The Freedom of the Translator: Issues in Correspondence and Shift

It is clear that the notion of “shift” plays a primordial role in leading the translator to predetermined targets, which explains why the majority of translated texts from the same sources are largely similar. In this article, we try to explore the different possibilities of choice between competing translations at the different levels of English and Arabic linguistic systems. We also attempt to highlight the important notion of ‘shift’ in translation by showing the different mechanisms that operate in the process of translating English into Arabic in our endeavor to keep meaning as close as possible in both languages. The term “equivalence” is elucidated to underscore its different manifestations at the textual level. In this respect, a brief look at the article “Asymmetric struggle for the hearts and mind of viewers: Can the media actually trigger sympathy towards terrorists?”, written by the Israeli journal-ist Ifat Maoz, gives the reader an idea about how translation is determined by the notion of shift and about the possibilities, if any, the translator may have recourse to for stylistic reasons. We found that a register analysis must be incorporated into the general conceptualization of the flow from the original text to the translated text in order to produce an acceptable translation. We also found that the only shift in translation that provides translators freedom is intra-systemic shift, a subset of class shift, addition shift, omission shift, and meaning shift. Finally, we discovered that expressive meaning is the most influential factor in translation as it may drive translators to depart from the source text to match the context of the target text in order to create a product that is reader-friendly.


Introduction
Translation, as an art as well as a skill and a science (Newmark, 1981), rests on three aspects: the foreign and native cultures, the two languages, and the writer and the translator respectively. "What truly distinguishes translation is that it takes place in the context of the relations between two cultures, two worlds of thought and perception" (Delisle, 1988, p. 74). It is believed that the concept of fidelity or faithfulness can be equated with equivalence as this latter is part and parcel of the very definition of translating or translation. The target text (TT) which does not not comply with the standards of equivalence is not desirable; therefore, word for word translation and extremely literal translation too faithfully reproduce every feature of the original while paraphrasing and adaptations take too many liberties in drifting away from the source text (ST).
The purpose of this article is to investigate the actual meaning and structure shifts that can occur during translation and to identify the various changes that occur during the movement from one language/culture to another. Specifically, we will examine an article entitled "Asymmetric struggle for the hearts and minds of viewers: Can the media actually trigger sympathy towards terrorists?" (Maoz, 2010), which was translated by the Moroccan translator Badia Elharraki in 2012. These alterations are illuminating because they illustrate the relevance of communicative aspects such as "purpose" and "audience," as well as the general backgrounds of the author and the translator, respectively. This article was chosen because it addresses the Israeli audience's understanding of the Israeli-Palestinian conflict in the Middle East during a talk with a Palestinian girl who was captured while trying a suicide attack. The overarching objective is to experiment with the small changes that occur in translation as one conceptual network is replaced by another. Our interest in articles dealing with media perception and analysis necessitated the current study as they provide a testing ground for the relevance of the main factors which are the purpose and the audience in translation. Linking shifts with both the freedom of the translator and register analysis is meant to fill a gap in research and constitutes a rationale for investigating this issue.
Considering that the current research is predominately centered on textual analysis, the technique may be considered descriptive as it is limited to the identification, classification, description, and interpretation of selected sentences from an article translated from English to Arabic. The ability to recognize, identify, and analyze the different types of shifts adopted by the translator as a case study depends on a somewhat subjective comparison, interpretation, and evaluation of the ST and the TT. Hence, this study is qualitative in character as we have tried to explain the shifts adopted by providing selected examples.
In this article, we underscore the pivotal notion of 'shift' in translation by displaying the various mechanisms that contribute to an acceptable translation from English to Arabic in our attempt to preserve the same meaning in both languages. The term "equivalence" will be scrutinized to exhibit its different manifestations at the level of the text. The research question we are mainly concerned with revolves around the issue of whether translators have an array of choices up their sleeve to render faithful meaning or whether they are bound to only one type of translation. In other words, if translators have alternatives, we should be able to characterize the level at which these translation choices are possible. In this regard, an examination of the article "Asymmetric struggle for the hearts and mind of viewers: Can the media actually trigger sympathy towards terrorists?", written by the Israeli journalist Ifat Maoz in 2010, will clarify the thorny issue of shift in translation and enumerate the possibilities available to the translator.

The problem of translating: A general background
When we analyze a text, we should take into account the fact that it has to be evaluated in different ways depending on many parameters and situational conditions that can either be uncontrolled by the writer like the ones related to the pragmatics of time and space or determined by a decision taken prior to the production of the text like the choice of lexis. Then, there are features which are dictated by social norms like text types and genre conventions. Finally, stylistic features are chosen by the writer out of several alternatives to produce a certain effect on the reader.
Given the fact that translation involves two linguistic systems with different structural and grammatical codes, two cultural and social contexts, and two independent communicative domains, opposing forces will pull the translating activity in opposing directions. There is a tension between the intrinsic and communicative aspects, and the semantic and pragmatic ones; thus the translator is under permanent and constant pressure from the conflict between form and meaning.
Most, if not all, translation theories highlight the crucial position of linguistics in building various translation definitions. Newmark (1981, p.5) clarifies the bond between linguistics and translation theory in the following way: "Translation theory derives from comparative linguistics, and within linguistics, it is mainly an aspect of semantics, all questions of semantics relate to translation theory". Different theorists hold different positions in as far as the translation process is concerned. Kelly (1979, p. 34) holds that "each stream of language theory corresponds to a theory of translation. All linguistic schools or trends devoted part of their work to translation problems". The focus has recently shifted towards the relationship between linguistics, culture and translation. First, theories that rely on comparative linguistics and semantics are said to be linguistic (Baker, 1998, p. 118). Second, theories with a communicative orientation are called sociolinguistic. Third, a newer orientation, which is gradually moving from the mainly sentence-oriented approach towards more "discoursal" and macro-textual approaches such as discourse analysis, is gaining momentum. Its aim is to afford the language user with the right means to handle features of language beyond sentence-oriented theories. In this respect, Hatim and Mason (1990, p. 35) state that "translation theory has indeed moved away from its sentence-based stage and has taken a discoursal and macro-textual orientation which not only operates on the textual and discoursal level in both source text analysis and transfer, but also considers translating as a dynamic communicative act, and restores to the translator the central role in a process of cross-cultural communication and ceases to regard equivalence merely as a matter of entities within texts".
At the theoretical level, translation in the past focused on achieving a translation theory based on source texts i.e., the emphasis was on the product rather than the process. Afterwards, translation embraced the era where the critical concept of "equivalence" was systematically scrutinized as in the early developments in linguistics; the art of translating was gradually influenced by new disciplines such as sociology, psychology and communication theory. Many translation theorists, such as Nida, Catford and Newmark have applied current linguistic theories to translation. Whereas Catford based his translation approach on structural linguistic theory and Nida based his translation theory on transformational generative grammar, Newmark adopted a more eclectic approach, borrowing from different linguistic theories. On another level, Hatim and Mason developed a more complete translation theory. Some linguists such as Halliday developed socio-linguistic theories (systemic-functional grammar/the theory of context, text and grammar), which were used by other writers as bases for translation theories.
In her analysis, Elharraki (2022) recommends a submission of both the ST and the TT to a rigorous register analysis where the notions of the field, mode, and tenor are key elements. She also opts for a critical discourse analysis (CDA) of the ST and the TT. Similarly, Bayani (2016) attempts to shed light on the relationship between language and ideology involved in translation generally, and on how Fairclough's (1989) and Halliday's (1985) frameworks are used. Both the ST and the TT are examined to identify the hidden ideological presuppositions in the texts. The study's main objectives are to develop a CDA framework to analyze the ST and the TT, as well as a text-linguistic analysis of the translators' decisions in the TT.
As far as ideology is concerned, Elewa (2019) confirms the fact that sociocultural distinctions affected translation and led academics in translation to investigate various sociocultural viewpoints utilizing a variety of data sources. Using CDA, he examines how Arabic quotations were translated in American and British publications and suggests that translators and editors who work for these newspapers frequently reflect the ideologies of their publishers and/or target readers by contrasting the cultural values, social systems, and ideologies reflected in the linguistic structures and strategies used by both newspapers to present quoted Arabic statements.
From a translation studies viewpoint, Munday (2007) explores key issues relating to ideology and language. His study investigates the concept of "ideology", how it is viewed in translation studies, and where it has mostly been connected to manipulation and power dynamics. This article concentrates on the translator's personal ideology. The main focus is on how ideology in all of its manifestations is expressed and presented textually in translation and how an analysis from monolingual traditions (like CDA and the tools of systemic-functional analysis) may not always be the best to identify and categorize the shifts that occur. Examples of political translations reviewed include speeches, other writings, and interviews with Latin American revolutionary leaders (Castro, Marcos, Chavez).
Every language use is an ideological practice that reflects specific ideologies (Anshori & Sumarlam, 2018). In fact, ideologies aim to influence a specific social group while still maintaining both the power over it and the status quo. Because translators who work in a particular culture at a certain moment perform translations, their cultural views may have an impact on their work. Translation is significantly shaped by ideology, and ideology, in turn, can be somewhat influenced by translation (Bian & Li, 2021).

Shifts in translation: Formal correspondence versus translational (textual) equivalence
The term "shift", as introduced by Catford (1965, p. 73), distinguishes between formal correspondence, which exists between source language SL, and target language TL categories that occupy approximately the same place in their respective systems, and translational equivalence, which holds between two portions of texts that are actually translations of each other. For him, a shift occurs if there are "departures from formal correspondence" between the ST and the TT, i.e., if translational equivalents are not formal correspondents. He has argued that there are two major types of shifts: level shifts and category shifts. On the one hand, level shifts are shifts between grammar and lexis as in the translation of verbal aspects by means of an adverb or vice versa. On the other hand, category shifts are further subdivided into structure shifts (e. g. a change in clause structure), class shifts (e. g. a change in word class), unit shifts (e. g. translating a phrase with a clause), and intra-system shifts (e. g. a change in number even though the languages have the same number system). Generally, shifts in translation are seen as the alterations which result from the attempt to deal with the systemic differences between the ST and TT (Baker, 1998). The concept itself has been extensively adopted in nearly all the approaches as the process of translation can be perceived as a transformation where a system is replaced with another by preserving a similar communicative function. Hatim and Mundy (2004, p. 27) link Catford's 'formal correspondence' and 'textual equivalence' to Saussure's distinction between 'langue' and 'parole. Whereas formal correspondence is associated with 'langue', textual equivalence is concerned with 'parole', i.e., all that which language users say, write, etc.……i.e., their actual performance in real situations.
Formal correspondence can be established when two languages operate at the same grammatical ranks (hierarchies of units) such as a sentence, clause, group, word, and morpheme. In doing translation, an attempt should be made to reproduce several formal elements including: (1) grammatical units, (2) consistency in word usage, and (3) meanings in terms of the source context. In other words, the translation is a literal one. In many cases, however, we cannot stick to these certain formal elements of the ST. Catford (1965) argues that formal correspondence could not be said to be set up between all linguistic patterns in the TL system and SL linguistic patterns in the SL system. This fact obliges us to opt for textual equivalence which is closely related to the notion of shift. We begin translation from formal correspondence to achieve textual equivalence, and then we start operating with shifts when the identical meaning is either not available or not able to ensure equivalence. Hatim and Mundy (2004, p. 29) hypothesize that grammatical changes are dictated by the obligatory structures of the target language: we are frequently forced to make adjustments like shifting word order, using verbs in place of nouns, and substituting nouns for pronouns. Next we will survey the different types of shifts we have encountered in the process of translating the article mentioned above.

1. Level shifts
Level shift occurs when a ST item at one linguistic level has a TT translation equivalent at a different level. It includes shifts from grammar to lexis and vice-versa. For example: ST: This study examines Jewish-Israeli responses TT: tunāqishu ddirāsatu lḥāliyyatu rudūda lfi'li lyahūdiyyati l'isrā'īliyyati BT: The study present The ST deictic 'this'-a grammatical term-has been substituted by a lexical word in Arabic i.e., 'this', which fulfills a grammatical function in English, has been replaced by the word alḥāliyyatu 'the present', which is a normal lexical word in the Arabic linguistic system. Thus, this study has been translated as 'the present study'. At this level, it is clear that the translator is free as he has both grammatical and lexical structures to choose from according to his favorite style. In the example above, we could have kept the same form in Arabic, namely hātihi ddirāsa (this study).

Structure shift
Languages are by nature different in terms of the realization of similar structures existing in these languages and the type of structures in each language. Structures where one element is typically obligatory, while other elements are optional like agreement between the head and its modifiers, are usually observed in some languages. The case of number and gender agreement between nouns and adjectives in the Arabic nominal group is a case in point. Nonetheless, languages vary widely in the restrictions they assign to this agreement. In the English nominal group, for instance, this agreement is observed between articles and nouns but overlooked between nouns and adjectives. By contrast, Arabic adopts such agreement in both cases. This is a potential area of structural shift in translation.
ST: we focused specifically on two major schemes or positions through which Israeli-Jews relate to the Arab-Israeli conflict TT: rakkaznā bishaklin mouḥaddadin 'ala moukhaṭṭaṭayni aw mawqifayni ra'isiyyayni narbiṭu min khilālihimā bayna ssira'i lyahūdiyyi l'isrā'iliyyi wal'arabiyyi l'isrā'iliy BT: on two schemes or two positions main The example above highlights the fact that a shift is obligatory as duality, which is unmarked in English in terms of inflection as we number'two' before the nouns 'schemes' and 'positions', is necessary in the Arabic code where duality is marked by the inflectional morpheme ayni 'two'. As far as word order is concerned, the replacement of SVO in English by VSO in Arabic is obligatory. This means that structure shift occurs when there is no formal correspondence between the two languages involved in the process of translation; hence, the translator is not free as he is obliged to fit in the ideas in the target language scheme without further options: ST: In the asymmetric struggle of terrorists with state actors, the media play a central role.

TT: tal'abu wasa'ilu l'i'lām dawran muhimman wamarkaziyyan fi ssirā'i lghayri mutakāfi'in bayna l'irhābiyyīn walmas'ūlina rrasmiyyīn fiddawla. BT: plays media role central.
Because of the systemic differences that exist between English and Arabic at the level of word order, shifts are bound to occur at the clause level. The English clause-structure (SVC) 'the media play a central role' is expressed as (VSC) in Arabic: tal'abu wasa'ilu l'i'lām dawran mouhimman 'plays media a central role'.
On another scale, structural shift may also take place at other levels. Consider the following structures that illustrate the change that occurs at the phrase level in so far as the order of the constituting elements is concerned: In all the above cases changing the order of phrasal constituents is inevitable and the translator is not free to make any modification at the syntactic level.

Class shift
Class shift occurs when the translation equivalent of a SL item is a member of a different class from the original item. Catford (1965, p. 119) defines class shift following Halliday's definition as "that grouping of members of a given unit which is defined by operation in the structure of the unit next above"; structure shifts leads to class shift because of the "logical dependence of class on structure". Basically, class shift occurs when one part of speech is translated as another part of speech in the target language as in the shift from an adjective to a verb, a verb to a noun, a noun to an adjective, etc. To illustrate, in the example below, we have an explicit class shift from a verb in English to a noun in Arabic:

ST: The different and conflicting themes and frames that emerged as the interview unfolded can help to better understand the responses of viewers to this interview. TT: awḍaḥati lmawḍū'ātu wal'iṭārātu lmukhtalifatu walmutaḍāribati llati barazat khilāla lmouqābalati 'alā imkāniyatihā lilmousā'adati 'alā fahmin afḍal listijābāti lmushāhidīna lihādihi lmuqābala BT: for a better understanding.
This movement from one class to another class crosslinguistically, as in the other types of shifts, is bound to occur too when there is no corresponding lexical category available. In the majority of cases, however, finding a direct lexical match is easy.

Intra-system shift
Intra-system shift has to do with the changes that take place within a system. The equivalence is said to occur at a non-corresponding level in the target language system as all languages have their specific systems of number, deixis, articles, etc. For example, Intra-system shift happens when a term is singular in the ST and its textual equivalent is plural, or vice versa (e.g., when a change in number occurs even though the two languages have the same number system). In our example we have expressed the same notion of dates by replacing numbers with lexical markers of number.
ST: In the 1970s and 1980s, several regional newspapers developed, and news broadcast stations began to emerge as part of daily life TT: aḥrazat 'iddatu suḥufin iqlimiyyatin taqadduman khilāla ssab'īnāti wttamānināti wabada'at maḥaṭṭātu bathin taḍharu kajuz'in mina lḥayāti lyawmiyyati BT: during the seventies and the eighties.
In this special case, although Arabic has the same numeral system as English, we do not have the option, as translators, to render the same form in dates. Thus, the translator is obliged to replace the numbers with lexical forms.
Another instance of intra-system shifts occurs when a term that is plural in the ST has a correspondent in the singular in the TT:

ST: ………. affects viewers' perceptions of the interviewed terrorist and their
emotional responses to her.

TT: tu'athiru 'alā ru'yati lmushāhidīna fi lliqā'ati ma'a l'irhabiyyīn wal'istijābati l'āṭifiyyati lahā BT: the perception of viewers the response emotional
In this special case, grammatically speaking, we can use the plural in Arabic too, but in terms of style the above translation is preferable, which means that the translator opts for it to respect general accepted usage, a choice which is stylistically oriented. Although the translator is free in this respect, he prefers to conform to the way users of Arabic tend to prefer singularization instead of the opposite.

Unit shift
Unit shift occurs when the translation equivalent of a ST unit at one rank is a unit at a different rank in the TT. It includes shift from morpheme to word, word to phrase, clause to sentence, and vice versa. In situations like these, the translator has no choice but to follow the general patterns already in use in his community to avoid an odd translation.
To sum up, textual equivalence may require changes in units, number, person, structure, etc …, which are most of the time compulsory when we process two languages as distant as the ones we have since Arabic and English belong to two different language families: Semitic and Germanic ones. In addition to the shifts surveyed, we have shift by addition, shift by omission, and shift in the meaning of words (Baker, 1992). Here are some examples that we have spotted in the process of translating the media article mentioned before.

Shift by addition
As its name suggests, shift by addition occurs when the translator opts for structures that have more words than the original ones in the ST.
ST: Israeli security TT: ajhizatu l'amni l'isrā'īliyyi BT: security services ST: Al-Jazeera TT: qanātu ljazīra BT: AL-Jazeera channel In these situations, the translator is free to stick to the same number of words in the ST and reproduce the above phrases as al'amnu l'isrā'īliyy and aljazīra (of course by taking into consideration the context), or express the same concepts by adding more words. This depends on the translator's approach, which may be a either a one relying on minimizing the number of words in the TT or a one expanding the TT.

Shift by omission
Unlike shift by addition, shift by omission occurs when the translator opts for structures that have less words than the original ones in the ST.
ST: Participation, therefore, has two moments: participation in discussion and participation in adoption. TT: almushāraka filmunāqasha wafī tabannīha BT: participation in discussion and its adoption As in the previous shift, the translator is free to drop words from the ST or keep them in the TT in accordance with his stylistic preference.

Shift in the meaning of words
By far, this is the area where translators have more freedom in translating depending on their personal and ideological orientation; that is why this category is the most used one by translators. Baker (1992) has classified meaning into four types: propositional, expressive, presupposed and evoked meaning. Since expressive meaning cannot be judged as true or false as it relates to the speaker's feelings and attitudes rather than to what words and utterances refer to, the translator has the possibility to attack the text from his own perspective. In the example below, the translator shifts the meaning of "monsters" to "giants", which shows that he has a positive attitude towards western media. SL: Al-Jazeera rose up to challenge "monsters" like CNN and BBC, the companies that Al-Jazeera itself listed as its major competitors TT: taqaddamati ljazīra litaṣila 'ila taḥaddi "l'amāliqa" mithla cnn wa bbc BT: Giants.
Although we strongly believe that translation has to be faithful to the original message of the source text, all translators and researchers know that perfect equivalence between the ST and the TT is illusionary and acknowledge the fact that in any translation there is some kind of loss; however, our role consists of minimizing the degree of loss in translation to the least possible degree.
Generally, the freedom of the translator raises fidelity issues at the expressive level as translation is "a textual thing" (House, 1981, p. 65), but also a form of cross-cultural communication (Gregory, 2001). This has compelled theorists in translation to investigate the relevance of registers. Translation studies used to focus more on translation equivalence than on other key issues like discourse and register analysis approaches which have benefitted a lot from recent developments in discourse analysis. Based on Halliday's systemic functional grammar, Munday (2008) assumes that these approaches focus on register and discourse level (Hatim & Mason, 1990). As we all know, the effects of discourse go beyond the linguistic content to embrace the social, political and economic implications of discourse in translation. It is commonly believed that ideology plays the role of an "invisible hand" in translation as there are factors which influence translation through the transmission of ideology between different nations and countries (Yan, 2007, p. 63). In general, research has uncovered the different manipulations in the TT that signal the translator's ideology. This actually creates tension between 'foreignization ' and 'domestication', in what Venuti (2008) refers to as the "translator's invisibility". This situation reminds us of the translation procedure called "modulation", which is defined by Gérard Hardin and Gynthia Picot (1990) as "a change in point of view that allows us to express the same phenomenon in a different way." It is applied when translation, although structurally acceptable, is considered unsuitable or unidiomatic in the TL (Vinay & Darbelnet, 1995). Unlike the other procedures, this one is applicable at the whole message level. In the translated article, the word "suicide bomber" that refers to a Palestinian woman depicts the difficulties translators may face with their audience. When Arab translators translate it as intiḥāriyya 'suicide bomber', they respect the principle of faithfulness, but offend the Arab audience. If they translate it as istishhādiyya 'a woman who sacrifices herself for the sake of a sacred cause', they violate the principle of fidelity to please the Muslim community! Because of these tensions, the translator is never free in the choices he takes as he is tightly caught in the web of culture and ideology.
Taking the producer of the translation into account, his background has a direct influence on his text. As a result, for an Arab translator, translating this text by keeping the original meanings seems like an inadequate solution for Arabs as translation has to take into account all the variables that participate in the production and the reception of the produced text and the translated one respectively, namely the general context. In the case of our article, its translation is embedded in a political context characterized by a long history of conflict between Arabs and Israelis, a conflict which surely has a huge impact on how we translate the Israelis texts. As a matter of fact, domestication remains a logical technique from an Arab audience perspective.

Conclusion
In order to see whether translators are free in their endeavor to convey meaning in the TT, a grammatical analysis has demonstrated that this freedom is not possible at some levels. At the level of category shift, freedom is denied to the translator at the level of structure shift as inflection and word order are language-specific; therefore, with a lack of formal correspondence between the two languages, the translator is compelled to adhere to the main scheme of the target language. In class shift, which occurs when the translation equivalent of a SL item is a member of a different class from the original item, this movement from one class to another class cross-linguistically, like in the other types of shifts, is bound to take place if there is no corresponding lexical category in the TT; otherwise, the translator is free to move from one category to another for stylistic considerations. As far as intra-system shift is concerned, in cases where Arabic has the same system as English and yet the translator cannot reach formal correspondence, as it is the case with dates where he is obliged to replace numbers with lexical forms, the translator is not free at all. However, within the same category (intra-system shift), the translator may have the possibility to choose between two alternatives as in the treatment of singular/plural forms, but he chooses to comply with the way users of Arabic tend to prefer singular forms instead of plural forms. The last shift in categories, unit shift, where morphemes become words, word become phrases, and clauses become sentences, and vice versa, we have no freedom in choosing a correspondent form.
Freedom in translation is more apparent in two areas of translation. First, in level shift the translator has a chance to opt for either a grammatical or lexical structure because they are both available. Second, in shifts by addition, shifts by omission, and shifts in the meaning of words translators are free too. As we stated before, expressive meaning provides a fertile ground for translators to reshape the ST according to personal, cultural, social, and ideological perspectives. This means that any decent translation analysis should be done in the light of an examination of the ST and the TT by comparing and contrasting the purpose and the audience of the original text with the purpose and the audience of the TT, by sticking to formal correspondence when similar structures refer to the same content, and by having recourse to a more dynamic translation when translatability and comprehensibility are at stake.
To sum up, in this article we have reached some important conclusions. First, in order to reach a good translation, it is obligatory to use a register analysis as a component in the overall conceptualization of the movement from the original text to the translated text. Second, the only shift in translation that gives freedom to translators is intra-systemic shift, a part of class shift, shift by addition, shift by omission, and shift in meaning. Finally, we have found that expressive meaning is the most influential element in translation as it may lead translators to drift away in some crucial cases from the ST and take some liberty in modifying elements to fit them in the context of the target text in order to make his product reader-friendly.