failed to maintain the IP in English, because they could not differentiate between the implicated ironic use of this formula and its conventional semantic meaning. Note the inappropriate renditions in (30): 30. a) Peace be upon you. b) Say you are in peace. c) You’ll be safe. Clearly, the inappropriateness of the translations in (30) bears witness to the fact that some student translators were not aware of the pragmatic norms of encoding Arabic ironic formulas. More specifically, such literal renditions ignore the fact that, in the case of ironic uses, the speaker’s intended meaning on the one hand and sentence meaning on the other can be seriously at variance. This being the case, the ironic meaning may be maintained in the TL by paraphrasing it as follows: 31. You won’t get much peace! Further, a contextualization of ironic formulas seems necessary to give the translator a chance to grasp the conversational implicature in such cases. Therefore, according to Levinson (1983:126), “ironies require particular background assumptions to rule out the literal interpretations”. Roughly speaking, ironies are pragmatically indirect speech acts that are used to perform various kinds of illocutions. By way of illustration, consider the formula in (32): 32. Abda said anxiously: “Let’s just remember that Khonfus is the ruler of Gebel’s people and that his family’s friendship is an answer to prayer.” Rifa’a: mubarakah calayk haaðihi-ṣ-ṣadaaqah [May this friendship be blessed upon you] 158 Used ironically, the formula in (32) is intended to be offensive and sarcastic. To observe the PP, however, the speaker adopts the strategy of irony to avoid being impolite. The context of (32) reveals that Rifa’a does not intend to perform a congratulatory act as many respondents (40%) and Stewart opted for in their translations, as can be illustrated by the respective renditions in (33): 33. a) Congratulations. b) Congratulations on this friendship. Notably, both translations do not observe the IP in English. That is, the ironic force of the utterance in Arabic is clear from its exaggeration in being overpolite, which is something that can be relayed in English using the same strategy, thus: 34. Bless this great friendship! 4.4 Euphemism One of the main problematic areas in relaying Arabic linguistic politeness in English involves euphemistic formulas, which begin as polite formulas but soon acquire the implicated sense of being impolite. Pragmatically, the producer utters some expressions euphemistically to disguise unpleasant subjects, e.g. taboos, by referring to them by means of apparently inoffensive formulas (Leech 1983:147). As a strategy of politeness, euphemism in Arabic may present much confusion to the translator in his/her attempts to find English equivalents (see Farghal2012; Farghal and AlManna 2014). Based on the data in this study, many student translators could not render appropriately the Arabic euphemistic formulas. By and large, the problem seems to be twofold. First, the relative culture specificity of some linguistically taboo expressions tends to confuse the translator and force him/her to shift from respecting the euphemistic norms in the TL to giving more importance to the intended meaning in the SL formula. By way of illustration, note the Arabic formula in (34): 34. Arafa with boundless confidence (as he introduces a medicine for sexual weakness to the Trustee): “A grain of that in a cup of tea, two hours before “cadam mu’aaxaðih” [without blame], and afterwards either you’ll be pleased with me or you can chase me away with your curse.” 159 Notably, 20% of the student translators who seemed to understand the implicated meaning of the euphemistic formula above (i.e. making love) attempted to apply the Arabic strategy of euphemism literally to English, thus offering renditions like: 35. a) before excuse me! b) before pardon me! c) before without blame! By contrast, many respondents as well as Stewart translated this formula by clarifying the implicated meaning, as can be illustrated in (36): 36. before making love! This explicitness may be due to the translator’s thinking that mentioning ‘making love’ in Arabic is a taboo, whereas it is relatively not in English. We believe, however, that the euphemistic strategy should be maintained in translation, as can be seen in (37): 37. a) before you know what! b) before that thing! Secondly, a difficulty may arise when some euphemistic formulas are parenthetically used right after the mention of what is assumed to be a taboo expression. For example, the Arabic formula in (38) was left out in the English translation by 25% of the student translators as well as Stewart: 38. Kadri the Trustee asked him: “Any news of your wife?” Arafa answered as he sat down beside him: “Stubborn as a mule, rabbana yiḥfað maqaamak.” [May God protect your status!] One should note that the deletion of this euphemistic formula distorts the pragmatics of the original message, as the inclusion of this euphemistic expression is meant to save the superior hearer’s face (Brown and Levinson 1978). Therefore, it is essential to maintain politeness through euphemism by rendering this formula into something like: 39. a) I’m afraid to say! b) Excuse my language! 160 4.5 Interjections Interjections are another significant yet neglected area of linguistics that may interrelate with politeness phenomena. As they are usually highly conventionalized and may be used in different situations to perform polite acts, interjections can be considered one kind of politeness formulas that may create obstacles to translators. Being highly language-specific, interjections, according to Kryk (1992:195), lack exact equivalents across languages. Therefore, the question arising here is how to overcome the difficulty of finding the English equivalent for the Arabic interjection, especially when the latter is used for politeness purposes. One of the main roots of the problem of translating interjections from Arabic into English seems to be the lack of theoretical linguistic research in Arabic regarding this area. Further, a close look at the participants’ translations revealed that many of them could not grasp the intended meaning behind the pragmatic use of some Arabic interjections. By way of illustration, consider the interjection in (40): 40. Qassem went into Qamar’s courtyard to collect the ewe, calling out: ya saatir [O, veiler!] Then he heard the door creaking open, and her voice saying: “Good morning, come in.” The interjection in (40) is meant to be a permission-taking formula to enter the house uttered by the man in an attempt to get the attention of the females inside before he could be let in. Unfortunately, 20% of the subjects translated this formula literally, ignoring the intended implicature, as can be illustrated in (41): 41. Oh, God! The problem here stems from the respondents’ inability to abide by the pragmatic criteria and failure to match linguistic choice with pragmatic functions. Adopting Stalker’s (1989:183) hypothesis that “Pragmatic choices are guided by communicative needs”, we can argue that these respondents’ do not possess the needed communicative competence to select the pragmatic system of using interjections in English as a TL. It should be noted that such an interjection is a complete utterance that performs a polite speech act. In their translations, 35% of the participants adopted a deletion strategy toward this 161 interjection, thus neglecting the context and the performance of a fully-fledged speech act that contributed to the understanding of the entire context of situation. Similarly, other participants (25%) demonstrated a different kind of communication breakdown in their renditions by trying to paraphrase the illocutionary force at the expense of politeness, as can be noted in (42): 42. a) I’m coming! b) I’m coming in! c) I’m here! Although the renditions in (42) may express some degree of politeness as the speaker can get the attention of those inside, this may be insufficient at the politeness level and unsatisfactory on the part of the hearer. The intuition of the native speakers of English stressed formulaicity and approved greeting ‘Hello’, the question form ‘Anybody here?’, or a combination of both as a polite attention getter for the interjection in (40). Only a small number of participants’ (20%) managed to render this interjection of politeness appropriately as: 43. Hello! Anybody here? More opaque to translating are primary interjections as opposed to secondary ones. Ameka (1992) defines a primary interjection as a little word or non-word that can constitute an utterance by itself, such as English Oops! and Ouch! A primary interjection can be used as a coutterance with other units as in (44): 44. Ouch! I have been stung by a bee! Likewise, Arabic possesses a similar class of interjections, which is still virgin ground insofar as research is concerned. By way of illustration, consider the Arabic interjection in (45): 45. Rifa’a found the door ready, open for him. To get their attention to his arrival, he said: “’iḥim” [an attention getter sound similar to ‘coughing’]. Then, she asked him to come in. Many student translations of this formula (60%) ranged from paraphrasing to deletion. However, some renditions (15%) offered inappropriate interjections such as Hum! English native speakers, by contrast, tended to opt for the idea that the speaker in this context should ‘cough’, which can be considered an appropriate equivalent. 162 5. Conclusion The primary concern of this paper has been to investigate the main pragmalinguistic problems that may hinder the process of translating Arabic politeness formulas into English. By examining some problematic areas that may have led to communication breakdown, we have attempted to explain possible sources of pragmalinguistic failure. Most importantly, it has been shown that the lack of linguistic and pragmatic competence will, most often, cause communication breakdown and distortion of the source message. That is, student translators may fall victim to pragmalinguistic failure as they attempt to convey the illocutionary force of politeness formulas. It has been argued that mistranslating Arabic politeness formulas reflects erroneous understanding of the pragmatic norms of English politeness on the part of translators. In some cases, they may ignore the way a certain politeness formula is used in its English context of situation. Further, they may apply the pragmalingistic norms of Arabic to English in an attempt to maintain the meaning of the Arabic formula. An Arab may adopt the strategy of euphemism, for instance, in certain formulas on some occasions, while the same situations may not call for the use of any euphemistic expressions in English. Therefore, translators need to identify pragmatically the way native speakers of English handle politeness and the way they operate the CP and the PP, even in cases of irony, threatening, or insulting. Also, it has been shown that student translators’ unfamiliarity with formulaicity of English has contributed much to the occurrence of inappropriate renditions, because some English formulas have fixed forms in accordance with context. Many student translators could not grasp the distinction between what is formulaic and non-formulaic. Hence, they have either maintained the Arabic formulaic expression or just paraphrased it into English. Finally, the analysis has indicated that student translators often adopt literal translation as a solution when they encounter difficulty in translating any formula. Thus, they try to maintain the semantic import of the Arabic formula at the expense of the intended illocutionary force, which makes their renditions opaque and unintelligible to English readers. What these translators need to do is, in Hatim and Mason’s (1990: 32) terminology, “to situationalise the text by relating it to its environment”. 163 Appendix 1 ترجم ما تحته خط إلى اللغة اإلنجليزية : .1فقال عتريس معلقا ً على ما سم من قصة الجبالوي : كان في الدنيا خير ،حتى أدهم لم يج يوما ً واحداً.وإذا (تمرحنة) العجوز ....تقول موجهة الخطاب إلى عتريس : -يسلم فمك يا عتريس ،كالم كالبرتقال السكري. .2فالتفت هدى (الهانم) إلى جبل متسائلة : ألديك ما تقول يا جبل؟ ...يهمني أن أعرف رأيك.فقال جبل وهو يداري ضيقه بالنظر في األرض : 164 -األمر منكم وإليكم يا سيدتي. .3وصا فرحات وسط الزحام : تعال اسم ما يقال وانظر كيف يعبث العابثون بآل جبل على آخر الزمان.فهتفت عبده جرعا ً :وحِّ دوا هللا والمسامح كريم. .4وأيقظه من تهويمته صوت عذب يقول : القهوة يا معلم قاسم.إلتفت وراءه فرأى "بدرية" تحمل الفنجان ،فتناوله قائالً : لم التعب؟فقالت :تعبك راحة يا سيدي. .5بلقيطي :إنك شديد الحذر ،ولكنك ستأنس إلي سريعا ً وتفضي لي بكل أسرارك. جبل :إن شاء هللا. فضحك جبل في نشوة طفل (وهو يفكر في طلب يد سيدة من البلقيطي) ...ثم قال باندفاع. 6 : يا معلم ،جبل يطلب القرب منك.فابتسمت عينا البلقيطي المحمرتين وقال :إنك الرجل الذي أعهد إليه بابنتي مطمئنا ً .2قال حنورة :ما أنتم إال حشاشون ال خير فيكم ،ولو م ّر أمامكم اآلن خنفس لسجدتم بين يديه .ثم وهو يلتفت إلى رفاعة " :ال تؤاخذنا يا بني" ،فليس على الحشاش حرج. .8قال علي :لرفاعة أصدقاء هزموا بطيخة فاختفى من الحارة. فقالت ياسمينة مقطبة : -بطيخة ال بيومي ،إذا تحديتم بيومي فقل عليكم السالم. .9وقالت عبدة :زارتني زكية زوجة خنفس ...ما أفخم مسكنهم ،المقاعد الوثيرة ،السجاد الفاخر ...ولنذكر فقط أن خنفس سيد آل جبل ،وأن صداقةأهله دعا ٌء مستجاب. 165 ٌ مباركة عليك هذه الصداقة. فقال لها رفاعة في ضجر : .10قال عرفة في ثق ٍة ال ح ّد لها مخاطبا ً المعلم (وهو يعطيه دوا ٌء للضعف الجنسي) : قمحة منه على فنجان شاي قبل "عدم مؤاخذة" بساعتين ،وبعدها إما ترضى عن محسوبك عرفة وإما تطرده من الحارة. .11وسأله قدري الناظر :هل من جديد عن زوجك؟ فأجاب عرفة وهو يتخذ مجلسه إلى جانبه :عنيدة كالبغل ،ربنا يحفظ مقامك. .12دخل حوش قمر ليأخذ النعجة وهو يقول ‘" :يا ساتر" ...وإذا بصرير الباب يُفتح وصوت الست يقول :صبا الخير! .13ووجد رفاعة باب المسكن مفتوحا ً فغمغم قائالً :إحِم! فأذنت له بالدخول َفد َ َخ َل. .14وعادت (سكينة) بلفافة فاعطته إياها وهي تقول : فطيرة بالهنا والشفا.فتلقاها بيديه قائالً :اشكري عني السيدة الكريمة. .15قدّمت عواطف للسنطوري فنجان الشاي .... فرشف السنطوري رشفة وقال :تسلم يديك الجميلة! Appendix 2 Choose the most appropriate Politeness Expression to fill in the blank relying on both context and your intuition as an English native speaker. For each blank, we first provide the Literal Meaning (LM) of the Arabic politeness expression. Secondly, we suggest five translations, so that you can choose the equivalent in English, if there is any. Otherwise, you can fill in (6) with your suggested rendition. 166 1. Atris said, following on from the story about Gebelawi, “There was some good in the world ; even Adham was never hungry, not even for a day. The old woman appeared at the door and said to Atris :ــــــــــــــــــــــــــ (LM) : “May your mouth be saved !” 1. Bless you! 2. You’re right! 3. Well said! 4. Thank you! 5. (No comment) 6. 2. Huda turned to Gebel, “Have you anything to say, Gebel?” He looked at the ground and said : (LM) : “The matter is from you and to you, Madam.” 1. It’s your problem. 2. It’s up to you, Madam! 3. The matter is in your hand. 4. It’s all up to you. 5. (No comment) 6. 3. Farhat shouted to the crowds : “Come and hear what people are saying, and see the latest game that’s being played with the honour of Gebel’s people.” Abda shouted wretchedly : ـــــــــــــــــــــــــــــــــــــــــــ (LM) : “Say Allah is One” 1. For God’s sake! 167 2. Believe in the One God. 3. Take it easy . 4. Stop it, please! 5. (No comment). 6. 4. A sweet voice roused him : “Coffee Mr. Qassem.” He turned and saw Badria holding out the cup to him. He took it and said : “Why the trouble, don’t bother yourself for me.” Bedria : ـــــــــــــــــــــــــــــــــــــــــــ (LM) : “Your tiring me is a relief.” 1. Not at all. 2. Don’t mention it. 3. It’s a comfort to serve you. 4. It’s a pleasure to take trouble for you. 5. (No comment). 6. 5. Balkiti : “You are very cagy, but you’ll soon get used to me and tell me all your secrets.” Gebel : ـــــــــــــــــــــــــــــــــــــــــــ (LM) : “If God permitted” 1. God willing. 2. Perhaps so! 3. I hope so. 4. I will, hopefully. 5. (No comment) 6. 168 6. Gebel said with an impetuousness (while he was thinking of Sayyeda, Balkiti’s daughter) : ـــــــــــــــــــــــــــــــــــــــــــ Balkiti : “You’re the man I’d gladly give my daughter to.” (LM) : “Gebel wants to be a relative of yours.” 1. Gebel wants to be your son-in-law. 2. I want to marry your daughter. 3. I’ll be honoured to marry your daughter. 4. I ask for your daughter’s hand. 5. (No comment) 6. 7. Hanoura said : “You are nothing but worthless hashish addicts. If Khonfus came now, you’d fall at his feet.” Then to Rifa’a : ــــــــــــــــــــــــــــــــــــــــــ. A hashish smoker has no inhibitions. (LM) : “Do not blame us my son” 1. Don’t blame us, my boy. 2. Forgive them son. 3. Please, forgive us. 4. I’m sorry for that. 5. (No comment). 6. 8. Ali : “Rifa’a has friends. They attacked Batikha, and he disapeared from view.” Yasmine frowned and said : “Batikha is not Bayyumi. If you defy Bayyumi, ــــــــــــــــــــــــــــــــــــــ (LM) : “Say peace be upon you” 1. Yo won’t get much peace. 2. You’ll be defeated. 169 3. Peace be upon you. 4. You are losers. 5. (No comment). 6. 9. Abda : “Zakia, Khonfus’ wife, visited me ... They became very rich... Let’s just remember that Khonfus is the ruler and that his family’s friendship is an answer to a prayer.” Rifa’a said : ـــــــــــــــــــــــــــــــــــــــــــ (LM) : “May this friendship be blessed upon you!”. 1. Congratulations on this friendship. 2. Bless this friendship. 3. You should be pleased with this friendship. 4. I should congratulate you. 5. (No comment). 6. 10. Arafa with boundless confidence (as he introduced a medicine for sexual weakness to the Trustee) : “A grain of that in a cup of tea two hours before “ ”ـــــــــــــــــand afterwards, either you’ll be pleased with me or you can chase me away. (LM) : “without blame” 1. making love 2. excuse me. 3. you know. 4. you do it, exuse me. 6. 11. Kadri (the Chief) asked Arafa : “Any news of your wife?” Arafa answered as he sat down beside him : “Stubborn as a mule, ـــــــــــــــــــــــــــــــــــــــــــ.” 170 (LM) : “May God protect your status!” 1. sorry to say that before you. 2. excuse me! 3. God preserve you. 4. thanks! 5. (No comment). 6. 12. Qassem went into Qamar’s courtyard to collect the ewe, calling out: “ ــــــــــــــــــــــــــــــ.” Then he heard the door creaking open, and her voice saying : “Good morning, come in.” (LM) : “O, veiler!” 1. Anybody here? 2. Can I come in? 3. Excuse me ! 4. I’m here ! 5. (No comment) 6. 13. Rifa’a found the door ready, open for him. To get their attention to his arrival, he said : ـــــــــــــــــــــــــــــــــــــــــــ Then, she asked him to come in. (LM) : “’ihim” (an attention getter sound.) 1. Anybody here! 2. Hello! 3. Can I come in ? 4. (He coughed) 171 5. (No comment) 6. 14. Sekina came back with a package which she gave to the guest saying : “A pancake, ــــــــــــــــــــــــــــ.” (LM) : “With happiness and being very well” 1. enjoy it! 2. I hope you’ll enjoy it . 3. I wish you a good appetite. 4. please! 5. (No comment). 6. 15. Awaatif brought Santouri a cup of coffee. Santouri took a sip and said (thanking her) : ـــــــــــــــــــــــــــــــــــــــــــ (LM) : “May your beautiful hands be saved!” 1. Bless your beautiful hands. 2. Thank you! 3. Preserved be your lovely hand. 4. Thanks! 5. (No comment). 6. 172 Media Translation: The Case of the Arabic Newsweek Mohammed Farghal & Mashael Al-Hamly Abstract The present project is a case study of the Arabic Newsweek, namely the translation of two constant features: a main column page and a stars news page. The purpose is to examine the two features from a translation assessment perspective over a period of about four months. The selection of the two genres is informed by the type of discourse employed in each, that is, the evaluative vs. the expository type (Hatim & Mason 1990). The corpus will be subjected to a close examination aiming at detecting perceptible errors (Hickey 2003), which, being lacking in naturalness, draw attention to themselves in the target language. The errors are first divided into general categories such as structural, lexical and discourse features, then they undergo finer classification and subsequently assigned to sub-categories like grammatical usage, collocations, cohesive ties, etc. The analysis of the data shows that the Arabic Newsweek version suffers from a variety of local and global perceptible errors relating to lexical, discoursal and grammatical usage. In terms of genre analysis, these perceptible features present themselves more frequently and seriously in argumentative/evaluative discourse than in expository/nonevaluative discourse. 1. Introduction Traditionally, emphasis has been placed on the translation of authoritative texts such as literary works and religious books (for example, see Nida 1964 and Newmark 1981; 1988). However, with the emergence of translation studies as a legitimate branch of the larger discipline of applied linguistics, the focus has been modified to include a plethora of discourse genres. These include journalistic material (both editorials and news reports), advertisements, and manuals, among other things (for example, see Hatim and Mason 1990 and Hatim 2000). Despite some methodological differences, most authors view translation as transferring the message or meaning of a source text into a target text (Jackobson1959; Nida 1964, 1969, 1977; Kachru 1982; Newmark 1981, 1988; De Waard and Nida 1986; Larson 1984; Farghal 1993a). Notably, the disagreements do not relate to the content but rather to the vehicle (form) by 174 which the message is conveyed, i.e., the type of equivalence which the translator adopts, for example formal (Catford 1965), functional (De Waard and Nida 1986), or ideational (Farghal 1994). Debates on various types of equivalence soon lose their appeal if they are not supplemented by a theory of context. The option for one type of equivalence rather than another is regarded as a correlative of the dynamic nature of contextual factors, including text type, audience and author, among other things. For instance, a creative metaphor should be relayed intact when translating a poem, regardless of the fact that it may be hard to process in the target language. The same metaphor, however, will call for a different treatment when encountered in a newspaper editorial, where the audience rather than the text type comes in as the most relevant contextual factor. As an act of communication, translation or interlingual communication, just like intralingual communication, involves, ipso facto, the presence of two parties: a producer and a receiver. The receiver, who is either a hearer (in the case of interpreting) or a reader (in the case of translating), may make judgments as to the quality of the oral or written input he/she receives. In most real-life situations, these judgments are passed independently of a source text, that is, the target text is evaluated on its own right as an existing text that communicates meaning autonomously in relevant contexts. Some receivers may not even perceive that what they have heard or read had involved translation activity, simply because the act of communication often overrides, and sometimes even omits, the tacitly superimposed translational intervention. However, the communicatively suppressed translational action comes to the forefront in scholarly and/or academic endeavors. As a result, systematic ‘translation quality assessment’ has emerged as an important area in translation studies. Given the nature of translation activity, one may not be surprised by the fact that existing models of translation assessment are sourcetext oriented (Tirkkonen-Condit 1989). This orientation often manifests itself in correspondence between forms (Catford 1965), effects (Nida 1964), or text-types (Reiss 2000), among other things. Taking the source text as a point of reference, Newmark (1988), for example, bases translation quality on the referential and pragmatic accuracy as key parameters of the target text. Similarly, House (1997) offers a model of translation assessment founded on a doublebinding relationship between the target text, on the one hand and the source text and the target culture, on the other. Clearly, priority is given to the source text and its cultural environment within these models (fore more details, see Burkhanov 2003:108-113). 175 Taking the English source text as a point of departure, Abdel-Hafiz (2002) selectively surveys the Arabic Newsweek for translational mismatches. Some of his examples come under imperceptible errors because they sound natural in Arabic despite the fact that they may deviate from the intended meaning in the source text. Some others are wrongly analyzed in the heat of searching for faults with the translation. As for the methodology, it confuses concepts with types of errors, e.g. error analysis sections are presented under rubrics such as formal equivalence, functional equivalence, managing, etc. as well as error types including linkage and grammatical problems. Added to this is the fact that there is no genre and statistical analysis of the data in this study. In a more practical development, Hickey (2003) presents a model of translation assessment based on potential judgments of lay readers. He divides translation features into imperceptible and perceptible. Imperceptible features escape the attention of the reader either because of their high quality in terms of accuracy and/or appropriateness or because of the reader's inability to access the original. By contrast, perceptible features are those which a lay reader “would, or could, consciously notice and suspect of being erroneous because they somehow or other draw attention to themselves” (p.22). Perceptible and imperceptible errors may seem to correspond to ‘overt’ and ‘covert’ errors (House 1922, 1992) but, in fact, they differ from them in that they do not represent a double-binding relationship between the target text and the source text. That is, lay readers can figure out perceptible errors without making reference to the source text. Based on this important distinction, we have opted for Hickey’s rather than House’s dichotomy. 2. Objectives of Study The present study aims to examine media translation into Arabic from a translation assessment perspective. Specifically, the study will examine the Arabic translation of two recurring Newsweek features in 15 issues of the Newsweek in Arabic (published by Kuwait Dar Al-Watan in collaboration with Newsweek international). The two features, a main column page and a stars news page, are meant to represent argumentative and expository discourse, respectively. It should be noted that the Newsweek news material is essentially evaluative, as this publication is a weekly magazine rather than a daily newspaper, where expository news reporting and news evaluation are quite common. This being the case, the closest segment to expository genre we could find is the stars news page. It is hoped that this contrast, though atypical, will bring out important insights into the type and frequency of perceptible translational errors in the two genres. The survey intends to lay hand on perceptible translation features as perceived by the two researchers functioning, so to speak, as lay readers. These perceptible errors are assumed to be 176 the cause of the oft-heard complaint made by the Arabic Newsweek readers about the quality of the Arabic used in terms of accuracy and/or appropriateness. The purpose is multi-faceted. First, the perceptible features will be specified and divided into general categories of errors, including grammatical, lexical, and discoursal problems. Second, each general category will be presented in sub-categories, for example, lexical problems will be sub-divided into errors relating to individual word meaning, collocations, and circumlocutions, among other things. Finally, a statistical analysis will be presented to show the frequency of each general category and subcategory in the two genres. 3. Significance of Study Much work has been done on theory building in translation studies in general (Nida 1964; Catford 1965; Newmark 1981; Larson 1982; De Waard & Nida 1986; Hatim & Mason 1991; Bell 1991; Neubert & Shreve 1992; Baker 1992; and Hatim 2000, among others). Similarly, a large amount of research has been conducted with respect to English/Arabic/English translation. In this vein, some studies have focused on researchers' intuitions, which are based on specific contextualized or de-contextualized material from the two languages (Al-Najjar 1984; Mouakket 1986; Saraireh 1990; and Farghal 1991, 1992, 1993c, 1994, among others). Others have drawn insights from empirical data stemming from student translations (Shunnaq & Farghal, this volume; Khalil 1993; Farghal & Shakir, this volume; Farghal 1995a, 2003; and Hatim & Mason 1997, among others). Some other studies have had recourse to existing translations, especially literary ones as a source of translation criticism (see, for example, El-Yasin 1996; Aziz 1999; Farghal & Naji, this volume). Clearly, there has been a great deal of research about theory-oriented translation models and more practice-oriented translation criticism. The presence of media translation in this voluminous literature does not go beyond anecdotal reference to de-contextualized segments from here and there, despite the fact that media translation is probably the most practicable type of translation in this age - the age of technology, news agencies, internet and globalization. From here comes the serious need for a fully-fledged study on media translation in order to explore its norms and quality. As for the choice of the Newsweek, it need not be explained, for this widely distributed international weekly is one of the more influential magazines in the world. 4. Methodology 4.1. Material The input material for this project consists of two constant full-page features: a main column page 'World View' written regularly by Fred Zakaria, who is a native speaker of American English and a stars news page, which is anonymously compiled. In the first magazine feature, the picture of the columnist is conspicuously displayed in the top right-hand corner in the English original and in the top left-hand corner in the Arabic translated version (See Appendix 1). The second feature clearly displays pictures of the stars on top of their news segments (See Appendix 2). 177 The choice of the first segment is motivated by the fact that editorials of this type usually involve argumentation that necessitates the employment of a language density not required in straightforward news reports. Therefore, column features, being highly evaluative, are potential ground for translation problems, a prediction that has been confirmed by pilot browsing into some copies of the Newsweek in Arabic by the present researchers. As for the selection of the second segment, it is meant to represent expository discourse in the absence of non-evaluative news reports in the Newsweek. The assumption here is that stars news genre, largely being nonevaluative, may involve less language density and subsequently a different picture of perceptible errors. 4.2. Procedure After the acquisition of 15 recent issues of the Arabic Newsweek along with their English originals, the Arabic corpus was carefully examined for perceptible features, which then underwent categorization and statistical analysis. In the course of analyzing the data, reference was made to the English versions only when need had arisen to strengthen arguments and draw further insights. Perceptible features were categorized into three main categories: lexical, discoursal and grammatical. Lexical errors include wrong collocations and idioms, wrong words, and circumlocutions and repetitions. Discourse errors fall within two categories: conjunction errors and phraseology errors. Grammatical errors feature problems with prepositions or pronouns, subject verb agreement, word form, and articles, among others. 4.3. Translators’ Background Two teams of translators are involved in the translation process; the first is the Washington DC team. This team consists of a group of Arabic native-speakers who are not necessarily based in Washington but are in charge of a first draft translation of the Newsweek. The second team is based in Kuwait and consists of translators who are also Arabic native-speakers with a bachelor degree in Arabic language and Literature. This team is responsible for proofreading/editing the first draft version of the Newsweek. Kuwait-based translators are not concerned with correctness of the English-Arabic translation but more with the correctness/grammaticality of the Arabic text. 5. Results The number of perceptible errors in the whole corpus is 266. The World View page claims 214 (80.45%) errors while the Stars News page contains only 52 (19.55%) errors. The maximum number of errors per page in the World View data is 22 against 6 in the Stars News data. A relatively similar ratio obtains regarding the minimum number of errors per page, viz. 7 errors in the World View data against only 1 error in the Stars News data. The overall breakdown of errors shows that discoursal features are the most frequent at 43.8%, lexical features come in second place at 35.7%, and finally grammatical features account for 20.5%. Bar graph 1 shows 178 this distribution. Overall Disribution of Types of Errors in the Corpus 21% 43% Discoursal errors Lexical errors Grammatical errors 36% Bar graph 1. However a different pattern of distribution of features obtains across the two genres, both in terms of typology and frequency. Typologically, discourse features come first in the World View data at 48.13%, while lexical features are the most frequent in the Stars News data at 48.1%. By contrast, discourse errors occupy second place in the Stars News data at 28.86% while lexical errors assume second place in the World View data at 32.7%. As for grammatical features, they come third in both genres at 19.17% in World View data vs. 23% in Stars News data. In terms of frequency in the corpus, features in the three categories are remarkably more frequent in the World View data than in the Stars News data, viz. discourse features account for 38.72% vs. 5.64%, lexical features for 26.3% vs. 9.4%, and grammatical features for 15.41% vs. 4.51, respectively. The distribution of errors across the two genres is shown in Bar graph 2 below. 179 Breakdown of Errors 120 100 80 World View Stars News 60 40 20 0 Lexical errors Discoursal errors Grammatical errors Bar graph 2. Within the category of lexical errors, wrong collocation/idioms score the highest frequency at 48.42%, wrong words come second at 42.11%, and finally lag circumlocution/repetition at 9.47%. Across the two genres, however, wrong words are more frequent than wrong collocations/idioms in the Star News data, viz. wrong words account for 56% while wrong collocations/idioms score 32%. Bar graph 3 below indicates the breakdown of lexical errors across the two genres. Breakdown of Lexical Errors 40 35 30 25 World View Stars News 20 15 10 5 0 Wrong collocations/idioms Wrong word Circumlocution/Repetition Bar graph 3. 180 The breakdown of discourse errors shows two main sub-categories: missing conjunctions and wrong phraseologies. Missing conjunctions are more frequent than wrong phraseologies in the corpus at 69.49% vs. 30.5%, respectively. Across the two genres, the distribution of discourse features is quite different, viz. missing conjunctions in the World View data score 66.95% while they only account for 2.54% in the Stars News data. As for wrong phraseologies, they score 20.34% against 10.17% in the Stars News data. The Bar graph below represents the breakdown of discourse errors across the two genres. Breakdown of Discoursal Errors 80 70 60 50 40 World View Stars News 30 20 10 0 Wrong phraseology Missing conjunction Bar graph 4. As for the data on grammatical errors, they contain seven subcategories including a miscellaneous one. Excluding the miscellaneous subcategory, the most frequent type of error involves prepositions at 30.19% and the least frequent error involves pronouns at 7.55%. Across the two genres, wrong/missing prepositions maintain the highest frequency at 29.27% in the World View data against 33.33% in the Stars News data. Similarly, pronoun errors maintain the lowest frequency at 5.66% in the World View data against 1.89% in the Stars News data. Bar graph 5 below shows the breakdown of grammatical errors across the two genres. Finally, a word should be said about zero versus near-zero translatability as represented by citing culture-sensitive material in source language versus transliteration. The expository text (i.e. the stars news page) featured a noticeable employment of zero-translation when the text included names of T.V. programs (e.g. West Wing, The Simpsons, Rock Talk), music albums (e.g. It had to be you, 8 Mile, Get rich or die trying), and/or movies (Lion King, Hulk, X Men), as well as titles of books (e.g. When the women come out to dance) and/or plays (e.g. Cabaret, Les Misérables). In some instances, Arabic translation or transliteration would accompany the English text, for example روايبةPattern Recognition would be accompanied with the Arabic 181 translation in brackets التعببرف البى األشبكال المميببزة, and كتبابHalloween would be followed by )(الهالوين. Breakdown of Grammatical Errors 12 10 8 6 World View Stars News 4 2 0 Wrong/missing preposition Wrong/missing pronoun Wrong word form Wrong/missing article Wrong agreement Micellaneous Bar graph 5. The argumentative text (i.e. world view page), on the other hand, featured the use of transliteration mainly when need had arisen. Names of magazines, organizations and airplanes are all transliterated; for example طبائرات، مؤسسة السباااك، منظمة ميمولاير، مجلة دي زايت األلمانية،نيوزويك و بريبديتور16أف. Only two instances of the combined use of both the English and Arabic texts were found in the argumentative text under study, for example كتبابThe Gathering Storm was accompanied with ) (العاصفة المتجمعةand كتابEnding the Vietnam war with ))انهاء حرب ايتنام. 6. Analysis and Discussion 6.1. Overview As sampled by the Arabic Newsweek, media translation from English into Arabic suffers from a variety of perceptible errors that seriously affect the quality of translation. Being uninterpretable and/or unnatural, these problematic features draw attention to themselves and, subsequently, cause the target reader to pause reluctantly and think hard in an attempt to reconstruct the linguistic reality in terms of form and/or content. Alternatively, the reader may opt out by abandoning the engagement altogether. On the one hand, problems in form, despite their being local errors (Hendrickson, 1980), may represent a serious source of annoyance during the decoding process. For example, the employment of a wrong preposition, which is a 182 matter of accuracy, or an odd collocation, which is a matter of appropriateness, in a text may create irritation, confusion or bad impressions on the part of the reader. On the other hand, content problems, being global in nature, often create dismay in readers because the reader cannot come to terms with the meaning of a certain part of the text. In this case, the reader’s inability to figure out meaning is usually caused by the translator’s failure to phrase out or code message properly. In many cases, however, form and content problems interact within some segments, thus effecting more confusion and dismay in target readers. Looking globally at the data across the two genres, it can be readily observed that evaluative discourse (represented by the World View page) presents more challenging ground to media translators than non-evaluative discourse (exemplified by the Stars News page), viz. perceptible errors in the World View page are four times more than those in the Stars News page (80.45% vs. 19.55%). This interesting finding may be attributed to the fact that evaluative discourse involves a language density that correlates with setting the tone rather than setting the scene (as in non-evaluative discourse) for the expression of propositions/ideas. Tone setting, it should be noted, necessitates a heavy use of conjunctions, collocations and hypotactic structures in an attempt to persuade the reader and bring about a change in his/her attitude. Scene setting, by contrast, is less demanding linguistically, as it mainly involves mere monitoring of states-ofaffairs; hence the emphasis is on conveying rather than evaluating information (for more information, see Hatim and Mason, 1990; Renkema, 1993) 6.2 Lexical Errors Lexical features come in second place claiming 35.7% of the errors in the entire corpus. In terms of genre, lexical errors are the most frequent in the Stars News data at 48.1%, whereas they score second place in the World View data at 32.7%. The frequency of sub-types, namely wrong collocations/idioms and wrong words, also differs in terms of genre, viz. wrong collocations/idioms come in first place scoring higher than wrong words which come in second place in the World View data, whereas the converse order obtains in the said sub-types in the Stars News data. As has been said above, this discrepancy may have to do with the language density across the two genres, where collocations are expected to play a more important role in argumentation or evaluative discourse (for more details, see Johnstone, 1991) than in expository or non-evaluative discourse. The following subsections will illustrate the types of lexical errors committed in the Arabic Newsweek with elaborate examples from the data. 6.2.1. Wrong Collocations/Idioms Collocations and idiomatic expressions are two important categories of English multi-word units. In fact, they constitute a major element in the make-up of the lexicon and, in effect, a significant part of lexical competence (Alexander, 1978; Aisensadt, 1981; Yario, 1980; Nattinger, 1980, 183 1988; Cowie, 1981, 1988; Strassler, 1983; Benson et al 1986; Baker and MacCarthy, 1987; anf Farghal and Obeidat, 1995, among others). In terms of translation, Shakir and Farghal (this volume) argue that collocations are more communicatively useful than idioms because they are more common in speech and are rarely replaceable by lexical substitutes, e.g. the collocation ‘public support’ is only awkwardly paraphraseable. By contrast, idioms are less common in speech and are readily replaceable by lexical substitutes, e.g. the idiomatic expression ‘bury the hatchet’ is replaceable by the lexical alternate ‘end the dispute’. Consequently, the translator’s failure to operate the Idiom principle in the target language will inevitably result in erroneous collocations but not necessarily erroneous idiomatic expressions, as lexical substitutes are always there as a workable solution in the case of idioms. That is, the Open Choice Principle does work as an alternate in the translation of idioms but it is practically ruled out in translating collocations (for the discussion of the two principles, see Sinclair, 1991). The results of the present study lend further support for the argument that collocations are more problematic than idioms in translation. As a matter of fact, there are only few cases of mistranslated idiomatic expressions in the data, namely 4 errors out of 38 errors in the World View data and none in the Stars News data. This being the case, we have decided to combine the two categories. The most frequent error in collocations is the choice of an inappropriate collocate for a word, which results in an awkward or unnatural combination in the target language. Assuming that translation is done by caring native speakers of the target language, as is the case in the Newsweek, the question that poses itself is: Why would translators sanction the presence of erroneous collocations in their translations? Or more pragmatically: Why would translators not reread their translations and subsequently naturalize such collocational deviations? The answer to the first question has something to do with the translator’s option for literal translation in cases where it does not work. It should be noted that while literalness remains one of the main methods of translation, it is of little use in the area of collocations, as collocations rarely correspond between different languages. This does not mean that some languages are richer in collocations than others, but it simply indicates that words may choose different collocates in different languages. For example, the [± dense] feature of a drink in English is collocationalized in terms of the features [±strong] and [±weak], hence the presence of the collocations ‘strong tea’ and ‘weak tea’, whereas the same feature is collocationalized in terms of the features [± heavy] and [± light] in Arabic, hence the presence of the Arabic collocations شا ثقيلand شا خفيف. The response to the second question may suggest two further indications. Firstly, it may be the case that translators rarely regard their translations as fully-fledged texts away from the source language texts; hence they overlook such deviations in the heat of 184 considering the translation an immediate reflection of the original in terms of content as well as form. In this case, the absence of appropriate translator training may be to blame. Secondly, it is possible that the translator is not adequately competent in the area of collocations in his first language, i.e., she/he does not possess an adequate feel that enables her/him to call up appropriate/natural collocations in the target language in the context of translation activity. This does not mean that some native speakers may lack collocational competence as it is part and parcel of general linguistic competence, but rather that this collocational competence may not be optimally functionalized in the process of translating. Thus, the translator’s collocational proficiency should be regarded as part of her/his translational competence rather than her/his linguistic competence in general. Below are five excerpts from the corpus featuring erroneous collocations as a result of choosing a wrong collocate and one excerpt (No. 6) involving an erroneous idiom. From now onwards, underlined material represents relevant erroneous segments, while bracketed material represents corrections which are not relevant to the discussion at hand (For more examples of erroneous collocations and wrong words, see Appendix 3): [و رغم] المعارك األخيرة بين، و رغم الحكومات السابقة المقسمة و الضعيفة، فرغم أسوأ ركود اقتصادي منذ جيل.1 (November 26, 2002) . فان تركيا قامت بعمليه تحرر اقتصادي واسعة النطاق،تركيا و أقليتها الكردية فبان الضبغط باتجباه، ديسبمبر8 فان لم تصل األمور إلى مرحلة المواجهة بعد وقت قصير من. إن الوقت قصير جدا.2 و لبن. كمبا أن الطقبس سبيجعل العمبل العسبكري أمبرا مسبتحيال حتبى الخريبف القبادم،القيام بعمل عسكري سيتالشى (December 3, 2002) .يكون من السهولة بمكان إعادة لعب هذا الفيلم ثانية [ف] اإلرهابيون اليبوم. كيف يمكن للمرء أن يجابه هذا التوجه؟ األمور ليست حالكة [السواد] كما تبدو للوهلة األولى.3 فمبواردهم هبي مبوارد يرثبى لهبل حبين تقبارن ببالقوى المجتمعبة.يتمتعون بقدر ضئيل مبن أسبباب التفبوق الجوهريبة (December, 10 . و يعني هذا قدر أكبر من العمليات الدولية و التنسيق التحقيقي.للحكومات التي تعمل معا ضدها 2002) حيبث كبانوا، أولئك الذين لم يستط ذووهبم البدف كبانوا يؤخبذون إلبى خبارج القريبة: و جاء في احد تقارير ميمولاير.4 و كانببت األسببالك الكهربائيببة تلصببق بأعضببائهم.يتعرضببون للضببرب و التعببذيب بالتيببار الكهربببائي لسبباعات طويلببة (November 12, 2002).الجنسية بدأت مسئولة العالقات العامة ليزي غروبمان التي عملت علبى صبيانة الضبرر النباجم عبن أشبخاص يطلقبون العنبان.5 (November 5, 2002) ....... ألهوائهم مثل ، فبان بوسبعها تحسبين اتفباق كلنتبون، مسبتخدمة مزيجبا مبن العصبى و الجبزر، و إذا تفاوضت اإلدارة بصورة حسنة.6 (November 14, 2002) . بصورة كبيرة جدا،الذي يعاني الكثير من نقاط الضعف و النقاط الخطرة Other things neutralized (i.e. ignoring bracketed material which relates to other types of errors(, the excerpts in (1) – (5) do not read well because of the erroneous collocations that feature wrong collocates. By way of illustration, the Arabic preposition منبذdoes not collocate with the 185 noun جيبلin (1) and similarly the Arabic noun فبيلمdoes not collocate with the verbal noun لعببin (2). To solve these problems, the translator should have naturalized the two collocations by choosing appropriate collocates such as سنوات طويلبة، عشرات السنين،عقود، etc. for منبذand the verbal noun عبرضfor فبيلم. Moreover, naturalizing the collocation in (2) should involve a further step by abandoning the corresponding Arabic collocation altogether for lack of naturalness in the text and subsequently opting for a more function-oriented Arabic collocation such as كتابة هذا السبيناريو ثانيبةor أداء هبذه المسبرحية ثانيبة. As for the excerpt in (6), it does not read well because of twisting the Arabic idiomatic expression سياسبة العصبى والجبزرةin a way that distorts the flow of discourse and puzzles the target reader. In some cases, the problem occurs because of the translator’s failure to employ a necessary collocate in the target language. The excerpts below illustrate this: عامببا – لمجلببة12 أدلببى صببدام بحببديث صببحافي الفببت – و هببو األول منببذ، ] و [ فببي وقببت سببابق مببن هببذا الشببهر.2 (December 3, 2002) .أسبوعية مصرية و ستجد أنظمة الحكم في الشرق األوسط – و غالبيتهبا أنظمبة غيبر ديموقراطيبة و غيبر مصبلحة [إصبالحية] إن مبن.8 (February 4, 2003) .الصعب عليها بصورة متزايدة البقاء في السلطة إن لم تنفتح و التببي، و الكثيببر منهببا تبببث مببن لببوس انجلببوس، و ينبغببي علببى واشببنطن أن تمببول المحطببات التلفزيونيببة الفضببائية.9 ( December 24, 2002) .أصبحت مصدر معلومات مرموقا بالنسبة إلى اإليرانيين المتعطشين In each of the underlined expressions above, there is a missing collocate that would not escape the attention of the lay Arab reader. Arabic discourse does not tolerate these expressions without post-modifying collocates, viz. the natural expressions are تنفبتح علبى غيرهبا (مبن،الفت للنظبر ) دول العبالمand المتعطشبببن للحريبةin (7), (8) and (9), respectively. It is the job of the translator to employ such natural expressions even when the source text segment does not manifest collocational behavior. To illustrate, consider the English text corresponding to (7) above: 10. Earlier this month Saddam gave a remarkable interview – his first in 12 years - to an Egyptian weekly. As can be seen, the translator managed to relay the English collocation ‘gave an interview’ correctly into أدلبى بحبديث صبحفيbut, unfortunately, failed to properly incorporate the attributive adjective ‘remarkable’ into the Arabic text because she/he was unable to call up the collocation that constitutes the natural habitat for the corresponding Arabic adjective, i.e., الفت للنظر. In other cases, though fewer in the corpus, the collocational problem shows a different directionality. To explain, some Arabic expressions in the data feature superfluous collocates, which are an immediate consequence of rendering English collocations literally. Below are two excerpts showing this kind of problem: 186 11. توماس كريتنسين من أم آي تبي (معهبد ماستشوسبتس،و يقول احد ابرز الخبراء األمريكيين في شئون آسيا الشرقية .(للتكنولوجيا) " إنه ما من شك في أن كوريا الشمالية تراقب مبا نفعلبه فبي العبراق عبن كثبب شبديدFebruary 18, 2003) 12. إن حقيقببة العببالم المعببولم هببي أن مجموعببات صببغيرة فببي العببالم اليببوم قببادرة علببى إحببداث قببدر كبيببر مببن المتاعببب .( واالضطرابDecember 10, 2002) It is clear that the erroneous Arabic expressions عن كثب شديد... تراقبin (11) and العالم المعبولمin (12) are literal renditions of the English collocations ‘watch…..very closely’ and ‘globalized world’, respectively. In other words, the translator was not aware of the constraint that, in contrast with English, the Arabic collocation يراقبب عبن كثببcannot be modified by an intensifier; in fact, the lexeme كثبب, which enjoys no existence in Arabic independently of the above collocation, is inherently emphatic, hence the impossibility of emphasizing it by an intensifier, i.e. the attributive adjective شبديد. Similarly, the translator has fallen victim to the pleonasm العبالم المعبولم instead of the familiar Arabic term العولمة. 6.2.2. Wrong Words Among lexical errors, the use of a wrong word is the second most frequent type of error at 42.11%, which is close to the frequency of wrong of collocations/idioms at 48.42% . It may be argued that this kind of error is a clear indication of a deficiency in the translator’s lexical competence. In many cases, this deficiency manifests itself in the translator’s inability to distinguish cognitive and/or near-synonyms in terms of lexical usage. Following are four excerpts involving lexical usage problems: فصورة الرئيس بوش باللونين األبيض و األسود فقط للحرب على اإلرهاب تفهمت. حمدا ّلل على الوضو األخالقي.13 على ما يبدو [فموقف الرئيس بوش البين تجاه الحرب على اإلرهاب تفهم على ما يبدو] كفا روسيا المعقد م (November 12, 2002) .الشيشان 14. .فالعراق يتطلب استراتيجية (عسكرية) متشددة في حين تتطلب ايران استراتيجية (سياسية) رخوة لقد. ويشر بيتسوان أن "األخبار و الصور من جمي أرجاء العالم بدأت تبث إلى منازل أبناء جنوب شرق آسيا.15 شاهدوا األحداث السياسية في البلقان و الشرق األوسط و الشيشان و هو ما أدى إلى رف وعيهم بشعور إسالمي . (November 5, 2002) ."موحد وعندما كنت أوضح الحاجة الملحة لتجريد صدام من أسلحة الدمار الشامل و الفرصة العظيمة لمساعدة العراقيين.16 كانوا يستمعون بنوع من السخرية و يجادلون في الحقائق و في وجهة نظري حول،على بناء دولة عربية عصرية .الدواف األمريكية و احتمال خلق أوضاع أفضل في العراق As can be noted, the erroneous choice of the underlined words is very clear. The fact that the Russians are the top dog while their subjects, the Chechens, are the underdog renders the use of the Arabic word كفباin (13) inappropriate – the correct cognitive synonym is صبراع. In (14), the 187 employment of the Arabic adjective رخبوةto modify اسبتراتيجية سياسبيةis inappropriate – the correct adjectives in this context are the cognitive and/or near-synonyms مرنبةor متسباهلة. For its part, (14) suffers from an erroneous use of the introductory Arabic verb يشبر, which is a cognitive synonym of the correct introductory verb يببين. Similarly, the last excerpt involves a mistaken use of the verb يجادلونinstead of the correct verb ;يشبككونthe two verbs may be argued to be near-synonyms. These examples, among others, point to a serious deficiency in the translator’s lexical competence, especially when it comes to differentiating between cognitive and/or near-synonyms in terms of correct lexical usage. In other cases, the phonological similarity between two Arabic words may be the source of creating confusion in translators. Once again, the translator’s insensitiveness to phonological form is a clear indication of premature lexical competence. Consider the following excerpts from the corpus: ، ابلغني دبلوماسي آسيوي رفي المستوى أخيرا بأنه قبل هذا الشهر لم يكن يفهم معنى عبارة "حين تستل سيفك.12 (November 18, 2003) ."استخدمه فان، و كما كان نجا اليابان االقتصادي بعد الحرب العالمية الثانية مثاال قويا اقتدت به دول أخرى في شرق آسيا.18 (February 4, 2003) . نجاحا معتدال في العراق يمكن أن يشد من ساعد المصلحين في المنطقة " "هل يمكنك أن توق لي على هذه؟، "أقدمت فتاة نحوي تحمل نسخة من الصورة و طلبت مني قائلة: يقول.19 (January 21, 2003) In (17) above, the translator mistook the Arabic adverbial ‘ أخيبراfinally’ for the correct adverbial ‘ مبؤخراrecently/lately’, which may have been caused by the similarity in phonological form. Similarly, the translator confused the sociological term مصبلحين/‘ مصبلحreformer/reformers’ with the political term إصببالحيين/‘ اصببالحيliberal/liberals’ in (18) , probably due to phonological similarity. Likewise, the translator employed the Arabic verb ‘ أقببدمتcommitted wilfully’ erroneously instead of the correct verb ‘ تقبدمتapproached/came near’ in (19). Once again, phonological similarity may have been a factor. It must be noted that solid lexical knowledge in translational competence should enable translation practitioners to neutralize phonological similarity between Arabic words and subsequently call up the correct lexical item in the process of translating. 6.2.3. Circumlocution Circumlocution is a lexical strategy whereby the translator falls short of coding meaning efficiently and effectively in the target language by describing or exemplifying the target word or phrase as a result of the translator’s being unable to maintain the same degree of lexical specificity between the source and target texts (Brown 2000;Cruse 1986). The resulting 188 rendition, though understandable, is usually awkward and repetitious. In terms of frequency, this type of error accounts for only 9.47% in the data. The two examples below are illustrative: سيؤدي تواصل عدم اليقين و أسعار النفط المرتفعة’ وانخفاض االستثمارات التجارية إلى تحطيم، وفي الداخل.20 (December 3, 2002). االقتصاد . ومن غير الممكن أن تساعد حقيقة أن جثة تشاندرا ليفي عثر عليها في وقت الحق في منتزه دوك كريك.21 (February 11, 2003) As can be observed, the messages in (20) and (21) above are coded awkwardly and repetitiously. To illustrate this point, consider the same messages encoded efficiently and effectively in (22) and (23) below, respectively: . ستؤدي الشكوك المتزايدة وأسعار النفط المرتفعة وانخفاض االستثمارات التجارية إلى تحطيم االقتصاد، وفي الداخل.22 . ومن غير الممكن أن يكون للعثور على جثة تشاندرا ليفي في وقت الحق في منتزه دوك كريك أي دور.23 6.3. Discourse Errors Discourse features are the most frequent in the corpus claiming 43.8% of the total number of errors. They fall within two sub-categories: conjunction errors and phraseology errors. The following sub-sections shed more light on the nature of these types of errors. 6.3.1 Conjunction Errors Arabic discourse is well known for its explicit paratactic nature, with a heavy use of conjunctions (Kaplan, 1966; Johnstone, 1991). The main function of Arabic conjunctions, wa ánd’ and fa ‘so’ in particular, is to make the text stick together and furnish a natural flow of discourse. Consequently, the translator from English, in which discourse relations are primarily asyndetic and hypotactic, into Arabic should be careful enough to cater for this discourse mismatch. That is to say, an Arabic translation should feature more conjunctions than the English original and, conversely, an English translation should include fewer conjunctions than the Arabic original. Apparently, the Newsweek translators are not adequately aware of this rhetorical fact, especially in evaluative discourse where the employment of conjunctions accounts for 69.49% within discourse errors. Following are two illustrative paragraphs: [ و] يمكن لمشاعر. يمكن أن تؤدي الحرب إلى استحثاث[تشجي ] هجوم إرهابي كبير يقوم به صدام أو آخرون.24 [ ف] إذا تفجر العراق من الداخل فان المنطقة بأسرها ستعاني انعدام.المسلمين في كل مكان في العالم أن تلتهب (February 4, 2003) .االستقرار 189 هو البدء في التحدث إلى كوريا الشمالية عن مناف تخفيض التوتر و بدء عالقة جديدة، إن الحل قصير األمد الوحيد.25 و كذلك. ]ف[ الضغط قد ينجح. أو باألحرى أن يتوقف عن فعل شيء. إننا نريد هذا النظام أن يفعل شيئا.معها (January 21, 2003) . [و] ليس أمامنا من خيار سوى أن نجرب النهجين.يمكن للحوافز The Arab reader will immediately notice the restrained flow of Arabic discourse in (22) and (23) above, as a result of the translator’s failure to optimize the use of conjunctions. Also, the lack of an adequate number of conjunctions constrains the impetus of the evaluative tone and consequently weakens the impact of the argument. To shed more light on the significance of this discoursal mismatch between Arabic and English, let us examine the English paragraph corresponding to (23) above (The null sign ø indicates the place where a conjunction and/or a discourse marker is needed in Arabic): 26. The only short-term solution is to start talking to North Korea about the benefits of de-escalating and starting a new relationship. Ø We want this regime to do something – or rather to stop doing something. Ø Pressure might work, so might incentives. Ø We have no option but to try both. (January 20, 2002) Clearly, the asyndetic ordering of the sentences in the English paragraph is appropriate and natural, as it conforms to the discourse norms in English, where it is possible to leave the intersentential relations implicit or suppressed (for more details, see Hatim 2001). In Arabic discourse, by contrast, these suppressed formal features are needed in order to render the discourse natural and appropriate. Note that the translator of (26) as (25) managed to fill the first gap with the emphatic particle إن, which makes up for the lack of a conjunction (e.g. ف cliticized to the pronoun نحنto give فنحنinstead of the combination of the emphatic particle and the clitic )إننا. However, she/he failed to fill in the other two discourse gaps with the conjunctions فand و, respectively and, in effect, offered an asyndetic text which natural Arabic discourse does not tolerate. 6.3.2. Phraseology Errors In terms of frequency, phraseology errors are less common than conjunction errors (30.5% vs. 69.49%, respectively). However, they are definitely more serious than conjunction errors because they jeopardize the conveyance of meaning in its primary existence. That is to say, the translator’s failure to code the intended meaning clearly and/or properly may cause a breakdown in communication, whereas his/her failure to employ conjunctions discoursally may only affect the flow and tone of the text, independently of its basic or primary meaning. Consequently, both types of errors affect the text globally but in different ways. To see how serious phraseology errors are, let us examine the four excerpts below: 190 .22حمدا ّلل على الوضو األخالقي .فصورة الرئيس بوش باللونين األبيض و األسود ،فقط للحرب على اإلرهاب تفهمت على ما يبدو كفا [صراع] روسيا المعقد م الشيشان(November 12, 2002) . .28و رغم أن قادت االقتصاد إلى الهاوية ،إال انه توجد اليوم أقلية قوية في إيران استفادت الثورة منها بشكل كبير من الثورة .فرجال الدين يستخدمون نصيبهم من النفط إلرضاء الكادر الديني و البيرقروطيين الفاسدين والطالب الثوريين و ضباط الجيش(November 24, 2002) . .29لقد فعلت إدارة بوش الكثير لتنفير العالم بأفعالها و لكن أيضا بأسلوبها و نبرتها(January 28, 2003) . .30وفي موقعه االلكتروني (حيث ينشر الحادثة منذ فترة) ,يفترض موبي وجود داف " :ربما ال يحبون (المهاجمون) الرجال النحيفين الصلعان الذين يؤدون أغاني لمطربين آخرين الغطاء؟ أنا متأكد من أن ذلك هو السبب".... )(December 24, 2002 In (27), the reader may wonder how to interpret or make sense of the Arabic text. It is clear that the translator has failed to convey the intended message given in (31) below, which corresponds to the English text in (32). .31حمبدا ّلل علببى الوضبو األخالقببي .فموقبف الببرئيس ببوش الواضببح تجباه الحببرب علبى اإلرهبباب تفهبم علببى مبا يبببدو صراع روسيا المعقد م الشيشان(November 12, 2002) . 32. Thank goodness for moral clarity. President Bush’s black-and-white picture of the war on terror has apparently made sense of Russia’s complicated struggle with the Chechens. The situation in (28) is not any better because the way the intended message is phrased out by the translator befogs the editorialist’s attack on the Mullas in Iran, as can be contrasted with the correct coding of the message in (33) below: .33ورغم أنهم (الماللي) قادوا االقتصاد إلى الهاوية ،إال انه توجد اليبوم أقليبة قويبة فبي إيبران اسبتفادت بشبكل كبيبر مبن الثببورة .فرجببال الببدين يسببتخدمون نصببيبهم مببن الببنفط إلرضبباء الكببادر الببديني و البيروقببراطيين الفاسببدين و الطببالب الثوريين و ضباط الجيش. For its part, the phraseology of (29) is awkward and hard to process by the reader. To see this, compare (29) with the correct coding of the message in (34) below: .34لقد فعلت إدارة بوش الكثير لتنفير العالم منها ،ليس بأفعالها فقط ،و لكن أيضا بأسلوبها و نبرتها. Similarly, the last excerpt in (30) suffers from a phraseology problem. Mobi’s words are meant to spell out the motive behind the attack but the translator’s rendition can hardly convey the intended message. Examine the rendition in (35) as a possible coding of the message: .35وفي موقعبه االلكترونبي (حيبث ينشبر الحادثبة منبذ فتبرة) ،يفتبرض مبوبي وجبود دافب " :ربمبا يكبون البداف أنهبم ال 191 يحببببببون الرجبببببال النحيفبببببين الصبببببلعان البببببذين يبببببؤدون أغببببباني لمطبببببربين آخبببببرين! أنبببببا متأكبببببد أن ذلبببببك هبببببو (December 24, 2002)".....السبب 6.4. Grammatical Errors Grammatical errors are the least frequent in the corpus at 20.5%. Being local in nature, grammatical errors affect the meaning of the text in specific ways and can be readily corrected by the careful reader. The fact that the Newsweek translators make some grammatical mistakes points to a deficiency in their grammatical competence (Canale, 1983). The following subsections throw some light on the types of grammatical errors in the data. 6.4.1. Wrong/Missing Prepositions The most frequent type of grammatical errors is the translator’s failure to employ a correct preposition by either using a wrong preposition or omitting the employment of a preposition altogether. This kind of error accounts for 30.19% among grammatical problems. Following are some examples from the data: واألهم من ذلك أنه أعلن في خطابه أن واشنطن ستنتقل اآلن إلى عملية ترويج "أكثر تعاطيا" لإلصال الديموقراطي.36 ( December 17, 2002) .في المنطقة – وبغض النظر عن وجهات نظر المرء األوليبة لناحيبة حسبم الموضبوع العراقبي – و قبد كنبت أحبد المؤيبدين لبذلك.32 (February 18, 2003) .... فإنني ال أرى كيف يمكن أن تتراج (October 1, . وعدد األسابي اإلضافية في السنة التي يعملها األمريكي المتوسط بالمقارنة مب األوربيبين الغبربيين.38 2002) As can be observed, examples (36) and (37) include wrong prepositions, i.e. لإلصالshould be م اإلصالand لناحيةshould be من ناحية. As for (38), the translator has failed to employ a preposition, i.e., المتوسطshould read في المتوسط. 6.4.2. Agreement Errors Errors in agreement came second in grammatical errors accounting for 15.1%. Below are two illustrative examples: (February 18, .) (يتبين أن "العواقب الوخيمة" هي أن األمم المتحدة تقوم بإرسال عشرات أخرى من المفتشين.39 2003) بالمائة من عدد سكان10 وهو ما يصل إلى، من المدنيين100,000 قتل الروس نحو، فخالل العقد الماضي.40 منهم وحول أكثر من رب تلك الجمهورية الصغيرة إلى ارض200,000 وشتت أكثر من،الجمهورية قبل الحرب .يباب بيئي 192 In (39), the numerical phrase عشبرات أخبرىdoes not agree in gender with the head noun المفتشبين, which is [± human] and [± masculine]. To observe gender agreement, the phrase should read عشبرات آخبرين مبن المفتشبين. As for (40), the lexical subject البروسdoes not agree in number with the subsequent verbs in the text. Apparently, the translator omitted number agreement on the basis of the absence of such agreement between the verb and its lexical subject in قتبل البروس, which is sanctioned in Arabic. However, the number agreement rule must be applied when the agreement is between a verb and a resumptive pronoun; hence the correct forms of the subsequent verbs should be شتتواand حولوا, respectively. 6.4.3. Wrong Word Form Sometimes the translator fails to derive the correct form from a particular word. Errors of this type account for 13.21% among grammatical errors. To illustrate, consider the two examples below: [ف] ال نستطي الذهاب. فأنه ليست لدينا سوى تهديدات قليلة تتسم بالمصداقية اآلن، دعنا نواجه الحقيقة اآلن، و لكن.41 (January 21, 2003) . الن كوريا الجنوبية ستعرض لدمار هائل في هذه العملية،إلى الحرب ضد كوريا الشمالية هببن روزالببين كببارتر و ببباربرا بببوش و بيتببي فببورد و نانسببي ريغببان و، و اجتمعببت خمببس سببيدات أوليببات سببابقات.42 (January 28, 2003) ..……..هيالري كلنتون In (41), the passive verb form ستعرضis inappropriate; the correct verb form is سبتتعرض, which is active in form but passive in meaning (for more details, see Farghal and Al-Shorafat, this volume). As for (42), the error is in the plural form of ‘ سبيدة أولبىFirst Lady’, which erroneously occurs as سبيدات أوليبات, instead of the correct form سبيدات أَول. In both cases, as we can see, it is a matter of deriving the correct word form by the translator, whether it involves derivational or inflectional morphology. 6.4.4. Wrong/Missing Definite Article Errors that involve wrong or missing definite article account for 9.43% among grammatical problems in the data. To illustrate, below are two examples: فقد بدأت الجماعات اإلسالمية مثل الحماس و الجهباد ممارسبة تنفيبذ العمليبات االنتحاريبة ضبد المبدنيين اإلسبرائيليين.43 (December 10, 2002) .العاديين و بأعداد كبيرة و قد أيد توني بلير العمبل األمريكبي رغبم حقيقبة أن غالبيبة داخبل مجلبس وزرائبه و برلمانبه و داخبل ببالده تعبارض.44 (February 18, 2003) .الحرب 193 In (43) the translator wrongly used the Arabic definite article with the proper noun denoting the Palestinian Islamic Movement ‘ حمباسHamas’. In (44), the translator failed to employ the Arabic definite article with the underlined generic noun, that is, the correct form is ‘ الغالبيةthe majority’ rather than ‘ غالبيةmajority’. 6.4.5. Wrong/Missing Pronoun Errors in pronouns score 4 occurrences (7.55% among grammatical problems). Consider the two examples below: و هبي دعبوة رفضبتها تركيبا بغبباء، إنها أوروبا نفسها هي ا لتي دعت تركيا إلى التقدم بطلب العضوية في السببعينات.45 (November 26, 2002) .في ذلك الوقت حيبث ازدهبرت سباللة، جنوب شرق آسياø ]] لكن منطقة واحدة كانت االستثناء لهذه القاعدة الكئيبة [لهذا الواق المر.46 (November 5, 2002) .لطيفة [معتدلة] من اإلسالم على مدى قرون بأتباع يعدون بعشرات الماليين In (45), the cataphoric resumptive pronoun cliticized to the emphatic particle إنredundantly duplicates the anaphoric resumptive pronoun ;هبيhence it should be deleted. For its part, the example in (46) lacks the pronoun هبيin the place of the underlined null symbol. Without this pronoun, the text does not sound Arabic. 6.4.6. Miscellaneous Errors This subcategory contains a mixed bag of grammatical errors which accounts for 24.52% in the data. Among these miscellaneous errors, we find problems relating to tense, negation, punctuation, complementizers, copula, etc. Following are four different types of miscellaneous errors: . عشبية حبرب الخلبيج،1991 هل هذا هو المزاج السائد اليوم؟ كال! انه في الواق وصف لألجواء السبائدة فبي ينباير.42 (January 28, 2002) ببل و حتبى، كان على الواليات المتحدة أن تشحن لقوات األمن الكينية ال أجهزة الكمبيبوتر و الشبيكات فحسبب....، .48 (December 10, 2002) .أجهزة التصوير ببل إنهبا أن تتمثبل فبي أن كوريبا، ليست أن كيم جونغ قبد أصببح فجبأة رجبال شبريرا بصبورة أكببر، و األزمة الحالية.49 (January 14, 2003) . مصنعا للبلوتونيوم، في غضون أشهر،ستصبح . إن اإلسالم لم يتمكن من عقد سالمه م عالم اليوم، و لكنه ال مفر منه، إنه واق غريب.50 (November 5, 2002) 194 In (42), there is a tense problem. The verbal Arabic Adjective السبائدةcan only make a reference to the present time. However, the reference in the text relates to January 1991; hence a finite Arabic verb in the past tense form must be used, viz. التبي سبادت. In (48), the placement of the negative particle الis awkward and hard going. To solve this problem, the negative particle may be placed immediately after the complementizer أن. For its part, the example in (49) features a redundant complementizer – the underlined أن. In order for the text to sound Arabic, the said complementizer must be deleted. Finally, (50) includes a punctuation error – the underlined comma should be replaced with a period. 7. Conclusion and Implications Based on an examination of the Arabic Newsweek, it is clear that professional media translation from English into Arabic suffers from a variety of perceptible problems including lexical, discoursal and grammatical errors. These errors, which draw attention to themselves, often create irritation, and even dismay, in Arab readers. In terms of seriousness, the errors range between rectifiable local problems such as odd collocations and erroneous grammatical features and more subtle, global problems like the translator’s employment of an unnatural flow of discourse and hard-going, or even uninterpretable, phraseologies. Collectively, such perceptible errors affect negatively the overall quality of translation activity, especially when we are dealing professional media translation, which, we believe, cannot be represented by a caliber higher than that of the Arabic Newsweek, given the well-known international status of the source English publication. In terms of genre analysis, evaluative/argumentative discourse proves to be more challenging than non-evaluative/expository discourse as indicated by both the quality and volume of perceptible errors in the corpus. This finding points to the argument that this would be an extremely serious problem in the case of the Newsweek because, in contrast with nonevaluative news reporting which is predominant in daily newspapers, the bulk of the Newsweek material is inherently evaluative. Lagging in time with respect to news coverage, weeklies are built around managing rather than just monitoring news, which renders their task more taxing in creating and subsequently translating discourse. What implications can be drawn from this study? First and foremost, native language 195 competence in Arabic should not be taken for granted in Arab translators. The Arabic Newsweek case shows clearly that the Arabic native speaker translators who are employed to do translation activity at two stages (primary translation in Washington and translation editing in Kuwait) are seriously deficient in Arabic language competence at various levels: discourse, lexis and grammar. This may have to do with the special status of Arabic, where a diglossic situation predominates the linguistic scene in all Arab countries. Arabic diglossia involves a Low Variety (any given local dialect) and a High Variety (Standard Arabic). It can be argued that educated Arabs speak their dialects natively and standard Arabic, so to speak, as a second language which is ‘pseudo-natively’ acquired through formal education and continuous contact with spoken and written material within the Arab culture. Vernacular Arabic, therefore, may occasionally interfere with performance in the standard written variety. The degree of competence achieved in standard Arabic may vary greatly from one individual to another and from one Arab country to another. This depends on the efforts exerted by individuals to excel in the standard variety as well as the degree of attention given to the role of standard Arabic in different Arab countries. This being the case, Arab translators' knowledge of standard Arabic should be carefully checked before entrusting them with translating into Arabic, as is the case with the Arabic Newsweek. A second important implication is that media bodies should ensure the presence of translator training on the part of candidate translators. It is not enough for a candidate to be a university graduate of a foreign languages department. Prospective translators should receive extensive translator training in order to develop their translational competence at the level of structure, lexis and discourse. It should be noted that translational competence may not replicate general language competence in any language pair. Consequently, a candidate who is highly proficient in Arabic and English may prove a failure in translation activity because interlingual communication involves skills that go well beyond general language proficiency. Such skills, we believe, are the output of professional translator training. Finally, but equally important, target language texts, which are the output of translation activity, should be treated as fully-fledged texts in their socio-cultural environment. To accomplish this goal, they should undergo optimal naturalization in terms of discourse and diction, independently of their corresponding source texts at the hands of professional editor translators. At the end of the day, a media translation should not sound as a translation; it should sound as a native text created for a public readership in the target language culture. 196 197 Collocations: An Index of L2 Interlingual Transfer Competence Mohammed Farghal & Abdulla shakir Abstract Collocations constitute a key component in the lexicon of natural language. Translators and/or interpreters should, therefore, possess a working syntagmatic competence alongside their paradigmatic competence. The results of this study show that collocations are a problematic area where student translators/interpreters frequently stumble when working into L2, viz. the percentage of inappropriate renderings of target collocations is 48.2% in translating and 66.8% in interpreting. The higher percentage of inappropriate collocations in interpreting is argued to be relevant to the time factor. The study further looks into the strategies adopted by the student translators/interpreters in their attempts to render target collocations. The strategies of reduction, synonymy, compensation, paraphrasing and calquing are shown to play a significant role. The unnaturalness of the product, the paper argues, comes as a consequence of the translator’s/interpreter’s inability to call up the relevant collocations in the target language. The failure to cope with collocations in the SLT results in mitigating the evaluativeness parameter, thus weakening the line of argumentation in the TLT. 1. Introduction Multi-word units in general and collocations in particular constitute a key component in the lexicon of natural language (Alexander 1978; Aisenstadt 1981; Yorio 1980; Nattinger 1980, 1988; Cowie 1981, 1988; Baker & McCarthy 1987; Sinclair 1987/1991; among others). Conventionally, multi-word-unit taxonomies variously include canonical forms such as idioms, proverbs, clichés, formulaic discourse markers, etc. and non-canonical forms best represented by collocations. Cowie (1981:224) defines a collocation as, "a composite unit which permits the substitutability of items for at least one of its constituent elements". An idiom, on the other hand, is "immutable in the sense that its parts are unproductive in relation to the whole in terms of the normal operational processes of substitution, transposition, expansion, etc". (Mitchell 1971:59). It should be noted that collocations are more communicatively useful than idioms for the simple reason that they are more common in authentic speech and/or materials and are rarely replaceable by lexical substitutes, e.g. 'to harbor evil thoughts' and 'public support' are only awkwardly paraphraseable; whereas, idioms are less common in authentic speech and/or materials and are readily replaceable by lexical substitutes, e.g. 'kicked the bucket' and 'bury the hatchet' are 198 replaceable by 'died' and 'end the dispute', respectively. Yorio (1980:440) rightly notes, "I think it is time we gave up exercises that ask students to 'use the following expressions in a sentence'. Most of the idioms that we ask our students to use are not, in fact, really necessary". The special status of collocations brings them to the fore in the teaching of translation and/or interpretation classes. While highlighting the supremacy of the Idiom Principle over the Open Choice Principle, Sinclair (1987:7) writes, "Most normal text is largely delexicalized, and appears to be formed by exercise of the idiom principle, with occasional switching to open choice principle". Consequently, natural language should be looked upon as fundamentally constituting syntagms rather than paradigms, thus viewed as being highly prefabricated or preconstructed rather than original or creative. The translator and/or the interpreter needs to possess a working syntagmatic competence in the TL, here L2, before being entrusted with the task of translating and/or interpreting into it. This syntagmatic competence embodies a good knowledge of multi-word units, particularly collocations. In point of fact, the translator's and/or the interpreter's internalization of the collocational restrictions in L2 constitutes a fundamental prerequisite for providing idiomatic/natural product. To bring the discussion to focus, we have to distinguish between translation and simultaneous interpreting. The translator has the entire discourse at his disposal when translating, whereas the simultaneous interpreter has access only to the current unit of interpretation. Levi (1967:1172) writes, "The process of translating has the form of a game with complete information - a game in which every succeeding move is influenced by our knowledge of previous decisions and by the situation which resulted from them". The simultaneous interpreter, by contrast, can only access the temporally relevant unit of interpretation which makes his task more challenging and, according to Alexieva (1990:2-3), more creative. Further, the translator has sufficient time to ponder the translation units ranging from the lexical item to the entire discourse, while the interpreter has only a few seconds to render the unit of interpretation. The simultaneous interpreter needs to be both quick-minded and quick-tongued, in order to optimize the use of his syntagmatic repertoire while working under great pressure of time. Lederer (1973), cited in Lederer (1978:328), talks about two psychological processes in interpreting : interpreting through primary meanings, i.e. using the word that comes first to mind as an equivalent to the SL word, and interpreting by the deliberate calling-up of a specific term to match a given word. It can be argued that the first psychological process is more common to interpretation in its first stages and often proves successful, while the second strategy affiliates mostly with professional interpreters, who can operate at a high level of lexical specificity. Due to the foregoing complications in simultaneous interpreting, many scholars argue for capturing sense or message meaning rather than linguistic or language meaning (Pergnier 1978; Seleskovitch 1978; Uhlenbeck 1978, among others). Put differently, form should be played down in favor of content in interpretation, for non-semantic information decays more rapidly than semantic information, that is, the message is retained while the form is forgotten (Le Ny 1978). This argument, however, collapses in the case of collocations where form and content become inseparable in that the interpreter cannot capture a collocation in the SL if he is not capable of calling up its counterpart in the TL. The interpreter's failure usually results in his 199 adoption of strategies of lexical simplification, namely, reduction, synonymy, compensation, paraphrase, and transfer/calquing. Other things being equal, it is assumed that the above strategies are used at a higher rate by the interpreter than by the translator owing to the time factor. The higher the rate of these strategies the less effective the translation and/or the interpretation is, and, consequently, the less natural the intended communication will be. The goal of achieving naturalness in modes preferred by the TL community can be achieved by means of building up a memory bank of collocations and other multi-word units which can be called up, activated, and functionalized as the verbal situation may require. This can be attained mainly through deliberate exposure to certain text types. Exposure to various text types is a basic step in the process of building awareness of how pertinent collocations and other multiword units can serve the purpose of the SL writer/speaker and how they can satisfy the anticipations of the TL receiver. This awareness-building process is a reflection of the process the human memory employs for internalizing and patterning incoming information so that an expectancy mechanism is developed and consequently activated each time a collocant is uttered. This expectancy mechanism enables the language user to activate his memory bank the moment a collocant is uttered so that the pertinent counterpart is called up to fill in a semantic slot in the particular discourse. Immediate activation of the memory bank and recall of relevant collocants are characteristic of LI discourse users. However, this immediacy is not always within the reach of L2 users. This awareness-building process is perhaps most relevant to translator and interpreter training programs. Student interpreters and translators need to be aware of how particular lexical items stored in isolation in memory can be synthesized in patterns where the occurrence of one item predicts the occurrence of the other. Unless such collocational patterns become part of the memory bank of the student interpreter/ translator, communication is certainly doomed to falter. In this context, we can speak of opaque and transparent collocations. Opaqueness and transparency here do not relate to unawareness of the meanings of individual lexical items comprising the collocation under consideration; rather, they relate to the student interpreter's/translator's inability to recognize how one lexical item restricts the co-occurrence of other lexical items that may collocate with it in a particular verbal setting. They involve the failure to recognize the collocational behavior of one lexical item within a given verbal setting as represented in the text type in question. 2. The Present Study 2.1 Subjects The sample is comprised of thirteen senior MA translation students who had completed most of the practical and theoretical translation courses in a translation program. 2.2 Material and Method The study is based on an Arabic newspaper editorial dealing with a topic of political appeal, viz., the Gulf War and the American-Iraqi relations. The editorial argues against the 'American 200 embargo on political dialogue with Iraq'. The text was recorded by one of the present researchers, and the subjects were asked to interpret it simultaneously in a language lab. Four months later, the students were provided with the same text to be translated in writing. Because of the long period between the interpretation session and the translation session, most of the subjects did not recall doing it previously. The students were not allowed to use dictionaries in the translation session. The rationale behind that was to urge the students to rely only on their memory bank in the process of searching for equivalent English collocations. Twenty-four collocations in the editorial were selected, and the subjects' renderings of these were noted both in the interpretation and the translation sessions. The subjects' renderings of the collocations were noted, analyzed, and categorized (see section 3 below). The analysis focused upon the problems the subjects faced in rendering equivalents of the target collocations while interpreting and translating, in addition to the strategies they adopted. A frequency count of each type of rendering was obtained and the equivalents provided were categorized in accordance with the strategies adopted. 2.3 Purposes of the Study The study attempts to answer the following research questions: 1. Does the time factor, inherent in the process of interpreting, influence the quality? and completeness of the target collocations provided in the interpretation session? 2. What strategies do student interpreters adopt in their attempt to overcome the influence of the time factor and fill in semantic gaps when they come across an opaque collocation 3. Do the student interpreters and translators adopt the same strategies in their attempts to compensate for shortage of equivalent target collocations? 3. Analysis and Discussion The results show that the performance of the subjects differed significantly as they switched from interpreting to translating. Whereas they rendered only 33.2% of the target collocations appropriately in interpreting, they rendered more than fifty percent of them (51.8%) appropriately in translating. This is understandable when the following factors are taken into consideration: The time factor: The time lapse allowed in simultaneous interpreting is usually short and the student interpreter has to perform more than one operation during it. He has to comprehend the semantic unit on hand and simultaneously search for equivalents in the TL. Unless equivalents are readily accessible and retrievable from the memory bank, the interpreter will inevitably provide lexical items that may not appropriately do the intended job, but which first come to his/her mind. 2. The insufficient interpreting experience of the students, taking into account the fact that graduate translation programs at Arab universities focus on teaching translation rather than interpretation (usually one or at most two courses are offered in a translation program). 1. 201 Apart from this finding, the student interpreters/translators seem to have resorted to the same cognitive strategies in their attempts to compensate for a shortage of readily accessible and retrievable L2 collocational patterns. The following strategies were adopted in both the interpreting and translating sessions. 3.1 Reduction This applies to cases where either an incomplete or a non-equivalent rendering of the target collocation is provided. This strategy scored a higher frequency (23.8%) in the interpretation session than in the translation session (17.2%). It includes three sub-strategies : generalization, deletion, and message abandonment (see Tables 1 & 2). Strategy Generalization Deletion Message Abandonment Percentage 8.7% 11.6% 3.5% Table 1. A Breakdown of the responses provided via reduction strategies in simultaneous interpreting. Strategy Percentage Generalization 9.2% Deletion 4.3% Message Abandonment 3.7% Table 2. A Breakdown of the Responses Provided via Reduction Strategies in Translation. 3.1.1 Generalization This seems to have been a persistent strategy in the interpretation session based on semantic and schematic considerations where context of situation and assumptions of shared knowledge tend to play a decisive role in the process of searching for target equivalents. Observe, for instance, the following generalized renderings of the target collocations provided in the interpretation session: (1) Target Collocation Received Rendition 1a military build-up American Forces 202 1b terrible consequences destructive war 1c economic embargo another embargo 1d foreign nationals hostages in Iraq 1e historic initiative peace initiative A look at the above examples reveals that the student interpreters were drawing more upon schematic knowledge and assumptions of shared knowledge of the intended recipients than on collocational patterns or pertinent stored lexical items. (1.a), for example, shows that the context of situation seems to activate responses commensurate with what the student interpreter assumes to be knowledge shared with the recipients and which, therefore, cognitively incorporates the intended semantic unit. It is worth noting here that the political position of most of the Jordanians during the Gulf War was anti-American-led coalition. This explains the motives behind the collocations provided. In (1.a) some student interpreters seem to have assumed that 'military buildup' in the Gulf was in other words driven and engineered by the US; hence 'American Forces' was provided as target collocation. In (1.c), 'economic embargo' is rendered as 'another embargo' - a response which derives from the student interpreters' awareness of the series of sanctions taken by the UN Security Council against Iraq before the war. The same can justifiably apply to (1.d) where 'foreign nationals' is rendered as 'hostages in Iraq', thus a cognitive addition based on existing knowledge of the political milieu replaces the target collocation. Had the recipients (the researchers, in this case) been alien to the political stance of the student interpreters, such responses could have been considered void of content if judged on a right-wrong scale. Another type of generalization is illustrated in the following examples: (2) Target Collocation Received Rendition 2a intimidation and threatening threats 2b military build-up armies 2c peaceful settlement settlement In these examples the student interpreter seems to have been considering the input of the message by picking up the key item acting as a super-ordinate element when compared with the other element(s) supposed to have co-occurred with it. It seems that she/he has assumed that since /tahdiid/ (threatening) and /tarhiib/ (intimidation) are cognitive synonyms, and due to the 203 time factor involved in interpreting, 'threatening' could be an adequate key item that incorporates both meanings. 3.1.2 Deletion This refers to cases where one or more elements of the intended/target collocation is/are deleted, not because the one provided semantically shares the content of the ones deleted, but most probably because of the student interpreters' failure to provide a complete one. A look at Tables (1) and (2) shows that this strategy was operative more in the interpretation session (11.6%) than in the translation session (4.3%). Observe the examples provided in (3) below: (3) Target Collocation Received Rendition 3a to direct severe criticism criticizing 3b apparent objective objective/goal 3c historic initiative initiative 3d pre-meditated aggressive aggressive plans intentions 3e domestic and foreign policy policy As can be noted, the absence of the target collocants in the received responses affects the intended force of the message and empties it of its emotive impact. Thus 'apparent objective' has been neutralized when rendered merely as 'objective', and 'historic initiative' seriously suffers when reduced to 'initiative'. The same applies to 'pre-meditated aggressive intentions' when reduced to 'aggressive plans' since the focus of the source text seems to be placed upon the element of 'pre-meditation' for inflecting military and economic damage on Iraq. 3.1.3 Message Abandonment This strategy has been resorted to in very limited cases, i.e. when the student interpreter/translator either failed to grasp the semantic unit in question, or when he/she failed to provide any of its constituent elements. These types of failure were not persistent in either the interpretation or the translation session (3.5% and 3.7%, respectively). 3.2 Synonymy Strategy Percentage Reduction 23.80% Synonymy 18.00% Compensation 12.20% Paraphrase 10.32% 204 Transfer/Calquing 2.20% Table 3. Percentage of responses rendered via different strategies in simultaneous interpreting (arranged in order of frequency). Strategy Percentage Synonymy 18.20% Reduction 17.20% Compensation 12.20% Paraphrase 4.30% Transfer/Calquing 0.03% Table 4. Percentage of responses rendered via different strategies in Translation (in order of frequency) Tables (3) and (4) show that searching for lexical items synonymous with the target ones was a major strategy in the processes of interpreting and translating, viz. it scored a frequency of (18%) in the former and (18.2%) in the latter. It has been noted that the deviant collocants in almost all cases in interpretation and translation are the modifying elements. Nouns, on the other hand, are appropriately rendered in the target language. This tends to support our argument in sections (1) and (2) above that the student interpreters' failure in most instances is likely to have stemmed from a shortage of collocational patterns pertinent to political discourse. It also seems that insufficient exposure to L2 political discourse is behind the student interpreters'/ translators' inability to recognize the semantic narrowing which entails the choice of a particular modifier of a noun rather than the ones provided. The examples in (4) below illustrate the above observations: (4) Target Collocation Received Rendition 4a terrible consequences bad consequences 4b pre-meditated aggressive aggressive intentions intentions 4c public support popular support/ common support 205 4d apparent objective explicit objective/ surface objective/ aiming publicly at 4e comprehensive peace overwhelming peace/ overall peace/ general peace The examples above show that the student interpreters'/translators' reliance on the open choice principle (as opposed to the idiom principle) for rendering the target collocations indicates that L2 lexical items have not yet been patterned as independent meaningful units retrievable in response to the discourse type in question. This state of unpatterned L2 lexical items seems to be behind the student interpreters'/translators' reliance on searching for synonyms in their attempts to maintain communication, thus exploiting all available means to that end. This also indicates that the students, perhaps due to insufficient exposure to comparable L2 discourse, have not yet developed a sensitivity to observe specialization brought about by the co-occurrence of particular lexical items in a given verbal situation. 3.3 Compensation The responses provided here represent all, or almost all, the items inappropriately combined to produce the target collocations. The examples in (5) below illustrate how the meanings of the target collocations are distorted as a result of the students' opting for lexical items whose co-occurrence does not meet the anticipations of the L2 recipients of the discourse. (5) Target Collocation Received Rendition 5a military build-up military gathering 5b terrible consequences the expected disaster 5c social welfare social entertainment 5d economic embargo economic supply 5e to thwart such plans to destroy such trends As the figures in Tables (3) and (4) show, the student interpreters/ translators seem to have tried to avoid paraphrasing opaque collocations except in very limited instances : only 4.3% of the collocations were rendered in paraphrased form in the translation session, while paraphrases scored 10.32% in the interpretation session. Paraphrasing, therefore, can be viewed as an attempt to maintain communication (Farghal and Obeidat 1995). However, heavy reliance on such a strategy will certainly impair the naturalness of communication and discourse flow. The instances provided in (6) below reveal how the target collocations are replaced by lavish and incompact paraphrases which, 206 predictably, do not fit the time lapse in simultaneous interpreting, and at the same time do not meet the target audience's anticipations of compact and idiomatic expression: (6) 6a Target Collocation harbor ill will Received Rendition wish to do no good/ hide evil date 6b comprehensive peace settling all disputes/ solving all problems and issues 6c impose economic ban trading with/use force to embargo stop economic relations 6d thwart/abort such plans hinder by force the implementation of such plans/stop from carrying out plans 3.5 Transfer/Calquing It is clear from the figures in Tables (3) and (4) that the student interpreters'/translators' reliance on calquing has been kept to a minimum and that the frequency of occurrence of transferred responses is justifiably negligible - a situation which indicates that LI interference in the processes of interpreting and/or translation was not effectively operative. Below are some examples of transfer: (7) 7a Target Collocation probe the Iraqi position Received Rendition check the pulse of Iraq 7b intimidation and threatening fearing and threatening 7c apparent goal superficial goal 7d thwart/abort such plans corrupt such plans 4. Conclusion Interpretation/translation teaching can greatly profit from the study of collocations. In particular, the idiom principle should be brought to the fore because the open choice principle can only be appropriately utilized when it is functionally coupled with an effective use of the idiom principle. Put differently, the interpreters'/translators' knowledge of paradigmatic relations such 207 as synonymy, antonymy, hyponymy, etc., should be supplemented with a competent syntagmatic repertoire that functionalizes and idiomaticizes their phraseology. The oftenheard complaint that a translation does not sound English, despite its having good wording and syntax, undoubtedly ensues from a deficiency in the area of collocations in particular and multiword units in general. Prospective interpreters/translators should, therefore, undergo extensive training that has collocations as a major concern. Not only naturalness is crippled by the erroneous rendering of collocations, but also the evaluativeness parameter (Farghal 1991, 2012). A careful analysis of the text (the editorial from which the collocations are extracted) reveals that the select collocations mark the progression of the line of argumentation and signal its emotive and evaluative trend. Collocations like "to direct severe criticism", "(its) apparent goal", "to show its good will", "intimidation and threatening", "harbor ill will", "historic initiative", "terrible consequences", "pre-meditated aggressive intentions", etc. are not to be interpreted/translated in isolation from the content and purposes of the text in question, for they reflect the thread of argument the author attempts to maintain. Misinterpreting or mistranslating the target collocations will certainly result in distorting the argument by neutralizing it when it is not meant to be neutral. The subjects' erroneous rendering of 66.8% and 48.2% of the target collocations in interpreting and translating, respectively, is a matter of concern. One of its major implications is that the student interpreters'/ translators' repertoire of collocations in the target language is still in need of much enrichment. To achieve this end, translation programs need to: a. allot sufficient time for lab practice where authentic recorded texts (lectures, seminars, speeches, etc.) are accessible to student interpreters; b. encourage student interpreters to build up their text-type dictionaries in which they record the collocations relevant to the type of text and argument they are dealing with. Opaque collocations can be disambiguated via consultation with the interpreter/ translator trainer, or by consulting specialized dictionaries; c. incorporate simulation sessions where one trainee is selected to act the role of an interpreter in, for instance, a press conference. The trainer can take the role of the interviewee, while a second trainee will act the role of the press reporter. The rest of the trainees will be the audience and their job will also include noting down the weaknesses of the interpretation provided. Immediate feedback from the trainer and the trainees can be very helpful. 208 Explicitation vs. Implicitation: Discourse Markers in Translation Mohammed Farghal & Abdullah Samateh Abstract The present paper aims to examine Blum-Kulka's (1986) claim that cases of explicitation in the target text (TT) correspond to cases of implicitation in the source text (ST). A corpus of three discourse markers (DMs) in an Arabic translation is examined against the DMs in the English ST. The findings show that there are three types of correspondence in DMs: explicitation to explicitation, explicitation to implicitation, and explicitation to zero equivalent. The paper concludes that the syndetic nature of Arabic discourse, unlike the asyndetic nature of its English counterpart, accounts for the presence of several cases of DMs which do not correspond to implicit DMs in the ST and whose sole function is to improvise smooth and cohesive discourse. 1. Introduction Discourse markers (DMs) play a considerable role in communication, "they impose a relationship between discourse segments they introduce and the immediately prior discourse segments," (Farhan and Fannoush 2005: 5), and thus achieve greater transparency as they "knit the discourse together […] and orient the reader," (Pym 2005: 33). Baker (1992: 190) also attributes similar values to their presence in discourse; she writes: Unlike reference, substitution, and ellipsis, the use of conjunction does not instruct the reader to supply missing information either by looking for it elsewhere in the text or by filling structural slots. Instead, conjunction signals the way the writer wants the reader to relate what is about to be said to what has been said before. 209 Put simply, they are cohesive devices that bind the textual elements and signal logical relationships within the text to ensure a natural and smooth flow of discourse. The translator, therefore, should be aware of their functions and usage, for the absence of such awareness could lead to altering the meaning potentials of translations. Arabic discourse, being syndetic, abounds in DMs and makes frequent use of them (Baker 1992; Hatim 1997b; Farghal and Al-Hamly, this volume; AlKhfaji 2011; Tahaineg and Tafish 2011; Farghal 2012, among others). Nevertheless, the existing literature tends to show that most Medieval Arabic grammarians devote much effort and space to the parsing aspect of DMs and pay scant attention to their textual functions (Abdel Hameed, 1965; Anees, 1966; Ansari, 1979; Hamad and Zu'bi, 1984; Fareh, 1998). That is, they engage themselves in classifying the particles into categories as per their syntactic properties, including ’adawaat al-rabt 'connective particles' أدوات الربط, ’adawaat al-‘atf 'conjunctions of sequence' أدوات العطف, and ’adawaat al-bayaan 'explicative apposition' أدوات البيان, but they largely overlook their semantic and pragmatic aspects. They perceive these DMs as cohesive devices whose sole function is to coordinate units in discourse (Al-Hmouz 2001). Thus, the semantic and pragmatic aspects appear to be played down despite their significance in facilitating information processing for the receiver. By contrast, the last few decades are marked by particular interest in the study of DMs, thus taking them beyond the borders of grammar and allowing their semantic and pragmatic dimensions to be accorded due attention as well (Al-Hmouz, 2001; Karin, 2005; Johnston,1990; Al-Batal, 1990; Kammensjo, 1993; Hamdan and Fareh, 1999; Muzni, 1983; Zajjaji, 1984; Crew, 1990; de Beaugrande and Dressler, 1981; Halliday and Hassan, 1976; McCarthy, 1991; Schiffring, 1987; Cantarino, 1974, 1975, 1976; Al-Jubouri, 1987; Williams, 1989; Holes, 1995; Stubs, 1983; Saeed & Fareh, 2006, etc). In particular, the Arabic conjunction wa 'and' و, has been subjected to elaborate analyses in order to reveal its multi-faceted functions in discourse (Abdel Hameed, 1965; Kamal, 1971; Zajjaji, 1984; Hamad & Zu'bi, 1984; etc.). Al-Jubouri and Knowles (1988) indicate that wa and fa 'so' are found to be the most recurring DMs in Arabic discourse. Cross-linguistically, more research is needed, especially in contrastive studies involving Arabic and English. The underlying reasons why DMs are cross-linguistically understudied is probably because their analysis in English is mainly speech-oriented since they are mainly approached from the perspectives of dysfluencies and language acquisition, etc. (Howell, et al., 1999; Bell, et al., 2009; Chung and Nebaker, 2007; Dworzynski et al., 2004, etc.), while in Arabic the focus is on the structure of written discourse (Tahaineg and Tafish 2011). Lately, however, studies conducted cross-linguistically have yielded significant results, where the functional polysemy of DMs is highlighted. Examples of researchers who have explored the multiple functions a single DM can perform in various contexts include (Cantarino, 1974; Fareh, 1998; Illyyan, 1990; Farhan & Fannoush, 2005; Tahaineg & Tafish, 2011, etc.). With reference to Arabic in particular, it has been reported that the Arabic wa is identified with multiple discoursal functions, namely, the resumptive, additive, alternative, comitative, adversative, and circumstantial functions. Likewise, the Arabic өumma 'then' ثمsignals meanings of sequence with a span of time, 210 sequence with immediacy or with a short span of time, resumption of discourse, adversative relationships, and consequential function. The Arabic fa has also been shown to encode several syntactic and semantic functions, namely, the sequential, explanatory, causal, resultative/consequential, resumptive, and adversative. It is worth noting that DMs can be single words like the ones cited above or phrases, e.g. bixtisaar 'in short' باختصار, fiimaa ’adaa 'except for' فيما عدا, muqaaranatan bi 'in comparison with' مقارنة ﺒetc., which fall beyond the scope of this study. Given its syndetic nature, Arabic discourse employs DMs lavishly; their recurrence brings about a high degree of textual cohesion and coherence in Arabic writing. By contrast, English can be asyndetic to a large extent, where non-finite phrases and punctuation may signal suppressed logical relations. Consequently, an Arabic translation is expected to outrank its English source in the use of these elements, prompting cases of DMs with zero source equivalents. For example, the cause-result relationship between 'Arabic syndetic nature' and 'the lavish employment of DMs' in the first sentence in this paragraph (bold-typed) is suppressed in English, whereas an Arabic translation would make it explicit by the use of a DM like bimaa ’anna or bisababi 'because'. In addition, the English semicolon separating the two main parts of the sentence calls for the use of the Arabic fa as a DM, in order to signal commentative material as well as naturalize and smooth the flow of discourse. Moreover, the DM wa would be required at the beginning of the sentence as a default DM to maintain a natural flow of discourse. In this way, we would have three explicit DMs in the Arabic sentence corresponding to zero DMs in the English sentence. Semantically, however, the logic of the sentence is based on an implicit causeresult relationship in English, which corresponds to an explicit cause-result counterpart in Arabic, and an implicit commentative relation in English signaled by punctuation, which corresponds to an explicit commentative relation signaled by fa in Arabic. 2. Purpose of Study The purpose of this study is to examine Blum-Kulka’s (1986) claim within her oft-cited Explicitation Hypothesis that instances of explicitation in the target text (TT) must correspond to instances of implicitation in the source text (ST). Looking at English and Spanish, Saldanha (2008) finds the claim invalid; she argues that there are instances of explicitation that are not necessarily instigated by implicitation in the ST. Building on the findings of Saldanha, the current study assumes that the same phenomenon might obtain between English and Arabic, as well. 3. Study Material Instances of explicitation vs. implicitation of DMs are extracted from the first five chapters of the English novel (The Woman in White by Wilkie Collins 1960/2010, Penguin Books) and its Arabic translation ðaatu al-ridaa’i al-’abyad (Beirut, Dar Al-Bihar 2003). The choice of the novel is solely motivated by the fact that it is a representative sample of professional fiction translation which is commissioned by a reputable publisher like Dar Al-Bihar. The data consists of 55 examples featuring fa, ’ið إذand bitaalii بالتاليemployed to introduce causal, resultative, adversative, resumptive, explanatory, and adverbial clauses. It should be noted that the Arabic DM wa 'and' 211 has been excluded from the study data, albeit it is the most common in Arabic, because it is usually used as a default conjunction which practically carries no or little semantic content when it comes to marking logical relationships. In fact, this DM is largely considered too light to carry semantic content independently of other more semantically oriented DMs. Hence, it is mainly used to enhance rather than replace such markers, e.g. it often occurs with bittaalii 'therefore' in wa-bittaalii 'and therefore' to consolidate the logical relation and smooth the flow of discourse. The analysis of the three study Arabic DMs will determine whether or not they always have corresponding elements in the ST, and if not, whether this could be attributed to the fact that Arabic discourse is overwhelmingly syndetic while English discourse is largely asyndetic. 4. Data Analysis and Discussion 4.1 The DM fa The data reveals that fa is the most frequently used of the three DMs under investigation, viz. 20 instances, making up 36% of the corpus. According to Al-Afghani (1970), Arabic fa can signal both sequential and additive functions. Medieval Bin Hishaam Al-Ansaari (2002/ d.761h) goes even further and argues that fa performs six different functions, namely, a coordinative, sequential, resultative/consequential, causal, and combinatory with sawfa in ' فسوفwill' as well as with ’in 'if' in فإن. The analysis shows that out of the 20 instances fa corresponds to implicitation in the ST in 11 cases (55%), to explicitation in 3 (15%), and to zero equivalent in 6 (30%). The Arabic extract in example (1) below involves many instances of fa which perform different functions, namely, the adversative, explanatory, resultative, causal, and resumptive functions: (1) "Except that we are both orphans, we are in every respect as unlike each other as possible. My father was a poor man, and Miss Farlie's father was a rich man. I have nothing and she has a fortune. […] I can never claim my release from my engagement, she went on. "Whatever way it ends it must end wretchedly for me." […] I have been made all the readier to comply with this request by a passage at the end of his letter, which has almost alarmed me. After mentioning that he has neither seen nor heard anything of Ann Catherick, he suddenly breaks off […] ، فبينما كان والدي رجال فقيرا. فنحن نختلف بعضنا عن بعض في كل األمور األخرى اختالفا كبيرا،"وفيما عدا أ ّ ّننا يتيمان " ،] ال يمكنني أبدا اإلقدام على فسخ الخطوبة...[ . وهي تملك ثروة كبيرة، فأنا فقيرة معدمة.كان والد اآلنسة فارلي رجال ثريا ". فكيدما انتهت هذه المسألة ستكون نهايتها حزينة جدا بالنسبة إل ّي،تابعت قائلة . وبخاصة بعد قراءتي للمقط األخير من رسالته الذي أثار الذعر والخوف في نفسي،]صمّمت على تنفيذ ما طلبه مني فورا...[ ]...[ توقّف فجأة عن التحدث في هذا الموضوع،فبع ما ذكر أنه لم يشاهد أو يسم أي شيء عن آن كاثريك 212 It is clear that fa in the first mention in فنحنcorresponds to neither explicitation nor implicitation in the ST. Also, it can be observed that the adversative relationship between the two clauses is, in fact, signaled in the ST as well as the TT by the DM except that and فيما عداrespectively in the first clause, which expresses the speaker's adversative attitude in the second one. The fa is supposed to enhance that adversative relationship as well as smooth the flow of discourse. Yet, the insertion of fa makes the translation sound redundant, due to its immediate recurrence in the following sentence. However, the second instance of fa is needed to enhance the contrast DM bainamaa 'while' by smoothing the flow of discourse and improvising cohesion, whereas the first one may be considered superfluous, for the adversative relation is signaled by fiimaa ’adaa 'except that' rather than by the fa. Note that the contrast in the source text is signaled by the multifunctional 'and', but the translator opts for using a stronger contrast marker because he probably feels that wa is not semantically strong enough to carry the contrast, it being commonly used as a default DM in Arabic. In fact, there are three instances in the above excerpt where wa is employed as a default DM whose sole function is to render the text cohesive rather than mark logical relations, viz. wa-fiimaa ‘adaa 'and except that', wa-hya 'and she', and wabixaasatin 'and especially'. Likewise, the resumptive fa in فبعدdoes not correspond to implicitation in the source text. The preceding paragraph is mainly about a letter from Mr. Hartright in which he prevails on Miss Halcombe to get him an employment outside London. He attributes his request to the fact that he has been watched and followed by some strange men ever since he returned to London. Consequently, his life might be in danger. That first paragraph concludes with the above sentence I have been made all the readier to comply with this request by a passage at the end of his letter, which has almost alarmed me. However, there seems to be a shift of topic in the following paragraph, bearing in mind that it presents new information which revolves around Miss Catherick rather than Mr. Hartright, but still within the context of the same discourse. In other words, the following paragraph is related pragmatically to the preceding one, and the Arabic fa is employed to introduce that relationship and, as a result, orient the reader. Thus, the fa is employed to naturalize the discourse and render it cohesive. By contrast, it could be noted that the use of fa in فأناand فكيفماcorrespond to and and the suppressed thus in the ST respectively. For example, the fa in فأناintroduces the clause that describes Miss Halcombe's poverty as the result of her father being a poor man, whereas Miss Farlie's wealth is the result of her father being a rich man. It, therefore, corresponds to an explicit resultative marker and in the ST. Similarly, the fa in فكيفماintroduces a cause-effect relationship between the first and the second clause. In the first sentence, Miss Farlie states that she could not afford to renege on her engagement to Mr. Percival and in the second one she spells out the cause or reason; it would make her life a misery. So, fa corresponds to an implicit causal marker because, thus, etc. in the ST and it is brought to the surface in the TT to orient the reader and smooth the flow of discourse. The reader would feel that something is missing if it is left out. Below are more examples involving different functions of fa. 213 4.1.1 Adversative fa (2) The state of my spirit little fitted me for the society of stranger; but the meeting was inevitable. ،وبالرغم من أنني كنت في حالة نفسية سيئة تجعل من الصعب عل ّي الدخول في حديث اجتماعي م شخص غريب .فق كان ذلك األمر ال مفر منه The fa in (2) serves as a coordinating element between the dependent clause and the independent clause whereby it introduces the second clause which stands in an adversative relation to the first one. This is signaled in the ST by the contrast marker but. The target text, nonetheless, employs two DMs i.e. the fa and بالرغم منalthough/despite the fact that. However, it is the بالرغم منDM that serves as the corresponding equivalent of the ST contrast marker; it can do the job with or without the fa marker. The function of the fa is to enhance the adversative relation and smooth the flow of discourse; its deletion would only result in a less assertive tone and less cohesive discourse. In this example, therefore, we have a case of explicitation in Arabic بالرغم منcorresponding to an explicit English but, as well as a case of Arabic explicitation fa corresponding to zero equivalent in English. By contrast, the following example (3) involves an instance where TT explicitation corresponds to ST implicitation: (3) May she not give it in the future? Never! If you still persist it in maintaining our engagement, I may be your true and faithful wife, Sir Percival, but never your loving wife. أال يمكنها منح حبها له في المستقبل؟ فإنني سأكون زوجتك الوفية والمخلصة يا سير، فإذا ما زلت مصراً على االستمرار في خطوبتنا.أبدا .بيرسيفال ولكني لن أكون أبداً زوجتك المحبة This excerpt conveys the contrasting attitudes of Mr. Percival and Mrs. Farlie. Mr. Percival anticipates a loving wife in Mrs. Farlie, but the lady cannot just afford to love him. This contrast in their attitudes is signaled in the ST by the suppressed contrast marker but or yet, and it can be readily worked out by the ST reader. That is, the text might read as Never! But/Yet if you still persist it in maintaining our engagement, I may be your true and faithful wife, Sir Percival, but never your loving wife. However, this suppressed concessive marker is brought to the surface as fa in the TT in order to enhance the conditional marker and smooth the flow of discourse and, as a result, it renders the text more explicit. Sometimes, the adversative fa involves an instance where the target explicitation corresponds to zero equivalent in the ST, as can be illustrated in (4) below: 214 (4) I was struck, on entering the drawing-room, by the curious contrast, rather in material than in color, of the dresses which the ladies now wore. While Mrs. Vessey and Miss Halcombe were richly clad, Miss Farlie was poorly dressed in plain white muslin. لفت انتباهي عند دخولي قاعة االستقبال التناقض الالفت للنظر باألحرى في نوعية وليس في ألوان قماش ، فبينما ارتدت السيدة فاسي واآلنسة هالكومب فستاتين ثمينين غاية في األناقة.األثواب التي ارتدتها السيدات .كان فستان اآلنسة فارلي المصنوع من الموسلين It is clear that the second sentence in the above example conveys two contrasting ideas; it provides a clear contrast between Mrs. Vessey and Miss Halcombe's elegant dress on the one hand and Mrs. Farlie's unusually poor dress, on the other. The contrast is marked by the ST explicit contrast marker while and its counterpart in the TT بينما. However, a degree of explicitness can be observed in the TT in that, while both texts use a corresponding contrast marker, Arabic employs fa along with the contrast marker بينماin order to enhance it and smooth the flow of discourse. So, the fa in this construction has no corresponding equivalent in the ST. However, its presence is necessary as an Arab reader would feel a discoursal gap if fa is not prefixed to bainamaa. 4.1.2 Explanatory fa The function of the explanatory fa is to signal that the second clause/sentence is an explanation, comment or illustration of the preceding one. Consider the following example: (5) But my duty did not lie in this direction – my function was of the purely judicial kind. . فق كان دوري مجرد القيام بدور الحكم،ولكن واجبي كان مختلفا تماما In this example, the clause my function was of the purely judicial kind serves as an explanation of the preceding one. The ST employs an emdash to indicate explanation, whereas the TT explicitly uses the DM fa to introduce the explanatory clause. In other words, both texts correspond explicitly in that they both use an explicit explanatory marker, yet they differ in the method adopted. While the ST employs a punctuation mark, the TT settles for punctuation along with a DM. This is indicative of the syndetic nature of Arabic discourse which, unlike English, prefers a highly frequent use of DMs to achieve text competence and facilitate the reader's understanding. For instance, if the fa is not combined with the Arabic particle قدto which it is often prefixed, the TT would sound unnatural and the reader would feel something is missing. By contrast, following is an instance where TT explicitation corresponds to ST implicitation: (6) The only sign I detect of the struggle it must cost her to preserve appearances at this trying time, expresses itself in a sudden unwillingness, on her part, ever to be left alone. Instead of retreating to her own room, as usual, she seems to dread going there. والصعوبة المتناهية التي تواجهها،واإلشارة الوحيدة التي أمكنني مالحظتها والدالة على ما يعتمل في داخلها . كانت رفضها البقاء بمفردها ولو لحظة واحدة،المحافظة على مظاهر العروس السعيدة 215 في .ً كان يبدو عليها أنها كانت تخشى القيام بذلك العمل كلّيا،ًفب ال من اللجوء إلى غرفتها كما كانت تفعل سابقا As can be observed in (6), the second sentence offers an explanation for Mrs. Farlie's unwillingness to be left alone; she is possessed by fear. The DM thus/in this way is left suppressed in the ST, yet it could be perceived by the reader. Thus, the text might be interpreted as Thus/In this way, instead of retreating to her own room, as usual, she seems to dread going there. The target text, however, settles for an explicit DM in order to signal explanation and smooth the flow of discourse. Without it, the reader would feel a missing link. Therefore, this is a case where TT explicitation corresponds to ST implicitation. The following example, by contrast, represents an instance where the DM fa has no corresponding equivalent in the ST: (7) I never saw my mother and my sister together in Pesca's society, without finding my mother much the younger woman of the two. On this occasion, for example, while my mother was laughing heartily over the boyish manner in which we tumbled into the parlor, Sarah was picking up the broken pieces of a teacup, which the professor had knocked off the table in his precipitate advance to meet me at the door. كانت األولى تبدو أكثر حيوية من الثانية،أضف إلى ذلك أنه خالل استقبال والدتي وشقيقتي لصديقي بيسكا كانت والدتي تضحك جذالً من، فمثل في هذه المناسبة.طريقة تهالكنا ترحيبها به وسلوكها تجاهه ّ تحطم نتيجة ارتطام بينما كانت سارة تقوم بلملمة بقايا فنجان شاي،بيسكا به وهو يسرع لاللتقاء على األريكة .بي عند الباب بطريقة This example indicates that the two texts correspond explicitly in using the explanatory phrase for example and ًمثالrespectively. However, a degree of explicitness can be observed in the TT in that, while both texts use an explicit explanatory marker, Arabic attaches fa to it to enhance the exemplification marker and smooth the flow of discourse. In this way, we have an explicit Arabic DM that corresponds to a zero equivalent in the ST. 4.1.3 Resultative/Consequential fa The resultative fa performs a consequential function between two clauses/sentences, whereby the second expresses a state of affairs or action that comes as a result of the first one. Consider the following example: (8) It is the great beauty of the law that it can dispute any human statement, made under any circumstances, and reduce it to any form. If I had felt professionally called upon to set up a case against sir Percival Glyde, on the strength of his own explanation, I could have done so beyond all doubt. ّ إن عظمة القانون هي قدرته المصّر بها في مختلف،الفذة على تفنيد اإلفادات واألقوال البشرية كافة فلو شعرت ولو لحظة واحدة أن واجبي المهني يدعوني.الظروف ونقضها بأي شكل من األشكال 216 لما ال شك فيه كنت أقدمت على ذلك العمل، بناء على أقواله،لرف دعوى ضد السير بير سيفال غاليد .في الحال The employment of fa in this example is triggered by implicitation in the ST. It serves as the Arabic equivalent of a ST implicit resultative marker like as a result, consequently, therefore, etc. which introduces the resultative proposition. It could be observed that the first sentence presents a set of factors about the law that would naturally result in Mr. Hartright being compelled to set up a case against Sir Percival Glyde. The fa is brought to the surface in the TT in order to signal consequence and smooth the flow of discourse. Hence, this is a case of TT explicitation that corresponds to ST implicitation. By contrast, the following is an instance where the ST and TT correspond explicitly: (9) The partial cleansing of the monument had evidently been accomplished by a strange hand [...]. The work of cleansing the monument had been left unfinished, and the person by whom it had been begun might return to complete it. ]...[ تأ ّكد لي أن عملية التنظيف الجزئي للنصب التذكاري قام بها شخص غريب عن المنطقة .فأغلب الظن سيعود ذلك الشخص إلتمام عملية التنظيف التي بدأها ولم ينهها Clearly, the second part of the ST sentence serves as the result of what has transpired in the first one and is introduced by the source resultative marker and, which is the equivalent of so/therefore, etc. in this context. The TT equally employs the Arabic fa, which performs a similar function in this context. So, this is an example where the ST and TT correspond explicitly in terms of the DM. However, the translator opts for fa rather than the additive DM wa, which formally corresponds to and in the ST, in order to highlight the resultative function which would, otherwise, be blurred by the choice of the often default additive wa. 4.1.4 Causal fa The causal fa indicates the cause of an action or a state of affairs. That is, it performs a causal relationship between two sentences whereby the second sentence is the cause of the first one. Consider the following example: (10) I can do little more than offer my humble testimony to the truthfulness of Miss Halcombe's sketch of the old lady's character. Mrs. Vessey looked the personification of human composure and female amiability. وبالفعل ال يسعني إضافة شيء إلى ما قالته اآلنسة هالكومب سوى القول إن شهادتها تلك صادقة . فق كانت السيدة فاسي تجسيدا حيًا لرباطة الجأش اإلنسانية واللطف النسائي. ومطابقة تماما ً للواق As can be seen, the fa introduces a cause-result relationship between the first sentence and the second one. In the first sentence, the speaker states that he cannot afford but endorse Mr. 217 Halcombe's account of Mrs. Vessey's character, and in the second one he spells out the cause or reason for doing so, that is, she embodies serenity and good humor. Therefore, the use of fa is triggered by an implicit causal marker like because in the ST which is brought to the surface in the TT to orient the reader and smooth the Arabic flow of discourse. The reader would feel something is missing if it is not employed and the cause-result relation would be lost. 4.1.5 Resumptive fa The resumptive fa, which mostly occurs paragraph-, clause-, and sentence-initial, establishes a link between the just concluded ideas/thoughts and the following ones. It signals the continuity of discourse, with a shift of topic whereby the addresser presents the receiver with new information. Thus, it concerns the pragmatic aspect of discourse. Consider the following example: (11) As soon as Miss Farlie had left the room, he spared us all embarrassment on the subject of the anonymous letter, by diverting to it of his own accord. He had stopped in London on his way from Hampshire […]. حتى بادر هو من تلقاء نفسه إلى اإلشارة إلى موضوع،ما أن غادرت اآلنسة فارلي الغرفة فق توقف في لندن في.الرسالة اللغز موفراً على الجمي صعوبة مفاتحته بذلك الموضوع .]...[ طريق عودته من هامبشير The use of fa in the example above is not prompted by implicitation in the source text. Rather, it is employed to smooth the Arabic flow of discourse and make the translation more explicit. In fact, there is a notable mismatch between the ST and the TT when it comes to resumptive fa. This could be attributed to the fact that this type of fa, as is the case here, indicates the continuity of the discourse with a shift of topic; it presents new information within the context of the same discourse. It could be noted that the example above revolves around the anonymous letter. However, while the first sentence concerns the sense of relief felt by those present when Mr. Percival touches on the issue of his own accord, the second one concerns how he comes to know about the letter itself in London. Thus, the following sentence is related pragmatically to the preceding one. Given the asyndetic nature of English and the use of the past perfect 'had stopped' in the sentence, the ST reader can easily perceive the connection. However, the fa is employed in Arabic to signal the continuity of the discussion, create a logical link between the preceding and following sentences, and smooth the Arabic flow of discourse. 4.2 The DM ’ið According to Al-Afghani (1970), إذmay perform a causal relationship and signal suddenness. His claim goes along that of Bin Hishaam Al-Ansaari (2002/d.761h), who also adds adverbial, appositional, and additive functions. 218 The data analysis shows that إذis the second most recurring DM of the three in the translation under investigation, viz. 19 instances have been noted, constituting 34.5% of the corpus (almost tying with fa). Nonetheless, it has been identified solely with the causal and adverbial function, and it may or may not correspond to implicitation in the ST. Out of the 19 instances, إذ corresponds to implicitation in the ST in 12 cases (63%), to explicitation in 3 cases (16%), and to zero equivalent in 4 instances (21%). Consider the following extract: (12) I have resolved to prolong our stay for another week at least. It is useless to go back to Limmeridge till there is an absolute necessity for our return. […] but he is obstinate – or let me rather say, resolute. 'Merriman, I leave details to you. Do what you think right for my interest, and consider me as having personally withdrawn from the business until it is all over.' […] This is sad, but his occasional reference to himself grieves me still more. He says that the effort to return to his old habits and pursuits grows harder […]. . إذ ال فائدة من العودة الى ليميريدج ما لم يطرأ أمر ضروري يحتم عودتنا،قررت تمديد إقامتنا هنا أسبوعا آخر ّ "سأترك لك االهتمام بالتفاصيل يا: إذ قال،موطد العزم أو باألحرى،] ولكنه رجل عنيد...[ . وإلى أن نتوصل إلى إتفاق اعتبرني كأنني انسحبت من المسألة بكاملها، افعل ما تراه مالئما لمصالحي.ماريان ولكن الذي زاد في حزني عليه ذكره العرضي إلى ما،] شعرت باألسف جرّاء هذا األمر...[ .]...[ إذ قال إنه قد صعب عليه كثيرا العودة إلى متابعة أعماله وعاداته المعتادة،آلت إليه أحواله As can be noted, the first mention of إذcorresponds to a causal marker because, which is suppressed in the ST and brought to the surface in the TT to orient the reader and smooth the flow of discourse. This is an example of Arabic explicitation that corresponds to English implicitation. The second sentence of the ST provides justification for the speaker's action in the preceding one. That is to say, it gives the reason why the speaker decides to prolong their stay for another week. So, the Arabic causal marker إذ, whose nearest equivalent could be since, because, etc. in this context, is used here as a corresponding element to the implicit causal maker in the ST. By contrast, in the second mention, إذis attached to a reporting clause and does not correspond to implicitation in the ST. In fact, the translation exhibits addition in two consecutive instances; addition of the reporting clause he said قال, which is left implicit in the ST and insertion of the adverbial DM إذ, which corresponds to zero equivalent in the ST. The addition of the reporting clause ' قالhe said' is intended for speaker identification, the reason being the lengthy discourse, in which the two lawyers, Mr. Gilmore and Mr. Percival, are the participants, is fairly marked by suppression of speaker identity. It should be noted that the discourse revolves around their respective clients, and they sometimes sound as if they were reporting what their clients say. Thus, the reporting clause linking Mr. Gilmore to the speech in the quotes is supposed to resolve any potential ambiguity regarding his identity. However, it would be better, even more acceptable in this context to substitute the phrase إذ قالby a more appropriate expression like 219 وتاب قائالto resolve the potential ambiguity. The utilization of إذ قال, which translates into something like when he said or as he said, makes the speaker sound as if he were quoting a client and thus failing to resolve the ambiguity, which may lead to altering the meaning potential of translation. This shows an erroneous case of employing addition. It also means that marked (unjustified) explicitation exists even at the level of discourse markers. In the third mention, إذequally performs an adverbial function because it may be translated as when or as, which is not implied in the source text. Nonetheless, the use of the adverbial إذis significant in that it smoothes the flow of discourse and renders it more natural. Without it, a gap would be felt in the Arabic text. Below are more examples of the functions of إذin the data. 4.2.1 Causal ’ið The causal إذintroduces a sentence that describes the cause or gives the reason for the action or state of affairs in the preceding one. Consider the following example: (13)He was evidently in search of me, for he quickened his pace when we caught sight of each other. . إذ ما أن شاهدني حتى أسرع نحوي بخطى سريعة،بدا وكأنه كان يبحث عني As can be seen, the second sentence of the ST provides justification for the speaker's claim in the preceding one. That is to say, it gives the reason why the speaker claims that He was evidently in search of me. This clause is introduced by the English causal marker for. Therefore, the Arabic causal marker إذis used here as a corresponding element to the explicit causal maker in the ST. This is a case where the language pair corresponds explicitly in the employment a DM. By contrast, the following is an instance of Arabic explicitation that corresponds to implicitation in the ST: (14)To tell you the truth, I am uneasy about Laura, she has sent to say she wants to see me directly […]. .]...[ إذ أرسلت في طلبي على وجه السرعة.أصارحك القول يساورني القلق بشأن لورا It is clear that the causal marker إذis brought to the surface in the TT to serve as the corresponding equivalent of an implicit counterpart because, for, etc. It can be noted that the second clause describes the reason why the speaker feels perturbed by Laura's request. The ST reader can easily perceive the connection as well as the suppressed causal marker. However, Arabic needs to bring the causal marker to the surface in order to orient the reader and naturalize the flow of discourse. The Arabic text would be incohesive without it. One should note that the translator has erroneously punctuated this DM with a period rather than a correct comma; this DM can only introduce a dependent clause that cannot stand on its own, just like a dependent because clause in English. 4.2.2 Adverbial ’ið There is a unanimous consensus among Arabic grammarians that the primary function of إذis an adverbial one (Medieval Bin Hishaam Al-Ansari 2002/d.761h). This type falls under the category 220 of what (Khalil, 1999, p. 252) refers to as "adverbial object, which is a noun, in the accusative case that denotes the time and place of the verb". Consider the following example: (15)Mr. Farlie's answer reached me by return of post, and proved to be wandering and irrelevant in the extreme. "Would dear Gilmore be so very obliging as not to worry his friend and client about such a trifle as a remote contingency?" "أيمكن للعزيز غيلمور: إذ قال،كان جواب السيد فارلي على رسالتي هائما ً جداً وبعيداً تماما ّ عن صلب الموضوع ّ أن "يتلطف بحيث ال يقلق صديقه ومو ّكله بمثل هذا االحتمال الضئيل التافه؟ The clause إذ قالis brought to the surface in the TT to be the corresponding equivalent of a parallel implicit clause in the ST. The reporting clause he said, to which قالcorresponds, is suppressed in the ST and the adverbial marker إذ, which could be translated as when or as is not implied in the ST. The reporting clause is intended for speaker identification by associating Mr. Farlie with the quoted question because Mr. Gilmore's narration is fairly marked by suppression of speaker identity. Thus, the reporting clause linking Mr. Farlie to the speech in the quote is supposed to resolve any potential ambiguity regarding his identity. The DM إذ, which corresponds to a zero equivalent in the ST, performs an adverbial function and, therefore, helps to keep the flow of discourse cohesive and smooth. However, it would be better, even more acceptable in this context, to substitute the phrase إذ قالby إذ سألsince the following quote is a question rather than a statement. The utilization of إذ سألis justified here because it makes the reader realize that the speaker is quoting the character he has just mentioned and, consequently, resolve any ambiguity that might arise from identity suppression. 4.3 The DM bittaalii Despite its being a pervasive feature of Arabic discourse, research on the DM بالتالي, which may be translated as therefore, consequently, thus, as a result, is almost nonexistent. Review of the existing literature yields no results, except for Al-Mu'jam-l-ghanni (E-version) by Abul-Azm. ً نت، من ث ّم, According to this dictionary, this DM communicates the same meanings as يجة لما سبق and إذ. This indicates that بالتاليonly performs the resultative/consequential function, which is, in fact, the only function it has been identified with in the translation under investigation. The data shows 16 instances of this DM, making up 29% of the corpus. It corresponds to implicitation in the ST in 8 cases (50%), to explicitation in 7 cases (43.75%), and to zero equivalent in 1 instance (6.25%). The resultative/consequential بالتاليfunctions to either establish a link between two clauses of a compound sentence where the second clause occurs as a result of the preceding one or to introduce a sentence that occurs as a consequence of the preceding one. Consider the following example: (16) Mr. Gilmore is the old friend of two generations of Farlies, and we can trust him, as we could trust no one else. وبالتالي يمكننا،علما ً أن صداقة السيد غيلمور م أسرة فارلي قديمة تعود إلى جيلين من الزمن 221 .الوثوق به كليا According to Quirk, et al. (1986), the English DM and signals multiple textual functions including the consequential one as in the above example. Since the Arabic DMs bittaalii and wa, which co-occur in the above example, can equally perform the same function, i.e. the consequential function, bittaalii, being more semantically oriented, could be seen as the corresponding element to the English DM and, while the addition of wa is meant to make that function more explicit and the discourse more cohesive. It should be noted that wa, which is usually employed as a default DM whose main function is to cater to cohesion in Arabic discourse, can carry the weight of a semantically-loaded DM alone (as in the example above). However, most writers in Arabic prefer to employ a more semantically-oriented DM (bittaali here) and, at the same time, keep wa as an enhancer of the logical relation as well as a cohesive marker. By contrast, there are instances where the target DM has no equivalent in the ST. Consider the following example: (17)There are no such things as ghosts, and therefore, any boy who believes in ghosts believes in what can't be. يكون بالتالي، فأي ولد يصدق وجود األشبا، وبالتالي، ال وجود على االطالق لشيء يسمّى باألشبا .يصدق أمراً اليمكن أن يحصل على االطالق As can be seen, there is optimal formal correspondence between the DMs of the TT and their ST counterparts in the first occurrence. In other words, the target DMs وبالتاليare the corresponding equivalents of the ST DMs and therefore. By contrast, the second mention of the DM بالتاليhas no corresponding element in the ST. As a matter of fact, it represents an erroneous case of employing this DM, given its occurrence in the immediately previous sentence and, therefore, it makes the translation sound redundant. The first bittaalii should be kept, while the second one should be deleted in order to avoid redundancy and offer natural Arabic discourse. A more natural version could be achieved by a rendition like يكون ق صدّق أمراً ال يمكن أن يحصل على االطالق, thus using the Arabic confirmatory particle قدinstead to naturalize and smooth the flow of discourse. 5. Conclusion The argument presented in this paper runs counter to Blum-Kulka’s (1986) hypothesis that instances of explicitation in the TT must correspond to instances of implicitation in the ST. BlumKulka does not seem to have taken into account the nature of different languages. The data indicates that this claim is valid in some cases but invalid in others, i.e. DMs may correspond to implicitation in some instances but may not in others. The discussion of three Arabic DMs fa, ’ið and bittaalii, which perform different discoursal functions including the adversative, explanatory, causal, resultative, resumptive, and adverbial function, shows that they may correspond to explicitation, implicitation and zero equivalent in the ST. The employment of DMs in Arabic discourse ranges between marking purely logical relations and rendering the discourse 222 more cohesive. Apart from its frequent use as a default DM, it is generally felt that wa is too light a DM to mark a logical relation; hence, it is mainly used to enhance other semantically oriented DMs. It can be argued that what obtains between Spanish and English also obtains between English and Arabic, as far as DMs are concerned (Saldanha 2008). This study demonstrates through authentic translational data that Arabic makes frequent use of DMs because of the syndetic nature of its discourse, unlike English whose discourse is equally asyndetic. This being the case, formal correspondence between English and Arabic in terms of DMs cannot be stipulated. Besides naturalizing and smoothing the flow of discourse, Arabic DMs facilitate the reader's understanding of the text through creating the necessary semantic and pragmatic links. Nonetheless, some erroneous cases of employing DMs in professional translation into Arabic may occur, something which renders the translation redundant and/or unnatural. 223 Lexical Reduction in Scientific Translation Mohammed Farghal & Mashael Al-Hamly Abstract The present paper is a case study of the rendition of English Reduced Forms (RLFs) in Majalat AlOloom (the Arabic version of Scientific American). It works with the assumption that RLFs can be problematic in Arabic translation because English commonly favors the employment of such forms while Arabic opts for reduced forms only infrequently. The purpose is to examine authentic Arabic translational data in an English text type (popular science articles) that usually abounds in the use of RLFs, in order to see how translators render them into Arabic. The data shows that professional scientific translators employ various strategies in rendering a variety of English RLFs. While Blended Forms and Complete Form + RLF are the most frequent RLFs in the English corpus, Translation Alone and Translation + RLF are the most occurring strategies in the Arabic corpus. The study offers both a qualitative and quantitative analysis of the data. 1. Lexical Reduction in English Reduction or subtraction (Kreidler 1998) is a major word formation process in English. It involves the removal of some parts of existing lexical items/phrases in order to create new ones. Essentially, this is motivated by the principle of economy but some reduced forms may also be instigated by the desire to improvise professionalism and euphony (Newmark 1988). Nowadays, English, which is widely known for its tendency to reduce words/phrases, holds the banner of science and technology and, consequently, offers scores of reduced lexical terms such as DNA (deoxyribo nucleic acid), AIDS (Acquired Immune Deficiency Syndrome), LCD (liquid crystal display), DVD (digital versatile/video disc), NASA (National Aeronautics Space Administration), etc. This will certainly make the work of translators more taxing, requiring them first to understand what the RLF means (which, in many cases, involves tracing the RLF back to its full form) and then to find an appropriate strategy to render it in the Target Language (TL). English RLFs take various forms. The use of the initials of the words in lexical phrases to construct new words stands out as the most familiar process. This involves two procedures which differ only in the pronunciation of the output. The first procedure offers acronyms which are pronounced as ordinary English words like NATO (North Atlantic Treaty Organization), FAO (Food and Agricultural Organization), NASA, AIDS, etc., while the second results in abbreviations which are pronounced as individual letters such as USA (United States of America), UN (United Nations), LCD, DNA, etc. (Yule 1985). A few common abbreviations feature Latin individual letters such as ie for id est, eg for exempli gratia, am for ante meridiem 224 and pm for post meridiem. Other familiar abbreviations may take the initials of two syllables in the word and have them stand for the whole word, e.g. TV for television and TB for tuberculosis, or the first and last letters in the word, e.g. Dr for doctor, St for saint, Jr for junior, and Sr for senior. Note that some of these abbreviations occur as written forms only and receive the pronunciation of the entire word in spoken communication, e.g. Dr and St, while others are spoken out as individual letters, e.g. Jr and Sr. The second most common English RLFs involve the application of clipping to exiting words in order to create new ones. The most common procedure in clipping is to cut off the first (or, to a lesser extent, the last) syllable in the word and have it stand for the whole word. Examples include gym, lab, bike, phone, and bra, among many others. In some cases, the clipped forms occur in writing only and are pronounced as full words, e.g. Eng for engineer or English and Rev for Reverend. Another familiar reduction process in English is the blending of two lexemes to produce a word representing a new concept. The standard procedure is to combine the onset consonants and/or syllable(s) of one word with the last syllable of another. Examples include brunch from breakfast and lunch, smog from smoke and fog, and autocide from automobile and suicide, among many others. A more common manifestation of blending in technical materials involves taking one part of a word (one or more syllables) and combining it to a full word. Examples include geopolitics, psycholinguistics, ecosystem, Afroasiatic, carboxide, morphophonemic, and biochemistry, among many others. 2. Lexical Reduction in Arabic Arabic, in contrast with English, employs RLFs on a small scale. Traditionally, the use of a modified version of acronyms was confined to phrases with a religious tinge (They all feature allah). The procedure involves employing verbs featuring the most salient and/or important sounds in a phrase/sentence to indicate the uttering of full phrases/sentences such as hallala ملل for the sentence laa ’ilaah ’illaa ’allah ' و إله إو هللاThere is no God but God’, kabbara كfor the utterance ’allahu ’akbar ‘ هللا أكGod's the greatest’, basmala صسالfor the phrase bismi-llaah ilraħmaan il-raħiim ‘ صس هللا ال اا ال ايIn the name of God, the most gracious, the most merciful’, and ħawqala اوصلfor the sentence laa ħawla walaa uwata illaa billaah ‘There’s no refuge nor strength but in God’, among a few others. Few of them are only written forms such as Prophet Mohammed’s honorific sal‘am صلعfor the sentence salaa ’allahu ‘alayhi wa salam صلى هللا عليه ‘ وولMay peace be upon him’. In modern standard Arabic, one might find a few abbreviations and blends. Familiar examples of written abbreviations which are spoken out in full include the Arabic letter . وfollowed by a dot for duktoor ‘doctor’ وك ور, the letter . مfollowed by a dot for muhanndis ‘engineer’ دان س, the letter مafter a date for miilaadii ‘AD’ ديًو, and the letter after a date for hijrii ‘Hijri /Islamic 225 calendar’ مج. Pure blends, for their turn, are very infrequent in Arabic; examples may include mutašaa’il ‘optimistic and pessimistic’ from mutafaa’il ‘ د فائلoptimistic’ and mutašaa’im د شائ ‘pessimistic’ and ’a‘di aa’ ‘ أع صاenemies and friends’ from ’a‘daa’ ‘enemies’ and ’asdi aa’ ‘friends’. However, due to the continued influence of English technical terms, Arabic sometimes employs blends where one part of the first word is attached to the second word in full to form what is called in Arabic ‘ الا كب الازجthe blended compound’ such as kahrumaγnaatiisii ‘ كا ودغناطيسelectromagnetic’ from kahrabaa’i ‘ كا صائelectric’ and maγnaatiisii دغناطيس ‘magnetic’ and ’afru’aasyawii ‘ أف وآويوAfro-asiatic’ from ’afrii ii ‘ أف قAfrican’ and ’aasyawii ‘Asiatic’ آويو. 3. Translation of English RLFs English, which is an international lingua franca nowadays, offered and is still offering scores of RLFs, particularly the employment of initials to stand for lexical phrases, in various fields and occupations including business, industry, tourism, science, technology, academia, etc. The expert as well as the translator is astounded by the huge number of these RLFs and the opaqueness encapsulated thereof (Spencer 1988). The general tendency to use RLFs in written as well as spoken English discourse, which is mainly motivated by economy, euphony, and prestige and/or professionalism (cf. Newmark 1988), may cause serious problems to translators in first interpreting SLFs and then in rendering them into the Target Language (TL). The kind of problems occurring depends on whether the language pair involves genealogically related languages, e.g. English and other European languages, or those genealogically unrelated languages, e.g. English and Arabic. In the former case, initialisms whose source is not English may vary across European languages, which are receptive to such lexical reduction, e.g. SSSR, USSR, UDSSR, and URSS in Russian, English, German and French respectively all refer to former Soviet Union. The variation emanates from the initials in the phrase in each language. However, more recently, there is a tendency in several European languages to borrow English initialisms formally. Examples include English internationalisms such as AIDS, DNA, DVD, LCD, etc. (González 1991; Bankole 2006). In the latter case, English and Arabic, the rendering of initialisms becomes more complicated because Arabic does not use the Latin alphabet in the first place, nor does it behave receptively of this lexical phenomenon in the second place (cf. Al-Qinai 2007). Translators of English initialisms into Arabic, therefore, may adopt different strategies in rendering them. By way of illustration, let us take the familiar English initialism DNA, which is variously translated into corresponding individual Arabic letters reflecting the English pronunciation ( )و ان ا, or an Arabic acronym ()ال ا, or an Arabic loan translation ( الحا النوو/ )ال يا الوراثي, or a combination of two or more of them. 226 4. Purpose, Material and Procedure The purpose of this study is to examine authentic Arabic translational data in an English text type (popular scientific discourse) which frequently employs the use of RLFs, in an attempt to see how translators render them into Arabic. Consequently, the study will offer a typology of the strategies used, along with a qualitative and quantitative discussion of each strategy. The material consists of 15 Arabic translations (See Appendix 1) of their Scientific American originals (See Appendix 2) appearing in Majalat AlOloom, a publication of Kuwait Foundation for the Advancement of sciences (KFAS), Kuwait. Each English article will be closely examined for RLFs, noting whether the RLF is used alone or alongside the full form. Then, the English data are juxtaposed with the Arabic translational counterparts and subjected to careful analysis. As is clear, the study material is a par excellence sample of scientific translation. Scientific translation, it is generally believed, focuses on the accurate conveyance of the content of scientific discourse. Ilyas (1989: 109) writes "In scientific works, subject-matter takes priority over the style of the linguistic medium which aims at expressing facts, experiments, hypothesis, etc. …… All that is required in fact is that of verbal accuracy and lucidity of expression". If the translator is required to improvise verbal accuracy and lucid expression, one wonders what remains of the linguistic medium Ilyas is underestimating in the above quote. There is no question that the translator's knowledge in the relevant area is a prerequisite in scientific translation (Nida 1964; Giles 1995; Saedi 1996; Al-Hassnawi 200). However, an effective rhetoric of scientific discourse should accompany the rendition of accurate scientific information. Sharkas (2009), in an article titled "Translation Quality Assessment of Popular Science Articles Corpus Study of the Scientific American and its Arabic Version", criticizes the translation quality which focuses on factual quality and noticeably underestimates textual adequacy which caters for cohesion and coherence. She rightly concludes that it is not enough to have science articles translated and edited by specialists in the relevant areas in the absence of translation training that sharpens their transfer competence. It should be noted that the present study is not meant to conduct a quality assessment study of the Arabic version of Scientific American. Other things being equal, the purpose of this investigation is to examine how professional specialist translators deal with English RLFs in Arabic translation as sampled by Majalat AlOloom. Therefore, the discussion, analysis and critique therein are solely confined to RLFs, the feature under investigation. 5. Categories of English Data 227 There are three categories of the English data extracted: 1. The RLF is employed immediately after the complete form in the first mention but alone in subsequent mentions, e.g. human accelerated region 1 (HAR1), Throughfall Displacement Experiment (TDE), net primary production (NPP), Resource Description Framework ((RDF), Universal Resource Identifier (URI), etc. 2. The RLF is employed independently of the complete form. This category falls into five subcategories: a) Abbreviations (pronounced as individual letters), e.g. DNA, RNA, HIV, U.S., etc. b) Acronyms (pronounced as words), e.g. NASA, AIDS, etc. c) Clipped (spoken and written) forms, e.g. lab, chimp, etc. d) Written Clipped forms, C (Celsius), Calif. (California), Colo. (Colorado), etc. e) Blended forms, e.g. iPhone, e-mail, ecosystem, biodiversity, geothermal, etc. 3. The RLF (employed alone) modifies a Generic term (such as virus, gene, experiment, etc.) in the first mention but alone in subsequent mentions, e.g. FOXP2 gene, PrERV1 virus, AMY1 gene, etc. Table 1: Number of occurrences for the categories of English data in terms of frequency. Categories of English Data Category type Number of occurrences Percentage Blended form 39 32.5% Complete form + RLF 30 25% Abbreviation 25 20.8% Written Clipped form 15 12.5% RLF + Generic form 5 4.2% Acronyms 4 3.3% Spoken Clipped form 2 1.7% Total 120 Table 1 above indicates that the total number of English RLFs in the corpus is 120 first mention instances (subsequent mentions are not reported) which are distributed over 7 categories. Noticeably, blended forms, complete forms + RLFs and abbreviations account for about 76% of the data. This comes as no surprise because the high technicality of the terminology in scientific discourse necessitates the employment of such techniques. 6. Strategies Employed in Translating RLFS The corpus instantiates 6 Arabic strategies in translating English RLFs as follows: 1. Translation Alone: The translator simply translates the English source term into Arabic. 228 2. Translation + RLF: The translator translates the word/phrase and follows it with the complete form and reduced form. 3. Arabic Generic Word + Borrowing: The translator borrows the English term but precedes it with an Arabic generic term to clarify the translation. 4. Borrowing Alone: The translator borrows the English term and writes it using Arabic alphabet. 5. Borrowing + Acronym: The translator borrows the English term in Arabic alphabet and follows it with the English acronym. 6. Translation + Borrowing: The translator translates as well as borrows the English term/phrase. Table 2: Number of occurrences of translation strategies in terms of frequency Translation Strategies Percentage Strategy Type Number of occurrences Translation alone 50 41.6% Translation + RLF 27 22.5% Arabic generic word + Borrowing 22 18.3% Borrowing alone 11 9.1% Borrowing + Acronym Translation + Borrowing Total 5 5 120 4.4% 4.4% Out of the six strategies noted in the translation strategies, the highest percentage is the translation alone strategy, followed by the use of translation with the reduced. The following discussion sheds more light on each of these strategies. 6.1 Translation Alone This is the most frequent strategy in the data (41.6%). The high frequency of this strategy derives from the fact that it is employed in a variety of RLFs where the translator expects the reader to be familiar with the Arabic translation independently of the English RLF. The variety includes: a) Clipped Forms such as: كاليفور يا ‘Calif.’ اواا ‘Neb.’ الشا ا ز دخ ‘chimp’ ‘lab’ b) Abbreviations such as: 229 الوو ات الا ح األد اي ‘U.S.’ ال ك ورا ‘Ph.D.’ جادع كاليفور يا ‘U.C.’ ص طا يا 'UK' c) Blended Forms such as: الانظود ال يئي ‘ecosystem’ ' اخ ياصي ف اإلايا الحيوbiostatistician'. األووط دنطق الغ الطاص الح ار األرضي ‘Midwest’ ‘geothermal energy’ Firstly, clipped forms may include both proper and common nouns. On the one hand, clipped forms of proper nouns are rendered in Arabic alphabet in full, e.g. كاليفور ياfor 'Calif.'. On the other hand, clipped forms of common nouns are translated into their Arabic full counterparts, e.g. دخfor 'lab'. It should be noted that clipped forms like these must be rendered into their Arabic full counterparts as there are no corresponding clipped forms in Arabic. Secondly, there are some English abbreviations pronounced as individual letters which must be translated into their Arabic full forms such as USA and PhD. Examples like these are never borrowed using Arabic alphabet. However, there are some abbreviations (not found in our data) which may lend themselves to borrowing through Arabic alphabet and are pronounced as individual letters such as و أ أfor 'CIA' and ص ص وfor 'BBC'. Thirdly, and more importantly, scientific discourse frequently features blended technical forms where the first syllable of the first technical word attaches to the full form of the second word to form a compound concept. The corpus shows that most of these terms are translated into Arabic phrases featuring two full forms or more, e.g. الانظود ال يئيfor 'ecosystem' and اخ ياصي ف اإلايا الحيوfor 'biostatistician'. As is clear, this strategy unpacks the meaning of the English blended term employing Arabic translation alone. 6.2 Translation + RLF Table 2 above shows that the second most frequent rendition of a technical RLF involves an Arabic translation of the complete form immediately followed by the English RLF. This strategy accounts for 22.5% of the overall data. It is clear that the translator endeavors to make sure that 230 the Arabic translation is processed correctly. The reader, the translator feels, may have difficulty processing the Arabic translation, in which case s/he can rely on the English RLF. The translator can look the RLF up to know what it stands for. The following examples involve this strategy: HAR1 ) الا سارع (الاعجل CO2 أكسي الا صون النااي ال ش ثا RDF لغ إطار عال وصف الاوارو In the first example, the translator gives an Arabic translation of the English complete form (human accelerated region 1) along with the RLF. However, instead of using the abbreviation alone in subsequent mentions (as the English text does), the translator employs the translation of the head noun in the phrase ( الناايregion) to preface the RLF in every subsequent mention, viz. HAR1 ‘ الناايregion HAR1’. This is meant to facilitate processing by the reader, albeit it duplicates one of the initials (the one for head noun). The second example offers an Arabic translation along with the RLF in the first mention but only the Arabic translation in subsequent mentions. Apparently, the translator feels comfortable with his Arabic rendition and expects the reader to process it easily, thus doing away with the RLF, i.e. CO2. The third example offers a five-word Arabic translation of a three-letter abbreviation (RDF) whose complete form (Resource Description Framework) is given in a footnote. The complete English form and the abbreviation are employed parenthetically to explain what the term ‘the primary language’ means viz. ‘… the primary language – the Resource Description Framework (RDF) – which is layered onto the basic HTML and …’ (Web Science Emerges, p. 79). The translator successfully uses the head noun ‘ لغlanguage’ to modify the name of the program. In the subsequent mentions, s/he uses the generic word followed by the RLF, viz. RDF اللغ. Interestingly, the complete form of the second abbreviation in the quote above (HTML) is not given in the English text, probably because the authors deem it a familiar RLF in web science. This abbreviation stands for HyperText Markup Language. The translator just borrows the abbreviation without rendering what it stands for. However, he employs a generic word to preface it, viz. HTML ‘ ط ق الlayer HTML’. Apparently, the translator mistook the L in the abbreviation to stand for ‘layer’ rather than the target ‘language’, hence the erroneous rendering. He should have looked up the complete form of the RLF in a specialized dictionary in order to render its content, viz. اللغ ال ديز للنص الفوص. 231 6.3 Arabic Generic Word + Borrowing (in Arabic or English alphabet) This strategy comes third in the data (18.3%. The translator here opts for prefacing the abbreviation or acronym with a generic word to facilitate the processing for the reader. In many cases, the Arabic generic word duplicates the last initial in the English RLF as can be illustrated in the two examples below: NASA ‘ وكالNASA agency’ HIV ‘ الفي وسHIV virus’ In the first example (NASA), the initial 'A' for generic word 'Administration' is further duplicated by the Arabic generic ' وكالagency'. Similarly, in the second example (HIV), the initial 'V' for the generic word 'virus' is duplicated by the Arabic generic word ' الفي وسvirus'. Cleary, the translator wants to naturalize the discourse as well as facilitate comprehension. In some cases, the source text employs generic words in the first mention of highly technical RLFs. However, the target text will employ the generic term in all subsequent mentions, e.g. ASPM ‘ الجينASPM gene’. This strategy compensates for the zero translation of such terms by the continued employment of the generic term. Besides, the terms would sound unnatural without the generic term. In few cases, the translator, not being certain that the generic word would do the job, may reinforce the generic word with a post-modifier immediately after the borrowing, as in: الا فجTNT ‘ داو الTNT explosive material’. 6.4 Borrowing alone There are a few examples where the RLF is borrowed alone whether in Arabic alphabet or English alphabet (9.1%). The following examples are illustrative: األ ر for 'AIDS' ال ا for 'DNA' ال ا for اليو يسيف for 'UNISEF' IEEE for 'IEEE' 'RNA' 232 Tg for 'Tg' As can be seen, English acronyms may be borrowed as words in Arabic alphabet, e.g. اليو يسيفfor 'UNICEF' and األ رfor 'AIDS'. Similarly, in two recurrent cases and due to editorial policy, the English abbreviations DNA and RNA are borrowed as Arabic acronyms ال ا و ال ا, i.e. they are pronounced as Arabic words (See strategy 2 above). The reason for the adoption of borrowing alone seems to stem from the universal or local (magazine-specific) familiarity with these terms. In some cases, rendering the abbreviation as is in English alphabet is indicative of the translator's inability to translate the sense of the RLF and, consequently, constitutes zero translation. The employment of 'IEEE' (which is an online database) and Tg (which is a measurement of 'thyroglobulin') in the Arabic text without any attempt to translate the concept is a serious weakness in the translation. The specialist translator should be able to combine borrowing with translation in order to facilitate comprehension. 6.5 Borrowing + Acronym This strategy is often applied in cases of familiar English internationalisms where the translator works with the assumption that the reader is aware of such abbreviations. However, he reinforces the borrowing in Arabic alphabet by mentioning the English RLF. It should be noted that the Latin alphabet is replaced with the Arabic in such borrowings, that is, transliteration, which employs the Latin alphabet, is not available when borrowing into Arabic, whereas it is adopted when borrowing from Arabic into English, e.g. جااوis transliterated as jihad and اااسis transliterated as Hamas. The data instantiates only few examples of such internationalisms (4.4%), two of which (DNA and AIDS) are highly recurrent in the corpus. The first (DNA), which is an abbreviation pronounced as individual letters in English, is systematically acronymized as ' ال اaddanaa' in Arabic. Apparently, this is an editorial decision to adopt this Arabic acronym as an equivalent for this recurrent term in science materials. It should be noted, while Arabic can readily acronymize English acronyms such as AIDS األ ر, UNESCO اليو ساو, and NATO الناتو, it is extremely uncommon to acronymize English abbreviations such as DNA. The usual practice is to either translate it (viz. )الحا النووor borrow it as individual letters in Arabic alphabet (viz. )و أن أ. It should also be noted that the Arabic definite article الis added to Arabic acronyms which only require the zero article in English. When adopting this strategy, the translator includes the English RLF only in the first mention, i.e. she/he only employs the Arabic acronym in subsequent mentions. 233 6.6 Translation + Borrowing The corpus contains few cases (4.4%) where translation is combined with borrowing in Arabic or English alphabet, as can be illustrated in the following examples: GM biodiversity agroecological Chromosome 4E ،الاحور جينيا ال يولوج،ال نو أ اولوج رراع 4E الي غ Apparently, the translator in the first three examples above opts for borrowing the technical part of the RLF in English alphabet instead of employing translation alone. While this decision is transparent in the first two items because the borrowed terms are as familiar as their Arabic counterparts, viz. الحيو، و ال نو، الاحور وراثيا, respectively, the decision in the third item is rather opaque as the Arabic counterpart, viz. صيئ رراعis more transparent to the Arab reader. As for the fourth item, the translator rightly retains the English abbreviation (4E) but unjustifiably translates the generic word 'chromosome' into الي غ, which is largely unfamiliar to the Arab reader. A better option is to borrow it in Arabic alphabet, i.e. ك ودوروم, which is a very familiar borrowing in Arabic. 7. Categories of English Data in Relation to Translation Strategies To shed more light on the relation between the most frequent English categories that were used in the source English text, i.e., blended forms and complete form + RLF (See Table 1), and the translation strategies employed in translating these categories, let us examine Tables 3 and 4 below. Table 3: Number of occurrences of English category Blended form in relation to translation strategies. Strategy Type Translation alone Translation + RLF Borrowing alone Arabic generic word + Borrowing Borrowing + Acronym Number of occurrences 29 5 3 2 0 Percentage 74.4% 12.8% 7.7% 5.1% 0% Borrowing clipped into full form Translation + Borrowing Total 0 0 39 0% 0% 234 As discussed in section 5 above, the use of the blended form was the most frequent category among the data under analysis. Having a closer look at those blended forms, it is found that the bulk of blended forms (74.4%) is rendered into Arabic by translation alone (As shown in Table 3). This simply indicates that what is familiarly known in Arabic as 'the blended compound' is not a priority by the Arab translator when translating blended forms. It can be argued that the main motivation here is to achieve transparency by avoiding opaque blended compounds. For example, the blended form 'biochemistry' will be translated into the transparent الايايا الحيوrather than the opaque Arabic blended compound ال يوكيايا. It should be noted, however, that English adjectival blended forms do not lend themselves readily to the strategy of translation alone, e.g. 'biochemical' is translated into the Arabic blended compound صيوكيااوrather than the awkward ( ايو كيااوin which two adjectives are employed) or the rather opaque blended compound, ايوكيااوwhere ايوfunctions as a blending syllable from ‘ ايوbiological’. Thus, while nominal blended forms lend themselves easily to translation into Arabic by rendering the blending syllable as a full adjective الحيوto post-modify the noun الايايا. This process is not usually available for adjectival blended forms where translation plus borrowing becomes more likely, albeit more opaque. Table 4: Number of occurrences of English category complete form + RLF in relation to translation strategies. Strategy Type Translation + RLF Arabic generic word + Borrowing Translation + Borrowing Borrowing + Acronym Translation alone Number of occurrences 12 12 Percentage 40% 40% 2 1 3 6.7% 3.3 10% Borrowing alone Borrowing clipped into full form Total 0 0 0% 0% 30 Also, as mentioned in section 5 above, the second most frequent English category was the use of English Complete Forms + RLF. This category (as shown in Table 4) is rendered by Translation + RLF (40%) and Arabic Generic Word + Borrowing (40%). Both strategies aim to facilitate the processing of such highly technical phrases by the addition of the RLF or the employment of an Arabic generic word. It can be argued that these two strategies constitute the key to dealing with discipline-specific English RLFs. The translator's main task is to achieve a good degree of transparency of such RLFs in Arabic scientific discourse. In few cases, the strategy of Translation alone will be sufficient where the transparency of the Arabic rendition is very high. 235 8. Translation Strategies in Relation to Categories of English Data This section discusses another aspect of the data, that is, the relationship between the most frequent translation strategies employed, i.e., translation alone and translation + RLF (See Table 2), and the categories of English data that they were derived from. Tables 5 and 6 below provide a detailed breakdown of these two strategies. Table 5: Number of occurrences of English categories in relation to translation alone strategy. Category type Blended form Abbreviation Written Clipped form Complete form + RLF Spoken Clipped form Number of occurrences 28 14 5 2 1 Percentage 56% 28% 10.% 4% 2% Acronyms Generic term + RLF Total 0 0 50 0% 0% Table 5 predictably shows that the most frequent input for Translation Alone is blended forms (56% / see discussion of Table 3 above). More interestingly, the Table indicates that the second most frequent input for Translation Alone is abbreviations (28%). To explain, abbreviations, which are pronounced as individual letters, fall into two categories: technical discipline-specific abbreviations, e.g. NPP for Net Primary Production and general, non-technical abbreviations, e.g. USA for United States of America. While the first category usually requires Translation combined with other strategies (see discussion of Table 4 above), the second category requires translation alone. Examples include الوو ات الا حfor USA and ال ك وراfor PhD. Table 5 also indicates that written clipped forms usually require Translation Alone (10%). Examples include دقاصلfor vs 'versus', دئوfor C 'Celsius', and كا ساسfor Kan 'Kansas'. Table 6: Number of occurrences of English categories in relation to translation + RLF. Category type Complete form+ RLF Abbreviation Blended form Written Clipped form Acronyms Number of occurrences 18 5 5 0 0 Percentage 66.6% 18.5% 14.8% 0% 0% Spoken Clipped form Use of RLF with a generic term Total 0 0 27 0% 0% 236 Table 6 shows that the most frequent input for the strategy Translation + RLF is Complete Form + RLF (66.6%). The employment of the RLF in the Arabic rendition is meant to ensure the comprehension of such highly technical terms. If the translation is used alone in the first mention, comprehensibility may become at stake. Examples include (TDE) دشار إرال الاطول الاطfor Throughfall Displacement Experiment (TDE) and (TLRS) الاس ق ًت الش يا صالحاجزfor Toll-like receptors (TLRS). As for the other less frequent inputs for Translation + RLF, they are Abbreviations (18.5%), e.g. CO2 ثا أكسي الا صونfor CO2 and Blended Forms (14.8%), e.g. ecosystem النظ ال يئيfor ecosystem. In both cases, the use of the English RLF aims to consolidate the comprehensibility of the usually transparent Arabic translations. 9. Conclusions 1. The data shows that English RLFs in popular scientific discourse require a special treatment when translated into Arabic. The most taxing problem is to render the technical RLFs in a comprehensible manner while preserving the scientific jargon. This usually necessitates combining translating with borrowing to ensure that the reader understands the RLF. 2. In the absence of translating, the strategy of borrowing usually involves the employment of generic words and post-modification to facilitate comprehension, 3. Familiar abbreviations are simply translated into Arabic complete forms. 4. Clipped forms usually recover their complete forms in Arabic translation. 5. Blended forms are usually unpacked and subsequently translated into Arabic complete forms. 6. Editorial decisions by a magazine may sometimes override mainstream tendencies as in the invented Arabic acronyms ال اfor DNA and ال اfor RNA. 7. Quantity-wise, it is clear that the rendition of most RLFs contributes significantly to the verification of the Explicitation Hypothesis (Blum-Kulka, 1986), which argues that explicitation is an inherent feature of translation activity, i.e., the translation is always more explicit than the original. 10. Pedagogical Implications 1. Student translators need to be made aware of the highly technical nature of scientific discourse. Unlike other types of discourse, scientific discourse is full of technical jargon and reduced forms that require a degree of scientific knowledge and/or background on the part of the translator. 2. Student translators need to be familiar with the different categories of lexical reduction in English in general and English scientific discourse in particular. 3. Student translators need to be made aware of the various translation strategies that specialist professional translators employ when rendering RLFs in scientific discourse. 237 )(Arabic Articles Appendix 1 الحقيقة الناصعة للاات ينا جاصلو سا .ت جا أاا ال ااون و ع ا لقاور رااو .الاجل .20الع وان . 8/0 وليو /أغسطس .2636ص .01 -51 الزراعة في المستقبل .الاجل .20الع وان .5/1دارس/أص ل .2636ص .51-85 بزوغ علم الوب للااتب ا جل شاصولت و تي صي رل .ت جا اات النج و ع ان الحاو .الاجل .21الع وان .32/33وفا /و سا .2665ص5-5 . نجوم تتكون من غيوم للااتب ت و .ت جا خح اواا و ع ان الحاو .الاجل .20الع وان . 8/0وليو/ أغسطس .2636ص .31 -5 هل العولمة تساعد فقراء العالم أم قد تضر بهم؟ للااتب ص صارون .الاجل .22الع وان أغسطس /و ا .2660 ثمن الطفرات الصامتة للااتب شادار و م وت .ت جا دحا اس ا اات و دحا وصس .الاجل 21الع وان .32/33وفا /و سا .2665ص.30-36 . ما الذي يجعلنا بشرا؟ للااتب صوورو .ت جا ا اا ع ال اي و ع ان الحاو . .الاجل .21الع وان .32/33 وفا /و سا .2665ص51-18 . الوقود البيولوجي (الكراسولين) :البديل المتاح للنفط مستقبال للااتب ميوص و و ل .ت جا ف}او العجل و جان خور و ال ح .الاجل .21الع وان .32/33وفا /و سا .2665ص.21-38 . تعزيز قوة اللقاح للااتب كاروون و كول دان .الاجل .20الع وان .0/1دا و /و يو .2636ص.53-12 . حرب نووية إقليمية ،و المعاناة عالمية .للااتب روصوك و توون .الاجل 20.الع وان .0/1دا و /و يو .2636ص. .13-52 أموال حقيقية من عوالم افتراضية .للااتب مياس .الاجل 20.الع وان .0/1دا و /و يو .2636ص.23-35 . التغيير المناخي :تجربة متحكم فيها .للااتب ولسشليا و و امل .الاجل . 20الع وان .8/0وليو/أغسطس .2636 ص.11 -58 . شركات التقانة الحيوية تخطط لتحقيق زراعة مستدامة .للااتب صور ل و فيش و فيشوف و كالين س .لاجل .20 الع وان .5/1دارس/أص ل .2636ص .10-16 تطور المعادن .للااتب مار .ت جا ع ا لقاور عاص و وعي دحفوظ .الاجل .20الع وان .5/1دارس/أص ل .2636ص .01-10 مدى اهمية الضحك .للااتب ا ان .الاجل .21الع وان .32/33وفا /و سا .2665ص..13-20 . )Appendix 2 (English Articles ‘Local Nuclear War’. By Alan Robock and Owen Brian Toon (January 2010). 238 ‘Climate Change: A Controlled Experiment’. By Stan n. Wullschleger and Maya Strahl (March 2010). pp. 78-83. ‘What Makes us Human’. By Katherine S. Pollard (May 2009). pp. 44- 49. ‘Web Science Emerges’. By Nigel Shadbolt and Tim Berners-Lee (October 2008). Pp 76-81. ‘The Naked Truth’. By Nina Jablonski (February 2010) . pp. 42-49. ‘Future Farming’. By Jerry D. Glover, Cindy, M. Cox & John P. Reganold ( April 2010). Pp. 82-89. ‘Does Globalization Help?’. By Pranab Badhan. ( April 2006). pp. 84-91. ‘Cloudy with a Chance of Rain’. By Erick Young (February 2010). pp. 34-41). ‘Grassoline at the Pump’. By George Huber and Bruce Dale (July 2009). pp. 52-59. ‘Boosting Vaccine Power”. By Nathalie Garcon and Michel Goldman (Ocotber 2009). pp. 72-79. ‘Local Nuclear Power’. By Alan Robock and Owen B. Toon (January 2010). pp. 74-81. ‘The Price of Silent Mutations’. By J.V. Chamary and Laurence D. Hurst (June 2009) . pp. 46-53. ‘Real Money from Virtual Worlds’. By Richard Heeks.(January 2010). pp. 68-73). ‘Biotech’s Plans to sustain Agriculture. By Borel, Fischer, Fischhoff and Galindez. (October 2009). pp. 86- 94. ‘Evolution of Minerals’. By Rober Hazen. (March 2010). pp. 58- 65). ‘Laughing Matters’. By Steve Ayan. (April/May 2009). pp. 24-31. 239 Major Problems in Student Translations of English Legal Texts Mohammed Farghal & Abdullah T. Shunnaq Abstract 240 This paper aims to show that the translation of English legal texts into Arabic may involve syntactic, layout, and tenor problems. It demonstrates that MA student translators face serious problems when dealing with syntactic discontinuity in a UN legal document. In particular, they find difficulty in properly incorporating English parenthetical non-finite clauses within Arabic legal discourse because Arabic usually favors finite clauses when rendering them. The translation task also offers evidence for many of the students' inability to deal properly with some layout features such as indentation, which is not used to mark paragraphing in the study text but rather to lay emphasis on key propositions. Finally, tenor features relating to the formality of legal lexis presents itself as another obstacle for many participants in the translation task. 1. Introduction Postgraduate translation students usually receive extensive training in both Gen eral and Technical Translation. Technical materials normally cover specialized literature emanating from different disciplines such as sciences, e.g. chemistry, physics; economics, e.g. accounting, public administration; political science; law; religion, etc. It is commonly taken for granted that every discipline has its own jargon or technical vocabulary. Halliday (1978) talks about Register as a defining feature of a language variety. To him, Register can be defined in terms of Field, Mode and Tenor. In technical materials, it is Field that comes to the fore as a defining feature. Field simply relates to Subject-matter which distinguishes one discipline from another. Apart from content, it is usually form represented by syntax, layout and tenor that guides the language user to the differentiation between one register and another. The translation of technical materials requires special expertise in t he relevant discipline in addition to the general language competence. This expertise presupposes an acquaintance with the content and form. The translator's failure to cope with content would obviously result in mistranslations which could be fatal in real-life situations such as court trials; whereas, his failure to cope with form would most likely result in register failures; hence lack of professionalism and precision. English legal texts exhibit certain patterns which are not found in other technical materials or other general varieties. The draftsmen's incessant effort to externalize intentions in their documents so as to avoid ambiguity inevitably brings about inherent peculiarities of legislative texts. These peculiarities are mainly established forms that are taken in their entirety from the standardized legal register, such as 'In Witness Whereof'. Crystal and Davy (1969: 194) write, "Therefore, much legal writing is by no means spontaneous but is copied directly from 'form books', as they are called, in which established formulae are collected". This being so, legal English exhibits a high degree of linguistic conservatism. 241 The features of legal English, as expounded by Crystal and Davy (p. 213), are many. First, the features of layout, by which attention is drawn to the parts of the documents which are crucial to meaning. Second, the grammatical characteristics such as the chain-like nature of some of the constructions, syntactic discontinuity (Bhatia 1983), and the minimal use of anaphora. Third, the careful interplay between precise and flexible terminology in vocabulary. Finally, the legal register's preservation at all levels of forms which have long been abandoned, such as `hereons' and `hereunders'. 2. The present study The present paper aims to highlight the problematic areas in translating a UN legal document (See Appendix) as encountered by MA translation students on their Comprehensive Examination. The examinees, numbering 13, had already received at least two years of translation training when they sat for the test. They were allowed free access to different reference books during the test. From looking at their translations, it has been noticed that the problematic areas basically fall into three categories: syntax-related problems, layoutrelated problems, and tenor-related problems. Despite the fact that these categories may interrelate and intersect, they are dealt with separately below in the hope of sensitizing the translator to the symmetries and asymmetries that ought to be attended to when engaging in English/Arabic legal translation. 2.1 Syntax The most problematic area for the examinees relates to the handling of the parenthetical material separating the matrix subject 'The General Assembly' from its multi-verbs in the complements, viz., 'invites', 'encourages', 'appeals', `invites', and 'requests', respectively. The ability to cope successfully with syntactic discontinuity, which is a characteristic feature of legal texts, constitutes a key gate to finding one's way smoothly through the text. In the study text, the whole parenthetical material consists in non-finite English clauses. This syntactic choice is often avoided in Arabic in favor of finite clauses (Farghal 2012). Being unaware of this common asymmetry, 7/13 of the examinees explored this option, thus mistakenly rendering non-finite English clauses into erroneous nonfinite Arabic clauses. Observe the data in (1) below: (1) Gravely concerned Commending Noting دع ص ( دع ّ ) ع صلقاا الش مثنية على مالحظة The renderings in (1) above involve formal equivalence; hence their erroneousness. A closer look at the English non-finite clauses in the parenthetical material reveals that they are inherently elliptical, viz., the circumstantial marker 'while' is categorically dropped in English legal texts. By contrast, Arabic does not allow the deletion of the circumstantial marker, thus imposing another constraint on the translator besides the finiteness parameter. This being the case, the translator needs to opt for both a finite and an overtly-marked clause in Arabic. Unfortunately, 5/13 of the examinees were unaware of the second 242 constraint, thus rendering the non-finite English clauses into unmarked finite clauses in Arabic. Observe the data in (2) below: (2) Gravely concerned وتع ع صلقاا الش Commending وتةن على Noting وتًاظ Noting with great appreciation وتق ّ ر It should be noted that the actions alluded to in the parenthetical material in the English text are all Perfective, whereas those called for in the matrix material are all Imperfective. The translator’s unawareness of the second constraint results in a serious distortion of the document in Arabic; that is, all actions, whether they be in the parenthetical or the matrix material, will be Imperfective. Apparently, the aforementioned structural asymmetries between English and Arabic have led 12/13 of the examinees to wrong syntactic choices which have serious repercussions on both the layout and the cohesion of the text. On the one hand, the flow and the integrity of the text are broken by the replacement of commas, for example, by full-stops. To illustrate, the shift from non-finite to finite clauses usually entails a shift from commas to full-stops, thus producing different textualizations. On the other hand, the failure to utilize a cohesive -tie may affect both the cohesion as well as the coherence of the text. To illustrate, the covertness vs. the overtness of the circumstantial marker in the English non-finite and the Arabic finite clauses in the parenthetical material plays a key role in the cohesion and coherence of the legal text under study. Consequently, a successful handling of parenthetical materials, which are characteristically nonfinite clauses in English, is an indispensable skill when engaging in English-into-Arabic legal translation. 2.2 Layout Layout refers to the sketch or plan of the text's physical appearance. Basically, this relates to paragraphing, indentation, and graphitic choices, viz., capitalizing, italicizing, underlining and bold-typing. On the one hand, these features are sometimes governed by language-specific constraints such as the standards of paragraphing and capitalizing in English. Thus, for example, capital initials are required in a variety of linguistically-defined phenomena, viz., sentence beginnings, proper names, names of rivers, lakes, organizations, etc., among other things. These linguistically-induced capitalizing rules are non-existent in Arabic; and, being language-specific, have no bearing on the process of translating. Layout features, on the other hand, can be of significance in the text, i.e., their employment will affect the meaning of the text; hence they are relevant to the process of translating. The translator needs to be aware of the employment of significant layout features in technical texts in general and legal texts in particular. The failure to do so may affect both the cohesion and coherence of the text. By way of illustration, the UN document on hand constitutes, formally as well as syntactically, a single sentence with only one full-stop appearing at the end of it. The splitting-up of this sentence into many sentences in the target 243 language affects the linguistic packaging of information in the text alongside the textual and visual coherence it encapsulates, viz., the unbroken format. Being unaware of this important feature in legal texts, 10 out of the 13 examinees in this study used many full-stops in their translations, thus interrupting the structural and semantic flow of the text. To illustrate, observe the example in (3) along with its erroneous rendering by one of the examinees in (4) below: (3) The General Assembly, Recalling its resolution 35/206 N of 16 December 1980, Gravely concerned about the inhuman oppression of millions of women and children under apartheid, … ن الياور ف كا ون260/11 ) الجاعي العاد تشي إلى ص ارما رص5( كما تعبّر عن قلقها الشديد تجاه االضطهاد الالإنساني.1980 أول ... ،لماليين النساء واألطفال تحت نظام الفصل العنصري Other things being equal, the replacement of the comma in (3) by the full -stop in (4) above disrupts the flow of the text. To appreciate this point, observe the appropriate rendering of (3) in (5) below: ) إن الجاعي العاد1( كا ون30 ن الاؤرخ ف260/11 إذ تشي إلى ص ارما 3586 األول وإذ ساورما القلق الش إرا اوضطااو الًإ سا لاً ي ... النسا واألطفال ف ظل الفيل العني As can be observed, (5) above preserves punctuation and indentation as two layout features. However, it fails to relay the emphasis thoroughly on the matrix subject, i.e., The General Assembly, and its subsequent verbals, i.e., Recalling and Gravely concerned, which is obtained not only by indentation, but also by italics. In typewritten translations, the emphasis obtained by italics can be readily relayed in Arabic by italics or bold-typing; but, in handwritten translations, this emphasis can only be relayed by underlining. Unfortunately, none of the examinees explored this option, thus failing to relay the role of italics in the original. To capture this role, therefore, (5) needs to be rewritten as (6) below (Assuming that it is handwritten rather than typewritten): ،إن الجمعية العامة كا ون30 ن الاؤرخ ف260/11 إذ تشير إلى قرارها 3586 األول لاً ي 244 إرا اوضطااو الًإ سا وإذ ساوما القلق الش )6( ... النسا واألطفال ف ظل الفيل العني Clearly, (6) above relays all significant layout features, viz., punctuation, indenta tion and italics, thus securing optimal layout equivalence in the translation of legal texts. 2.3 Tenor Tenor is one of the three aspects that define Register, the other two being Field and Mode (for example, see Halliday & Hasan (1985: 29-44)). Tenor refers to the relations among the text producers and receivers, especially the level of formality, viz., colloquial, standard, formal, highly formal, etc. As a genre of technical materials, legal texts exhibit their own features of formal styles at various linguistic levels in both English and Arabic. We have already looked at syntactic and layout features that are problematic to student translators of English legal texts into Arabic. Let us now look at Tenor as it relates to the translator’s lexical choice, which is of direct impact on the Tenor of legal texts. Legal texts are characterized by the employment of a highly formal style relating to all word classes, that is, verbs, nouns, adjectives and particles. To illustrate the importance of the Tenor problem, observe the data in (7) below: تدعو/تطلب من (7) a. Appeals to all Governments حقوق/ وض/حالة/مشكلة/مأساة b. the plight of women and children c. generous contributions مساهمات كريمة/مساهمات كبيرة تحت d. under Apartheid First, in (7a), 5/13 of the examinees had tenor problems when they rendered the highly formal 'Appeals' by less formal Arabic verbs. Second, in (7b), the tenor problem was more serious in that 9/13 of the examinees rendered the highly formal 'plight' by less formal Arabic nouns. Third, in (7c), 6/13 of the examinees rendered the highly formal 'generous' in "generous contributions' by less formal Arabic adjectives. Finally, 7/13 of the examinees 245 had tenor problems relating to the highly formal preposition 'under' in 'under Apartheid' when they rendered it by the much less formal preposition. To appreciate the tenor problems, observe the appropriate rendering of the boldfaced items in (7) above in (8) below, respectively: (8) a. تناش b. دحن c. وخي d. ف ظل Tenor problems, it should be noted, don't greatly affect the propositional content of the text; rather, they result in register problems, causing lack of professionalism and precision; hence their importance in legal translation. 3. Conclusion The present study has demonstrated that the translation of English legal texts into Arabic involves many syntactic, layout and tenor problems that should be attended to when engaging in this activity. A merely general translating ability would here fall short of employing working and/or convincing legal translations. The fact that translation practitioners in general and student translators in particular need to be sensitized to the special problems encountered in the translating of legal texts becomes of central importance. Thus it should not be taken for granted that a competent translator of general materials is of necessity a competent translator of legal texts, because there are many subtleties relating to syntax, layout and tenor that are peculiar to legal texts. These subtleties need to be brought to the fore in English/Arabic technical translation classes. The paper has shown that Arab student translators of English legal texts stumble when it comes to handling syntactic discontinuity embodied in the parenthetical material. Most of them are also unaware of layout features in legal texts that have serious bearings on the process of translating. Further, a high percentage of them encounter tenor problems in their lexical choice, thus failing to capture the formality parameter which is integral to the translating of English legal texts. Hence, it is of great importance that translation training programs consider problems of syntax, layout and tenor of both SL and TL. More specifically, stu dent translators should be made aware of the main characteristics of the legal text, namely: 1. The extraordinary long sentences of the legal text which are entailed by the presence of lengthy parenthetical material. 2. Indentation doesn't signal paragraphs in the strict sense; rather, it is a layout feature meant for singling out, i.e., emphasizing, key propositions. 3. The heavy use of lexical repetition instead of anaphora, e.g. the key Noun Phrase 'Women and Children' is repeated several times in the English legal text in hand. 4. The initiation of the Arabic text needs an emphatic particle `inna, which is an almost exceptionless tradition in Arabic legal texts. 246 APPENDIX Women and children under Apartheid The General Assembly, Recalling its resolution 35/206 N of 16 December 1980, Gravely concerned about the inhuman oppression of millions of women and children under apartheid, resulting in the killing, detention and torture of schoolchildren protesting against discrimination, the enforced separation of women from their husbands and mass starvation in the reserves, Commending the Special Committee against Apartheid and its Task Force on Women and Children for giving special attention to the plight of women and children under apartheid, Noting the wide observance of 9 August 1981 as the International Day of Solidarity with the Struggle of Women in South Africa and Namibia, Noting with appreciation the establishment of the International Committee of Solidarity with the Struggle of Women in South Africa and Namibia, 1. Invites all Governments and organizations to observe 9 August annually as the International Day of Solidarity with the Struggle of Women in South Africa and Namibia; 2. Encourages the Special Committee against Apartheid to intensify activities in support of women and children oppressed by apartheid, and authorizes it to organize conferences, seminars and missions for this purpose; 247 3. Appeals to all Governments and organizations to provide generous contributions to the projects of the national liberation movements and front-line States for assistance to refugee women and children from South Africa; 4. Invites the co-operation of all Governments and organizations with the Special Committee in promoting solidarity, with and assistance to the women and children of South Africa in their struggle for liberation; 5. Requests the Secretary-General to ensure the closest co-operation by the Centre against Apartheid and the Centre for Social Development and Humanitarian Affairs as well as the Department of Public Information of the Secretariat, with a view to maximum publicity for the plight of women and children under apartheid and their struggle for national liberation. 102nd plenary meeting 17 December 1981 248 ii List of References Abdel-Fattah, M. M. (2005). On the Translation of Modals from English into Arabic and Vice Versa: The Case of Deontic Modality, Babel, 51(1), 31-48. Abedel-Hafiz, A. (2002). Problems in Translating English Journalistic Texts into Arabic: Examples from the Arabic Version of Newsweek. IJAES 3 (1+2): 69-94. Abdel-Hameed, S. (1965). Buloogh Al-Arab fi Al Waw fi Lughat Al- Arab. Cairo: Maktabat al-Kulliyah Al Azhariya. Adonis, A. (1978). The Age of Poetry [in Arabic]. Beirut: Dar Al-Awdah. Aisentadt, E. (1981). Restricted Collocations in English Lexicology and Lexicography. ITL Review of Applied Linguistics 53: 53-61. Al-Afghani, S. (1970). Al-Moojaz fi Qawaaed Al-Lughat Al-Arabia wa shawaa hidiha. Beirut: Dar AlFikr. Al-Batal, M. (1990). Connectives as cohesive elements in a modern expository Arabic text. (ed.) Mushira Eid and John McCarthy, Perspectives on Arabic Linguistics II. Amsterdam / Philadelphia: John Benjamins. Alexander, R. (1978). Fixed Expressions in English: A Linguistic, Psycholinguistic and Dialectic Study. Anglistik und Englischunterricht 6: 171-181. Alexieva, B. (1990). Creativity in Simultaneous Interpretation. Babel 36(1):1-6. Al-Hassnawi, A. (2000). Aspects of Scientific Translation: English into Arabic Translation as a Case Study. www.TranslationDirectory.com . Al-Hayek, I. (1996). An Approximate, Plain, Straightforward Translation of the Honorable Quran in English. Damascus: Dar al-Fikr lil-Tiba'ah wal-Nashr. Al-Hmouz, M. A. (2001). Al-Rasheeed fi al-nahw al-Arabi. Dar Alsafa'. Jordan: Amman. Ali, A. Y. (1934/2006). The Meaning of the Glorious Quran. Vol. 1, Beirut and Cairo: Dar AlKitab al-Lubnani wa Dar Al-Kitab Al-Masri. Al-Jubouri, A. (1987). Computer-aided categorization and quantification of connectives in English and Arabic. Birmingham: University of Aston. iii Al-Jubouri, A. and Knowles, F. (1988). A computer–assisted study of cohesion based on English and Arabic corpora. Proceedings of the 12th International ALLC Congress, Geneva, Geneva: ALLC. Al-Kharabsheh, A & B. Al-Azzam (2008). Translating the Invisible in the Qur`an. (2008): 1-18. Babel. 54 AlKhfaji, R. (2011). Essays in Arabic Text linguistics and Translation Studies. Amman: Amwaj for Publication and Distribution. Almanna, A. (2013). Quality in the Translation of Narrative Fictional Texts from Arabic into English for the Purposes of Publication: Towards a Systematic Approach to (Self-) assessing the Translation Process. Unpublished Ph.D thesis: University of Durham. Al-Najjar, M. F. (1984). Translation as a Correlative of Meaning. Bloomington: Indiana University. [Doctoral Dissertation] Al-Masri, H. (2004). Semantics and Cultural Losses in the Translation of Literary Texts, unpublished Ph.D thesis: Purdue University. Alpert, M. Translatabilty. In Mona Baker (ed.), Routledge Encyclopedia of Translation Studies. London & New York: Routledge, 2001 (online pdf file pp. 273-276). Al-Qinai, J. (2007). Abbreviation and Acronymy in English Arabic Translation. Meta 52(2): 376-389. Al-Qinai, J. (2008). On Translating Modals. Translation and Interpreting Studies, 3(1&2), 30-67. Al-Udhari, A. (trans.). (1984). Victims of a Map. London: The Thedford Press. Al-Wahy, A. Idiomatic False Friends in English and Modern Standard Arabic. Babel 55 (2): 101-123. Ameka, F.(1992). “Interjections: The Universal yet Neglected Part of Speech”. Journal of Pragmatics 18:101-118. Arberry, A. J. (1955/1996). The Koran Interpreted. London: George Allen & Unwin. Arndt, H. and Janney, R. (1985). “Politeness Revisited: Cross-Modal Supportive Strategies”. IRAL 23(4):281-200. Anees, I. (1966). Min Asrar Al-Lugha. Cairo: Anglo-Egyptian Bookshop. iv Asad, M. The Message of the Qur’an. England: The Book Foundation, 1980/2003. Austin, J. (1962). How to Do Things with Words. Oxford: Clarendon Press. Awad, R. (1985). The Beginning and the End. Cairo: American University in Cairo Press. Translation of N. Mahfouz novel bidaayatun wa-nihaayah (1942-43). Aziz, Y. (1999). Cross-cultural translation and Ideological Shifts. International Journal of Translation 11(1 and 2): 29-84. Badran, D. (2001). Modality and Ideology in Translated Political Texts. Nottingham Linguistic Circular, 16: 47-61. Baker, M. (1992). In Other Words: A Coursebook on Translation. London & New York: Routledge. Baker, M. and M. McCarthy (1987). Multi-word Units and Things Like That. (MS), English Language Research, University of Birmingham. Bankole, Adetola. (2006). Dealing with Abbreviations in Translation. Translation Journal 10(4): 1-11. Bayor, D. (1987). Translator/Reader. Translation Review. (23): 30-33. Bell, A., Brenier, J. M., Gregory, M., Girand, C., and Jurafsky, D. (2009). Predictability effects on durations of function and content words in conversational English. Journal of Memory and Language, 60, 92-111. Bell, R. T. (1991). Translation and Translating: Theory and Practice. London and New York: Longman. Benson, M. et al. (1987). The BBI Combinatory Dictionary of English. Amsterdam/Philadelphia: John Benjamins. Bentahila, A. and Davies, E. (1989). “Culture and Language Use: a Problem for Foreign Language v Teaching”. IRAL 27(2):99-112. Bhatia, V. (1983). An Applied Discourse Analysis of English Legislative Writing. University of Aston, Birmingham. Blum-Kulka, Sh. (1986/2004). Shifts of Cohesion and Coherence in Translation. In Venuti, L. (ed.), The Translation Studies Reader (2nd ed.), 290-305. London and New York : Routledge. Bolinger, D. (1952). Linear Modification. F. Householder, ed. Syntactic Theory 1. Harmondsworth: Penguin Education, 1952. 31-50. Bolinger, D. (1963). The Uniqueness of the Word. Lingua 12. 113-136. Bolinger, D. (1980). Language: The Loaded Weapon. London and New York: Longman. Boud, D., Keogh, R. and Walker, D. (1985). Reflection, Turning Experience into Learning. London & New York: Routledge. Boulton, M. (1977). The Anatomy of Poetry. London: Routledge and Kegan Paul. Brown, H. Douglas (2000). Principles of Language Learning and Teaching. New York: Longman. Brown, P. and Levinson, S. (1987). Politeness : Some Universals in Language Usage. Cambridge : Cambridge University Press. Burke, C. L. (1981). “Preparing Elementary Teachers to Teach Reading”. In K. S. Goodman (ed.), Miscue Analysis: Application to Reading Instructions. Urbana [1]: ERIC/RCS, NCTE, pp. 15-30. Burkhanov, I. (2003). Translation: Theoretical Prerequisites. Wydawnictwo Uniwersytetu Rzeszowskiego Rzeszow. Canale, M. (1983). From Communicative Competence to Communicative Pedagogy. In: J. Richards and R. Schmid (eds.), Language and Communication. London: Longman, 2-27. Candlin, C., Crompton, P. & Hatim, B. (2015). Academic Writing Step by Step: A Research-based Approach. London: Equinox vi Cantarino, V. (1976). The syntax of modern Arabic in prose. Bloomington: Indian University Press. Cantarino, V. (1975). The syntax of modern Arabic in prose. Bloomington: Indian University Press. Cantarino, V. (1974). The syntax of modern Arabic in prose. Bloomington: Indian University Press. Carell, P. (1983). "Some Issues in Studying the Role of Schemata, or Background Knowledge in Second Language Comprehension". Reading in A Foreign Language 1 (2) pp. 81-92. Catford, J. C. (1965). A Linguistic theory of Translation: An essay in Applied Linguistics. Oxford: Oxford University Press. Cauer, P. (1896). Die Kunts des Ubersetzens. Weidmann: Berlin. Chafe, W. (1970). Meaning and the Structure of Language. Chicago: University of Chicago Press. Chomsky, N. (1964). Aspects of the Theory of Syntax. Cambridge, Mass.: MIT Press. Chomsky, Noam. (1968). Language and Mind. New York: Harcourt, Brace and Jovanovich. Chomsky, Noam. (1977). On Wh-movement. In P.W. Culicover et al.(eds.), Formal Syntax. Academic Press, 71-132. Chung, C. and Nebaker, J. P. (2007). The psychological functions of function words. Social Communication, 343-359. Coates, J. (1983). The Semantics of the Modal Auxiliaries, London: Croom Helm. Cole, P. & Morgan, J.L. (eds). (1975). Syntax and Semantics: Speech Acts. New York: New York Academic Press. Collins, W. (1860/2010). The Woman in White. Penguin Books. Translated into Arabic by Dar Al-Bihar, vii Beirut, 2003. Cowie, A. (1981). The Treatment of Collocations and Idioms in Learners Dictionaries. Applied Linguistics 2 (3): 223-235. Cowie, A. (1988). Stable and Creative Aspects of Vocabulary Use. In: M. Carter and M. McCarthy, Vocabulary and Language Teaching, 126-139. London: Longman. Cruse, A. (1986). Lexical Semantics. Cambridge: Cambridge University Press. Crystal, D. and Davy, D. 1969. Investigating English Style. London: Longman. pp.191-217. Halliday, M. 1978. Language as Social Semiotic. London: Edward Arnold. pp.31-35. Dagut, M. (1981). Semantic “Voids” as a Problem in the Translation Process. Poetics Today 2: 61-71. Darwish, Ali. (2005). English Verticality in “Square” Arabic Translation. Translation Monitor 2(1):1-17. Davies, E. (1982). “A Contrastive Approach to the Analysis of Politeness Formulas”. Applied Linguistics 8(1):75-88. de Beaugrande, R and Dressler, W. (1981). Introduction to Textlinguistics. London: Longman. de Haan, F. (1997). The Interaction of Modality and Negation: A Typological Study, New York: Garland Publishing. de Waard, J. and Eugene N. (1986). From One Language to Another: Functional Equivalence in Bible Translating. Nashville: Thomas Nelson Publishers. Dworzynski, K., Howell, P., Au-Yeung, J. & Rommel, D. (2004). Stuttering on function and content words across age groups of German speakers who stutter. Journal Multilingual Communication Disorder, 2 (2), 81-101. viii El-Hassan, S. (1990). Expressing Modality in English and Standard Arabic. Journal King Saudi University, 2 (2), 149-166. El-Yasin, M. K. (1996). "The Passive Voice: A Problem for the English-Arabic Translator". BABEL 42 (1). Durkin, Dolores. (1978). Teaching Them to Read. London: Alyn and Bacon Inc. Elyas, A. E. (1987). The Thief and the Dogs (An English Translation of Najeeb Mahfouz’s alliṣṣu wa al-kilaab, 1973, Beirut Dar Al-qalam). Jeddah: Dar Al-Shorouq. El-Yasin, M. K. (1996). The Passive Voice: A Problem for the English-Arabic Translator. BABEL 42 (1): 73-80. Encarta ® 97 Encyclopedia © (1993-1996). Reading Activity. Microsoft Corporation. Fareh, S. (1998). The functions of AND and Wa in English and Arabic written discourse. Papers and Studies in Contrastive Linguistics, 34, 303-312. Farhan, Z. and Fannoush, T. (2005). Difficulties of translating discourse markers from English into Arabic. ADAB AL-RAFIDAYN, 4, 1-27. Farghal, M.. (1991). Evaluativeness Parameter and the Translator from English to Arabic and Vice-versa. Babel 37:138-151. Farghal, M. (1993a). Managing in Translation: A Theoretical Model. Meta (38)2: 257-267. Farghal, Mohammed. (1992b). The Arabic Topic-Comment Structure. Journal of King Saud University 4(1): 47-62. Farghal, M. (1993a). The Translation of Arabic Cognate Accusatives into English. Tujuman 2(2):66-92. Farghal, M.. (1993b). The Translation of Arabic Cognate Accusatives into English by MA Student Translators. Interface 8: 25-41. ix Farghal, M. (1994). Ideational Equivalence in Translation. In: Beaugrande, Robert, et al. (eds.), Language, Discourse and Translation in the West and Middle East. Amsterdam & Philadelphia: John Benjamins, 55-63. Farghal, M. (1995a). Lexical and Discoursal Problems in English-Arabic Translation. META 40 (1): 54-61. Farghal, M.. (1995b). The Pragmatics of 'inšaallah in Jordanian Arabic. Multilingua 14 (3): 253-270. Farghal, Mohammed. (2003). Schemata and Lexis in Translation. (2003). Perspectives: Studies in Translatolgy 11(2):125-133. Farghal, M.. (2004).Two Thirds of a Boy are his Uncle’s: The Question of Relevance in Translation. (2004). Across Languages and Cultures 5(2): 257-269. Farghal, M. (2009). Al-Tariq (Arabic translation of The Road by C. MacCarthy, 2006, Penguisn Books). Ibda'at Alaamiyah [World Masterpieces], 380, Kuwait National Council for Culture, Arts, and Literatures, Kuwait, pp. 269. Farghal, M. (2012). Advanced Issues in Arabic-English Translation Studies. Kuwait: Kuwait University Press. Farghal, M. (2013). Khraa'it (Arabic translation of Maps by N. Farah, 1986, Penguin Books), Ibda'at Alaamiyah [World Masterpieces], 395, Kuwait National Council for Culture, Arts, and Literatures, Kuwait, pp. 433. Farghal, M. and A. Borini. 1997. Pragmareligious Failure in Translating Arabic Politeness Formulas into English: Evidence from Mahfouz’s Awl a d H a ritna. Multilingua 16(1). 77-99. Farghal, M. and A. Shakir (1992). The Interpretation of Advertisements by English native and NonNative Speakers As Viewed from a Schema-Theoretic Perspective. Abhath Al-Yarmouk, 10 (1): 21-41. Farghal, M. and H. Obeidat. (1995). Collocations: A Neglected Variable in EFL. IRAL 33 (4): 315-331. Farghal, M. and M. Al-Hamly. (2007). Lexical Collocations in EFL Writing. The Journal of Asia TEFL 4(1): 69-94. Feder, Marcin. (2003). Machine-Assisted Human Translation: Its Position. Perspectives: Studies In Translatology11(2):135-143. x Fillmore, Charles J. 1968. The Case for Case. In E. Bach and R. Harms (eds.), Universals in Linguistic Theory. New York: Holt, Rinehart and Winston, 1-88. Fowler, R. (1996). Linguistic Criticism. Oxford University Press, USA. Geertz, C. (1973). The Interpretation of Cultures. New York: Harper Colophon, Basic Books. Georges, A. & Barakt, G. (2000). A linguistic Approach towards the Teaching of English Grammar. Aleppo University Journal of Research 38, Syria. Giles, D. (1995). Basic Concepts and Models for Interpreter and Translator Training. Amsterdam: John Benjamins. González, F. R. (1991). Translation and Borrowing of Acronyms: Main Trends. IRAL XXIX(2): 161-170. Goodman K.S. (1962). A Linguistic study of Cues and Miscues in Reading. Elementary English 42: 639-660. Grimes, J. (1976). The Thread of Discourse. Mouton: The Hague, Paris. Grice, H.P. (1975). Logic and Conversation. In Cole & Morgan, PP. 41-58. Gust, E. A. (1996). Implicit Information in Literary Translation: A Schema-Theoretic Perspective. Target 8(2):239-256. Gust, E. A. (1991). Translation and Relevance: Cognition and Context. Oxford: Blackwell. Halliday, M. A. K. (1970). Functional Diversity in Language as seen from a Consideration of Modality and Mood in English. Foundations of Language. International Journal of Language and Philosophy 6: 322-61. Halliday, M.A.K and Hassan, R. (1976). Cohesion in English. London: Longman Group Ltd. Halliday, M. and Hasan, R. 1989. Language, Context, and Text: Aspects of Language in a Social-Semiotic Perspective. Oxford University Press. Hamad, A. and Zu'bi, Y. (1984). Al-Mu'jam Al Wafi fi Al Nahu Al-Arabi. Amman: Arts and Culture Department. Hamdan, J. and Fareh, S. (1999). The translation of Arabic wa into English: some problems & implications. Dirasat: Human and Social Science, 26 (2), 590-603. Hatim, B. (1984). "A Textlinguistic Model for the Analysis of Discourse Errors: Contributions from Arabic Linguistics", in Monaghan, J. (ed.) Grammar in the Construction of Texts. Prentice-Hall. xi Hatim, B. (1997a). Communication Across Cultures: Translation Theory and Contrastive Text Linguistics. Exeter: University of Exeter Press. Hatim, B. (1997b). The Translator as Communicator. London and New York: Routledge. Hatim, B. (2012). Teaching and Researching Translation. London: Longman. Hatim, B. and I. Mason. (1990). Discourse and the Translator. London and New York: Longman. Hemingway, E. (1952). The Old Man and the Sea (translated into Arabic al-shaykh wa al-baHr by Munir Ba’labki, 1985, Beirut: Dar Al-Malayeen). Penguin books. Hendrickson, J. M. (1980). Error Correction in Foreign Language Teaching: Recent Theory, Research, and Practice. In: K. Croft (ed.), Readings on English as a Second Language. Cambridge, MA: Winthorp Publishers. Hickey, L. (2003). "The Reader as Translation Assessor". Studies in Translation 1(1): 59-92. Holes, C. (1984). A Textual Approximation in the Teaching of Writing to Arab Students: A contrastive Approach, in Swales, J. and Mustapha, H. (eds.) English for Specific Purposes in the Arab World. pp. 228-243. Holes, C. (1995). Modern Arabic: Structures, Functions, and Varieties. London: Longman. Hönig, H. G. (1998). Positions, Power and Practice: Functionalist Approaches and Translation Quality Assessment. In: C. Schäffner (ed.), Translation and Quality. Philadelphia: Multilingual Matters, 6-34. Hosseini, K. (2003). The Kite Runner. USA: Riverhead Books. Trans. into Arabic as ʻadaa'u aṭṭaa'iratu al-waraqiyyah by M. Fayyadh (2010). Damascus: Dar Al-Nashr watTawzee. House, J. (1977). A Model for Translation Quality Assessment. Tubingen: Narr. House, J. (1997). Translation Quality Assessment: A Model Revisited. Tubingen: Narr. Howell, P., Au-Yeung, P., and Sackin, S. (1999). Exchange of stuttering from function words to content words with age. Journal of Speech, Language, and Hearing Research, 42, 345-354. xii Hutchins, W. John (1991). Why Computers Do Not Translate Better. Paper presented at the conference on Translating & the Computer 13, London. Hutchins, W. John (2001). Machine translation and human translation: in competition or in complementation?. International Journal of Translation 13:5-20. Hymes, D. (1972). Models of the Interaction of Language and Social Life. In: Gumperz , J. and D. Hymes, Directions in Sociolinguistics. New York: Holt, Rinehart & Winston, 35-71. Ibn-Hishaam, Jamal Al-Deen Ibn Hishaam Ansaari (d. 761h) (2002). Mughni Al-Labiib ʻan Kutubbi AL-ghaariib, (ed.). Al-Khatib, A. L. Kuwait: Al-Turaath Al-Arabi. Illyyan, M. (1990). Transitional words and the Arab translator. Irbid-Jordan: University. [MA Thesis]. Yarmouk Ilyas, A. (1989). Theories of Translation: Theoretical Issues and Practical Implications. Mosul: University of Mosul. Irving, T. B. (1985/1993). The Qur’an: The Noble Reading. Iowa: The Mother Mosque Foundation. Ismail, A. (1972). Modern Arabic Poetry [in Arabic]. Beirut: Dar Al-Awdah and Dar Al-Fikr. Ivir, V. (1977). Lexical Gaps: A Contrastive View. Studia Romanica et Anglica Zagrabiensia 43:167176. Ivir, V. (1981). Formal Correspondence vs. Translation Equivalence Revisited. Poetics Today 2(4). 5159. Ivir, V. (1991). Procedures and Strategies for the Translation of Culture. International Journal of Translation 1&2:48-60. Jakobson, R. (1959). On Linguistic Aspects of Translation. In Brower, R. (ed.) On Translation. Harvard University Press, pp. 232-239. Johnstone, B. (1990). Orality and Discourse Structure in Modern Standard Arabic. In Eid, M. (ed.) Perspectives on Arabic Linguistics I, 215-233. Amsterdam/ Philadelphia: John Benjamins, Johnstone, B. (1991). Repetition in Arabic Discourse. Amsterdam: John Benjamins. xiii Kachru, Y. (1982). Toward Defining the Notion ‘Equivalence’ in Contrastive Analysis. TESL 5: 82-98. Kamal, A. H. (1971). Huroof Al-Ma'ani. Taif: Al-Ma'rif Bookshop. Kammensj, H. (1993). Connectives in MSA and/or ESA: Suggestion for Research. Gothenburg: University of Gothenburg, Department of Oriental Languages. Kaplan, R. (1966). Cultural Thought Patterns in Intercultural Education. Language Learning 16 (1+2): 120. Kaplan, R. (1982). Contrastive Rhetoric Revisited. Paper Presented at the 1982 TESOL Conference, Honolulu. Karin, C. R. (2005). A Reference Grammar of Modern Standard Arabic. Cambridge: Cambridge University Press. Khafaji, R. (1996). Arabic translation Alternatives for the Passive in English. Papers and Studies in Contrastive Linguistics 31: 19-37. Khalil, A. (1993). Arabic Translation of English Passive Sentences: Problems and Acceptability Judgments. Papers and Studies in Contrastive Linguistics 27: 169-181. Khalil, A. M. (1999). A Contrastive Grammar of English & Arabic. Jordan: Jordan Book Center. Khalil, E. N. (2000). Grounding in English and Arabic News Discourse. Amsterdam: Benjamins. Khan, M. and Hillali, T. (1999). The Noble Qur’an: Interpretation of the Meanings of the Noble Qur’an in the English Language. Saudi Arabia: Darussalan publishers and Distributors. Koller, W. (1978). Kritik der Theorie der Ubersetzungskirik. IRAL xvi (2): 89-107. xiv Komissarov, V. (1987). The Semantic and the Cognitive: A Problem in Equivalence. META XXXII , 4: 416-420. Krazeszowki, T. (1971). Equivalence, Congruence and Deep Structure. In G. Nickel (ed.), Papers in Contrastive Linguistics. Cambridge: Cambridge University Press, pp. 37-48. Kreidler, Charles, W. (1998). Introducing English Semantics. London and New York: Routledge. Kress, G. (1982). Learning to Write. London: Routledge and Kegan: Paul. Kroll, B. (1984) "Writing for Readers: Three Perspectives on Audience". College Composition and communication. 35 (2) pp. 172-185. Kryk, B. (1992). “The Pragmatics of Interjection: The Case of Polish no”. Journal of Pragmatics 18:193207. Larson, M. L. (1998). Meaning-Based Translation: A Guide to Cross-Language Equivalence. Larham: University Press of America. Lederrer, M. (1978). Simultaneous Interpretation - Units of Meaning and Other Features. In Gerver, D. & Sinaiko, H. (eds.) Language Interpretation and Communication. Plenum press, New York. pp. 323-332. Leech, G. (1983). The Principles of Pragmatics. London : Longman. Lemke, J. L. (1985). “Ideology, Intertextuality and the Notion of Register.” In J. D. Benson and W.S. Greaves (eds.), Systemic Perspectives on Discourse, Vol.1, Norwood, N. J.: Ablex Publishing Corporation, pp. 275-294. Le Ny, J. (1978). Psychosemantics and Simultaneous Interpretation. In Gerver & Sinaiko (eds.), pp. 289298. xv Levi, J. (1967). Translation as a Decision Process. In: To Honor Roman Jakobson. The Hague. pp. 11711182. Levinson, S. (1983). Pragmatics. Cambridge: Cambridge University Press. Lewis, C. Day. (1969). The Poetic Image. London: Jonathan Cape. Lewis, A. L. (1997). The Practical Implications of a Minimum Machine Translation Unit. BABEL 43(2):138-150. Lyons, J. (1977). Semantics. Cambrdge: Cambridge University Press. Mahfouz, N. (1959). Awlad Haritna. Beirut : Dar Al-Adaab. Martin, J. R. (1985). Factual Writing: Exploring and Challenging Social Reality. Victoria: Deakin University Press. McCarthy, M. (1991). Discourse Analysis for Language Teachers. Cambridge: Cambridge University Press. Meijer, Siety. (1993). Attitudes towards Machine Translation. Language International 5(6):11-13. Mitchell, T. (1971). Linguistic 'goings-on' : Collocations and Other Lexical Matters Arising on the Syntagmatic Record. Archivum Linguisticum 2:35-69. Moffette, J. (1968). Teaching the Universe of Discourse. New York: Harvard University Press. Montoro, K. B. (1976). “Syntactic Acceptability, Syntactic Change, and Transformation”. In P. David Allen and Doroth J. Watson (eds.), Findings of Research in Miscue Analysis: Classroom Implications. Illinois: Library of Congress pp. 34-38. Mouakket, A. (1986). Linguistics and Translation: Some Semantic Problems in Arabic-English xvi Translation. Washington, D.C.: Georgetown University. [Doctoral Dissertation]. Munif, A. (1992). mudin-il-malh: taqaasiim-il-layl wan-n-nahaar (Cities of Salt; Variations on Day and Night). Beirut: Arab Corporation for Studies and Publishing. Mustafa, H. “Quran (Koran) Translation”. In: M. Baker (ed.), Routledge Encyclopedia of Translation Studies, pp. 200-2004. New York: Routledge, 2001. Nattinger, J. (1980). A Lexical Grammar for ESL. TESOL Quarterly 14: 337-344. Nattinger, J. (1988). Some Current Issues in Vocabulary Teaching. In: M. Carter and M. McCarthy, Vocabulary and Language Teaching, 62-82. London: Longman. Neubert, A. and Gregory M. Shreve (1992). Translation as Text. Kent: The Kent state University Press Newmark, P. (1981). Approaches to Translation. Pergamon Press: London. Newmark, P. (1988). A Textbook of Translation. London: Prentice Hall. Newmark, Peter. (1982). The Translation of Authoritative Statements META 27(4). Newmark, P. (1991). About Translation. London: Multilingual Matters. Nida, Eugene. (1964). Toward a Science of Translating with Special Reference to Principles and Procedures in Bible Translating. Leiden: E.J. Brill. Nida, E.(1969). Science of Translation. Language 45: 483-498. Nida, E. 1994. Translation: Possible and Impossible. turjuman 3(2): 147-163. Nida, E. (1997). The Nature of Dynamic Equivalence in Translating. BABEL 13: 99-103. Nida, E. & W. D. Reyburn (1981). Meaning Across Cultures. New York: Orbis Books. xvii Nida, E. & Ch. Taber (1982). The Theory and Practice of Translation. Leiden: E.J. Brill. Obeidat, H. and Farghal, M. (1994). On the Status of the Equational Sentence in the Grammar of Arabic. Abhath Al-Yarmouk 12(2): 9-35. O’Shea, B. (1996). Equivalence in Danish News Subtitling from Northern Ireland. Perspectives 4(2). 235-254. PACTE Group. (2000). Acquiring Translation Competence: Hypotheses and Methodological Problems of a research Project. In A. Beeby, et al. (eds.), Investigating Translation. Amsterdam: John Benjamins. Page, W. D. (1976). “Semantic Acceptability and Semantic Change”. In P. David Allen and Doroth J. Watson (eds.), Findings of Research in Miscue Analysis: Classroom Implications. Illinois: Library of Congress pp. 38-44. Page, W. D. (1985). “The Author and the Reader in Writing and Reading”. Research in the Teaching of English 8: 170-183. Palmer, F. (1986). Mood and Modality. Cambridge: Cambridge University Press. Pergnier, M. (1978). Language Meaning and Message Meaning: Towards Sociolinguistic Approach to Translation. In Gerver & Sinaiko (eds.), 199-204. Perkins, M. (1983). Modal Expressions in English. London: Frances Pinter. Picthall, M. (1930/2006). The Glorious Koran. London: George Allen & Unwin. Pickering, J. and Jeffry D. H. (1982). Literature. New York: MacMillan. Pym, A. (1992). Translation and Text Transfer. Frankfurt am Main: Peter Lang. Pym, A. (2005). Explaining Explicitation. Tinet,4, 29-43, retrieved from www.tinet.cat /~apym/online/translation/explicitation _web.pdf . Quirk, R. et al. (1986). A Comprehensive Grammar of the English Language. Essex: London Group Ltd. Rabin, C. (1958). The Linguistics of Translation. In: Smith, A. H. (ed.), Aspects of Translation. London: Secker & Warburgy, 123-145. xviii Reiss, R. (1981). "Type, Kind and Individuality of Text: Decision Making in Translation". Poetics Today 24: 121-131. Renkema, J. (1993). Discourse Studies. Amesterdam/Philadelphia: John Benjamins. Rousch, Peter D. (1976). “Semantic Word Relationships”. In P. David Allen and Doroth J. Watson (eds.), Findings of Research in Miscue Analysis: Classroom Implications. Illinois: Library of Congress, pp. 52-55. Sa'deddin, M. (1987). Three Problem Areas in Teaching Translating to native Arabic Literates. Anthropological Linguistics 29 (2), pp. 181-192. Saedi, L. (1990). "Discourse Analysis and the Problem Translation Equivalence". META XXXV (2): 289297. Saedi, L. (1996). Translation Principles vs. Translator Strategies. Meta 4(3): 389-392. Saeed, A. T. and Fareh, S. (2006). Difficulties encountered by bilingual Arab learners in translating Arabic 'fa' into English. The International Journal of Bilingual Education and Bilingualism, 9 (1), 19-31. Saldanha, G. (2008). Explicitation Revisited: Bringing the Reader into the Picture. Trans-Kom, 1 (1), 2035, accessed at www.trans-kom.eu. Saraireh, M. (1990). Some lexical and Semantic Problems in English-Arabic Translation. Madison: The University of Wisconsin. [Doctoral Dissertation] Schäffner, C. (2003). Third Ways and New Centres: Ideological Unity or Difference? In: M. CalzadaPerez (ed.), Apropos of Ideology Manchester: St. Jerome, 23-42. xix Schäffner, C. (1998). Skopos Theory. In: M. Baker (ed.), Routledge Encyclopedia of Translation Studies. London: Routledge, 235-238. Schiffring, D. (1987). Discourse Markers. Cambridge: Cambridge University Press. Schön, D. (1983). The Reflective Practitioner. How professionals think in action, London: Temple Smith. Searle, J. R. (1981). Expression and Meaning. Cambridge : Cambridge University Press. Seleskovitch, D. (1978). Language and Cognition. In Gerver & Sinaiko (eds.), 333-342. Shakir, A. and M. Farghal. (1997). When The Focus of the Text is Blurred: A Textlinguistic Approach for Analyzing Student Interpreters Errors. META 42 (4) 629-640. Sharkas, H. (2009). Translation Quality Assessment of Popular Science Articles: Corpus Study of the Scientific American and its Arabic Version. Trans-kom 2(1): 42-62. Shunnaq, A. T. (1994). ‘Monitoring’ and ‘Managing’ in Radio News Reports. In de Beaugrande , et al. (eds.), Language, Discourse and Translation in the West and Middle East, Amsterdam: John Benjamins, 103-114. Sifiano, M. (1993). “Off-record Indirectness and the Notion of Imposition”. Multilingua 12(1): 69-79. Sims, R. (1976). “What We Know about Dialects and Reading”. In P. David Allen and Doroth J. Watson. (eds.), Findings of Research in Miscue Analysis: Classroom Implications. Illinois: Library of Congress, pp. 128-132. Sinclair, J. (1987). Collocation: A Progress Report. In: R. Steele and T. Threadgol (eds.), Language Topics: Essays in Honour of Michael Halliday, Vol. 3, 319-331. Amsterdam: John Benjamins. xx Sinclair, J. (1991). Corpus, Concordance, Collocation. Oxford: Oxford University Press. Spencer, Richard H. (1989). Translatibilty: Undertranslatibilty and Usabilty by Others. Computers in Human Behavior 4: 347-354. Sperber, D. and D. Wilson. (1986). Relevance. Oxford: Basil Blackwell. Stalker, J. (1989). “Communicative Competence, Pragmatic Functions, and Accommodation”. Applied Linguistics 10(2) : 182 -193. Stewart, P. (1981). Children of Geblawi (Translation of Mahfouz’s awlaad Haaritna). London: Heinemann. Strassler, J. (1982). Idioms in English: A Pragmatic Analysis. Tubingen, Gunter Narr Verlag. Stubs, M. (1983). Discourse Analysis: The Sociolinguistic Analysis of Natural Language. Chicago: Chicago University Press. Swales, J. (1990). Genre Analysis. Cambridge: Cambridge University Press. Tabakowska, E. (1990). Surely There Must Exist a Polish Equivalent: On the Inadequacy of Dictionary Explications. Target 2(2): 199-218. Tahaineh Y. and Tafish R. (2011). Pitfalls encountered by bilingual Arab learners in translating the Arabic discourse marker 'θumma' into English. Theory and Practice in Language Studies, 1 (3): 226-238. Taylor, C. (1998). Language to Language: a practical and theoretical guide for Italian/English translators. Cambridge: Cambridge University Press. xxi Theroux, P. (1993). Cities of Salt: Variations on Night and Day (trans.) . Beirut: Cape Cod Scriveners. Thomas, J. (1983). “Cross-Cultural Pragmatic Failure”. Applied Linguistics 4 (2): 91-112. Tirrkonen-Condit, S. (1989). Theory and Methodology in Translation Research. In: S. Tirrkonen-Condit and S. Condit (eds.), Empirical Studies in Translation and Linguistics, pp. 3-18. University of Joensuu: Joensuu. Van Dalen, D. (1973). Understanding Educational Research. New York: McGraw – Hill Book Company. Vermeer, H. J. (2000). Skopos and Commission in Translational Action (A. Chesterman, Trans.). In L. Venuti (ed.), The translation Studier Reader London: Routledge, 221-232. Uhlenbeck, E. (1978). On the Distinction between Linguistics and Pragmatics. In Gerver & Sinaiko (eds.), 185-198. Widdowson, H. G. 1971. The Deep Structure of Discourse and The Use of Translation. In Corder, S. P. and Roulet, E.(eds.), Linguistic Insights in Applied Linguistics, 61-81. Paris: Didier. Wided, B. (2010). Modality in English, French and Arabic Biomedical Discourse: A Contrastive Study of Drug Information Leaflets. M.A. Dissertation, Mentouri University, Constantine. Wilss, W. (1982). The Science of Translation. Tubingen. Narr. Yorio, C. (1980). Conventionalized Language Forms and the Development of Communicative Competence. TESOL Quarterly 14: 433-442. Yule, George. (1985). The Study of Language. Cambridge: Cambridge University Press. Zayed, S.H. (1984). A Pragmatic Approach to Modality and the Modals: with Application to Literary Arabic. Ph.D. Thesis, University of Edinburgh, UK. xxii xxiii View publication stats