Historical Overview of Equivalence in Translation Studies

The concept of equivalence has been of specific concern to many Western translation scholars. It was an essential feature of translation theories in the 1960s and 1970s, which sited it within the framework of structural linguistics. This paper will primarily discuss the notion of equivalence in a comprehensive way according to translation scholars since the 1950s. The first part of this paper will review the earlier debate on meaning and equivalence as one of the key linguistics issues in the 1950s. The general view of equivalence at that time attempted to explain how the source text (henceforth ST) and target text (henceforth TT) share some kind of similarity. This is followed by further efforts to explain this term according to the perspectives and approaches of several theorists. The second part focuses on contemporary translation scholars investigating equivalence in broader terms. Contemporary theorists have shifted from looking at differences in each type of equivalence and the effect of meaning to another perspective where equivalence is seen as a relationship between two texts.

Introduction

This paper will primarily discuss the notion of equivalence in a comprehensive way according to translation scholars since the 1950s. It is true that the definition of ‘equivalence’ is not as easy as it is supposed to be, but it remains a rather controversial term. Chesterman considers it “the big bugbear of translation theory, more argued about than any other single idea” (1997, p. 9). Other researchers regarded it in different ways–either as an important condition for translation, a category to give the translation a more suitable descriptive state, or a barrier that disrupts progress in translation studies (Kenny, 2011, p. 96).

The first part of this paper will review the earlier debate on meaning and equivalence as one of the key linguistics issues in the 1950s (Munday, 2016, p. 59), followed by further efforts to explain this term according to the perspectives of several theorists and approaches. The second part will focus on contemporary translation scholars investigating equivalence in broader terms.

Historical Overview of Equivalence

The concept of equivalence has been of specific concern to many Western translation scholars. It was an essential feature of translation theories in the 1960s and 1970s, which cited it within the framework of structural linguistics (Pym, 2014, p. 7). The general view of equivalence at that time attempted to explain how the source text (henceforth ST) and target text (henceforth TT) share some kind of similarity (Panou, 2013, p. 1). The term itself means ‘equal value,’ assuming the source text and the translation have the same value (Panou, 2013). In order to answer the question of what kind of relationship existed between them, various classifications were proposed, which gave birth to different definitions of equivalence (Panou, 2013). The following section will give a summary of the views of the most important scholars who discussed the concept of equivalence profoundly and provided the field with new paths to follow, namely, Jakobson (1959), Nida and Taber (1969), and Newmark (1981).

Roman Jakobson

Roman Jakobson was one of the most influential linguists of the second half of the 20th century. He was famous for dividing translation into three basic types, or what he referred to as ways of interpreting a verbal sign. These are intralingual, interlingual, and intersemiotic (Munday, 2016, p. 59). Under the second type, Jakobson discussed the basic problems that can occur in relation to linguistic meaning and equivalence (Jakobson, 1959, p. 233). He explained that we could not find full equivalence between code-units of source and target languages because of culture-based meaning related to each code-unit and, therefore, cannot be identical (Jakobson, 1959).

Jakobson also underlined that lacking a full equivalent is a principal concern in linguistics. This is so important, according to Jakobson, because no linguistic units can be delivered without a translation. In other words, translating meaning through the science of language from sign to sign in the same system or from one system to another is a key feature in interpreting a linguistic specimen (Jakobson, 1959, p. 234). Jakobson added that when we compare two languages, we should examine the ability to get mutual translation through extensive practice in translation activities through linguistics science (Jakobson, 1959).

Jakobson’s view of equivalence is affected by linguistic theory and his semiotic approach to translation (Leonardi, 2000). In this approach, the translator has to recode the message of the source text in the first place and then convey it into an equivalent message in the target text (Leonardi, 2000).

Eugene Nida and Charles Taber

In Toward a Science of Translating, Nida agreed with Jakobson that there is no identical equivalence, and thus, the translator should look for the closest possible equivalence (Nida, 1964, p. 159). He then introduced two basic types of equivalence: formal equivalence and dynamic equivalence (Nida, 1964). Nida’s view was influenced by theoretical concepts from semantics and pragmatics, allowing him to adopt a fairly systematic approach to exploring the field of translation studies (Panou, 2013, p. 2).

Nida said formal equivalence “focuses attention on the message itself, in both form and content” (Nida, 1964, p. 159). Formal equivalence gives significant attention to the message of the source language; therefore, the message of the target language, along with all its various components, has to match the source message as closely as possible (Nida, 1964). The translator who uses this approach should try to give preference to the message of the SL, covering the following: grammatical units, the usage of each word and the sense of the terms included. Formal equivalence is concerned with the culture of the SL. It always puts TL culture at the forefront compared with SL culture to seek accuracy and correctness according to these standards (Nida, 1964). From this, we can see clearly that the major aim of formal equivalence is to retain the key features of the original text whenever possible.

Nida later co-authored The Theory and Practice of Translation with Taber and used’ formal correspondence’ instead of ‘formal equivalence’. They also provided a more detailed discussion of formal and dynamic equivalence (Nida & Taber, 1969). They argued that if the grammatical and stylistic features of the target text are misrepresented, the message will not be understood fully, and it will be hard to reach the target audience in a clear form.

Dynamic equivalence, by contrast, involves translating the meaning of the source text so that the target text will have the same impact on the target context audience as the source text had on the source text audience (Nida & Taber, 1969, p. 200). Thus, the form of the original text will sometimes change, but if a translator follows certain rules, the translation is considered faithful, and the text itself is preserved. These rules include back transformation in the source language, contextual consistency in transferring meaning and transformation in the target language (Nida & Taber, 1969). Nida emphasised that “the relationship between receptor and message should be substantially the same as that which existed between the original receptor and the message” (Nida, 1964, p. 159).

It is obvious from Nida’s discussion that he generally favours dynamic equivalence since he considers it to be normally a more effective translation procedure than formal equivalence (Panou, 2013, p. 3). Nida’s main mission in translation is to produce the same effect on different audience categories in the target language as in the source language, which is much closer to the idea of dynamic equivalence (Panou, 2013). Nida and Taber stated that dynamic equivalence is not only a correct communication of information but also involves an expressive factor that makes sure people can understand and feel the meaning as if they are the audience of the source text. They also argued that many neglect or do not realise the importance of this most essential feature of dynamic equivalence (Nida & Taber, 1969, p. 25).

Additionally, Nida and Taber (1969) draw attention to the need to set priorities in translation (p. 14). These priorities help guide what should be preferred in each specific choice in language form. The priority states that contextual consistency has priority over word-to-word consistency, which is more or less related to the linguistic form of a language. Secondly, dynamic equivalence is a priority over formal correspondence (Nida & Taber, 1969). They point out that the second priority is based on the audience’s reaction, another piece of evidence given by Nida’s favouring dynamic over formal equivalence. Then, priority number three favours the oral form of a language over a written form that is specifically related to Bible translation. The last priority emphasises that acceptable and common forms to the audience have priority over prestigious forms. This priority focuses on the social background of the audience and looks at the effects that this can have on their understanding of the translation (Nida & Taber, 1969).

We can conclude that the major influence that Nida and Taber have had is the shift in the field of translation studies towards producing a systematic analytical procedure for translators (Panou, 2013, p. 3). These procedures can be applied to all kinds of texts. They also shed light on the audience’s important role and what to expect in each culture.

Discussion on Nida’s Work

The concepts of equivalence that Nida proposed have been criticized heavily by many scholars for various reasons. The following lines will summarise the most prominent reactions relevant to this thesis. Ernst-August Gutt (2000) argued that dynamic equivalence may limit the translation and mislead readers for functional reasons since one of the primary purposes of this type of equivalence is to make the target text as natural as possible such that the reader cannot guess it was a translation in the first place. This can mislead the reader, and it would be better if the translator favoured what worked for the readers (Gutt, 2000, p. 10).

Edwin Gentzler (2001) has criticized Nida’s work with a special focus on using the term ‘science’ to refer to his methodologies in translation. After a detailed review of almost all of Nida’s work in Toward a Science of Translating, Gentzler argued that Nida’s contribution did not provide the groundwork for what is generally understood in the West in general as a ‘science’, indicating that there is no scientific content in translation methods. For example, he commented that Nida’s concept of dynamic equivalence can only be presented as a useful model of the kind of translation that aims to manipulate a text to serve the audience’s interests. Consequently, this model could be used in specific genres, such as propaganda and advertising. Gentzler suspiciously asks how Nida could apply it in certain religious contexts (Gentzler, 2001, p. 59).

Finally, some scholars addressed the topic of equivalence effect. van den Broeck (1978) argues that it is impossible to determine equivalent effects in relation to different contexts. He argues that we cannot elicit the same effect and reaction from readers from clearly different cultures and sometimes different eras (van den Broeck, 1978, p. 40).

Peter Newmark

Following on from the work of Nida, Newmark approached the translation process by dealing with problems encountered during translation (Panou, 2013, p. 4). His well-known books Approaches to Translation (Newmark, 1981) and A Textbook of Translation (Newmark, 1988) clearly define what he means by ‘equivalence’. Instead of ‘formal correspondence/equivalence’ and ‘dynamic equivalence’, he used the terms ‘semantic translation’ and ‘communicative translation’ to describe the two types of translation methods he recognised (Newmark, 1981). He defined semantic translation as an “attempt to render, as closely as the semantic and syntactic structures of the second language allow, the exact contextual meaning of the original” (Newmark, 1981, p. 39), while communicative translation refers to “attempts to produce on its reader an effect as close as possible to that obtained on the readers of the original” (Newmark, 1981).

The main difference between these two types is that semantic translation focuses more on meaning, while communicative translation focuses on the effect made on the second reader (Panou, 2013, p. 4). The difference is demonstrated in the figure Newmark gave about the replacements of each term (Newmark, 1981, p. 39). In the figure, He represented semantic translation as a replacement for faithful, literal and source language-based translation. At the same time, communicative replaces idiomatic, free and target language-based translation. Newmark suggested that his new terms could narrow the gap between source and target languages even though this gap will remain a basic issue in translation theory and practice (Newmark, 1981).

Newmark commented that communicative translation tends to be smoother, simpler and direct to a particular language register. In contrast, semantic translation is more complex, concentrated and has many details (Newmark, 1981). Moreover, he viewed communicative translation as a translation that uses more generic forms, while semantic translation is over-translation because it is more specific than the original (Newmark, 1981).

It is worth emphasising that Newmark did not favour one type over the other but showed that each type of equivalence is suitable in certain contexts, even though it is not always possible to identify the best method (Newmark, 1981). Generally, he assumed that informative texts, which include recommendations, instructions and rules, maybe better translated communicatively. By contrast, texts containing definitions or explanations should be translated semantically (Newmark, 1981, p. 46). He also believes that literal translation is the best approach in both categories. However, he mentioned that in case of conflict, communicative translation should be considered to avoid any sort of abnormal or odd translation (Newmark, 1981).

Current Developments in Equivalence

Contemporary theorists have shifted from looking at differences in each type of equivalence and the effect of meaning to another perspective where equivalence is seen as a relationship between two texts (Kenny, 2011, p. 96). Those who support equivalence-based theories try to relate source text to target text as well as acknowledge the role of the topics of the text and translators in making and maintaining this relationship (Kenny, 2011).

In this section, the works of Pym (2014) and Dickinset al. (2017) will be reviewed with a particular focus on cultural equivalence.

Anthony Pym

Pym starts his contribution with equivalence, explaining from the value point of view that equivalence is a matter of ‘equal value’ (Pym, 2014, p. 6). He argues that what we say in one language can share or have the same value when translating it into a target language. This value can be on the level, the form, the function or something in between them. Thus, Pym divides the ‘equivalence paradigm’ into two ‘sub-paradigms’; the first relates to the idea that “things of equal value are presumed to exist prior to anyone translating” (Pym, 2014). He calls this type natural equivalence and contrasts this with the other type of directional equivalence. This second type refers to “an asymmetric relation where the creation of an equivalent by translating one way does not imply that the same equivalence will be created when translating the other way” (Pym, 2014, p. 24). He clarifies that he will use the term ‘natural equivalence’ to refer to theories that examine equivalence as a two-way process. That means seeing the possibilities of translation and back translation of each equivalence. In contrast, the term ‘directional equivalence’ will be used for those cases where an equivalence is being assessed from one side more than the other (Pym, 2014, p. 27).

Natural Equivalence

Pym (2014) argued that natural equivalence is a very simple idea, but it has become more complex as a term and a theory. The initial explanations of this type of equivalence assume that it is determined by function only, such that the translator should understand the function of the source text very well in order to produce the right equivalence (Vinay & Darbelnet, 1995). This assumption was followed by Nida’s famous division of equivalence, where formal correspondence looks for an equivalence with the same ‘form’ as the source text. In contrast, dynamic equivalence tries to imitate the cultural ‘function’ of the source text (Pym, 2014, p. 8). After that, many translation scholars developed theories involving equivalence according to various language levels (Pym, 2014). However, Pym comments that equivalence in practice is much less straightforward than these approaches would suggest. In this regard, he gives examples of the translations of various game show names, such as ‘Who Wants to be a Millionaire’. He shows that the translations of these in other languages do not fit notions of natural equivalence (Pym, 2014).

Pym considers a set of procedures for maintaining natural equivalence proposed by previous theorists (Pym, 2014, p. 11). He first presents the seven general types of natural equivalence according to Vinay and Darbelnet, referring to these as ‘translation solutions’ as a guide to follow to achieve natural equivalence in any context. The seven solutions are loan, claque, literal translation, transposition, modulation, correspondence and adaptation. As we can see, these types go from the most literal to the most recreative solution or from the easiest way to the hardest one (Pym, 2014, p. 12). Pym also stresses that when a translator chooses to follow these procedures, he/she should not start with the first solution in this list but with a literal translation. Then, decide whether to go up or down until determining the right equivalence (ibid). Furthermore, he comments that even though Vinay and Darbelnet could theorize the desirability of natural equivalence, they also accepted the need for translators to find other solutions (Pym, 2014, p. 14).

At the sentence level, Pym reviews Vinay and Darbelnet’s set of solutions, which they call prosodic effects (Pym, 2014, p. 14). They provide many solutions under this category, but Pym prefers to concentrate on six as they are the most important and common. Vinay and Darbelnet’s proposals are built on the fact that they might help the translator when confronting problems at the level of the phrase or the sentence (Pym, 2014). The first solution is amplification, which means using more words in translating than the first text. It is usually used when the grammatical structure of the target text requires more words than the source text. The second solution is reduction, which is the opposite of amplification–using fewer words in the target text than in the source text. Explicitation is the third solution: the translator explains and specifies what is implicit in the source text. The fourth solution is implicitation, which is the opposite of explicitation. The last two solutions are generalization and particularization; generalization involves using a more general term (superordinate/hypernym) in the target text compared to the source text, while particularization means using a more specific term (hyponym) in the target text than in the source text (Pym, 2014).

However, neither list of solutions is as clear as it looks when we provide each type with examples. In real translation, Pym shows that sometimes, we find several categories explaining the same equivalence relation and other situations where no solution can fit that equivalence (Pym, 2014, p. 14). Pym also raises a serious question about the validity of these solutions with respect to Asian languages such as Chinese, Japanese and Korean. He argues that when translating between English and these languages, the translator cannot start with literal translation as a basic solution as he/she would when translating between European languages but rather has to start with transposition or modulation. He notes that these Asian languages are syntactically dissimilar to Germanic or Romanic. Accordingly, when looking at a wide range of languages, it is hard to consider literal translation as a default solution in translation. Thus, we might regard literal translation as fitting only with European languages. Pym concludes, “The classical linguistic theories of equivalence require more work if they are to be extended beyond cognate languages” (Pym, 2014).

Directional Equivalence

In relation to directional equivalence, Pym investigated several theories and models which do not see the relation between equivalence as natural but rather as a result of specific decisions made by the translator (Pym, 2014, p. 24). If we translate from one language into another, we do back-translation, and the result may not be the same as the source text. Moreover, if two translators translate the text, we cannot guarantee they will produce the same target text. Unlike natural equivalence, this type is usually divided into two opposite kinds: free translation and literal translation (Pym, 2014).

Pym presents Andrew Chesterman’s (1996) thoughts about the two kinds of similarity as they have some relevance to directional and natural equivalences. Chesterman pointed out that we should look at the relation between source text and translation in terms of similarity rather than equivalence (Pym, 2014, p. 26). Thus, he represented two types of similarity, the first of which he calls ‘divergent similarity’. This refers to the way the translator understands the task of translating. Here, the translator produces a new text as a target text with similarities but cannot replace the start text. Directionality is the key that leads the translator from the source text to produce a new one, but the translator cannot do a back translation with the same tool (Pym, 2014). The other type is ‘convergence similarity’. This refers to how a translation is understood by its audience; they should have the same expectations of the target text as readers of the source text. Chesterman concluded that these types of similarity should be able to replace notions of equivalence. Pym comments that these types of similarity more or less cover both natural and directional equivalence: convergence similarity is related to the ideal meaning of natural equivalence that can operate very well in both directions, i.e., from source text to target text and vice versa (Pym, 2014). Pym comments, “If natural equivalence forms one side of the paradigm, directional equivalence would be the other” (Pym, 2014, p. 27).

Pym has also analysed the work of Otto Kade (1968). Kade proposed that there are four models or modes of equivalence at word or phrase level (Pym, 2014, p. 28):

1. ‘One-to-one equivalence’, e.g., in cases of a fixed technical term where there is a single expression in the target text to describe a single expression in the source text.
2. There is ‘one-to-several’ or ‘several-to-one’ equivalence. This refers to situations where the translator has one item in the source text corresponding to many items in the target text and vice versa.
3. There is ‘one-to-part equivalence’, meaning only a portion of an equivalence is available in the target text.
4. There is ‘one-to-none’ or ‘nil equivalence’, i.e., there is no available equivalence in the other text.

Pym comments that one-to-one equivalence is natural since the relationship involves two-way equivalence. However, one-to-several and one-to-part equivalence has the characteristics of directional equivalence because back translation will be less likely to yield the source text word or phrase. Nil equivalence is problematic, though it most reasonably fits into directional equivalence (Pym, 2014).

According to Pym, Chesterman and Kade were perhaps the most prominent scholars who touched upon directional equivalence. However, he notes that many other equivalence theories are directional (Pym, 2014, pp. 30–31). He claims they cannot be natural since they do not provide solutions for their types nor arrange equivalence about linguistic levels (Pym, 2014). Two examples of these types of equivalence are Nida and Taber’s (1969) formal vs. dynamic equivalence and Newmark’s (1981) semantic vs. communicative equivalence.

James Dickins, Sandor Hervey and Ian Higgins

Dickinset al. (2017) contribute to equivalence from a pedagogical perspective, presenting their ideas to student translators without focusing on previous theories and scholars. They believe that equivalence is a particularly confusing concept in translation studies.

According to Dickinset al. (2017), theories of equivalence in translation can be divided into two main types: descriptive and prescriptive. Descriptive equivalence refers to “the relationship between ST features and TT features that are seen as directly corresponding to one another, regarding the quality of the TT.” By contrast, prescriptive equivalence is “the relationship between SL expression and the canonical TL rendering of it as required, for example, by a teacher” (Dickinset al., 2017). Prescriptive equivalence is found in any theory which prescribes a certain type of translation for a certain type of text.

Dickinset al. (2017) argue that a particularly influential variant of prescriptive equivalence is Nida and Taber’s (1969) dynamic equivalence, with its notion of equivalent effect. They point out, however, that a drawback of the equivalent effect is that it might give the student translator carte blanche to translate as he or she wishes.

Dickinset al. (2017) also point out that achieving the same (i.e., equivalent) effect in both the source and the target texts can cause a problematic issue in translation (p. 17). In practice, it is difficult to indicate the exact relationship between the audience and a text as many parameters, such as cultural differences, can influence the full understanding of the message. Thus, they argued that the ‘sameness’ that equivalence usually implies may not help in translation but can be considered a barrier in both theory and practice. For this reason, they suggest that it is useful not to seek to maximize sameness when looking for equivalence between source and target text but to minimize the difference between the two texts. The translator should save as much of the source text as possible in the target text (Dickinset al., 2017). They also state that the transfer from source to target text involves some differences since the source language differs from the target language.

In his article More on equivalence, Dickins (2007) further develops the idea of minimizing difference, relating this to the multi-level (multiple matrix, level and rank) approach in the Thinking Translation series. Dickins starts by considering the relationship between similarity and identity (i.e., sameness) by putting them on a continuum where each term exists at one endpoint, such that two things may be less or more similar to one another up to the point where they are identical (Dickinset al., 2017, p. 1). He emphasizes the vagueness of the notion of ‘similarity’, such that we cannot say where similarity begins and ends: some things are very similar to one another, other things less so. Using similarity as a feature of equivalence is thus somewhat intrinsically problematic (Dickinset al., 2017). He also argues that identity (sameness) should be regarded as exceptional in real translation; mostly, a source text word or phrase and a target text word or phrase will be similar to one another (in any respect one has chosen to consider them), rather than identical (Dickinset al., 2017).

Textual Matrices Model

Having established that equivalence involves anything from similarity (or at least ‘significant similarity’) to complete identity, Dickins (2007) presents an in-depth analysis of equivalence based on the schema of matrices, levels and ranks implicitly established in the Thinking Translation series and proposed there as a tool for translation analysis (Dickinset al., 2017). We shall first review the model’s components (Dickinset al., 2017, p. xviii) with a special focus on the cultural matrix.

The scheme of textual matrices is divided into five matrices, one of which is the formal matrix, which has several levels involving a number of ranks. The five matrices are as follows:

1. The cultural matrix,
2. The varietal matrix,
3. The semantic matrix,
4. The genre matrix,
5. The formal matrix.

The formal matrix, which covers linguistic form, comprises six levels: the phonic/graphic level, the prosodic level, the grammatical level, the sentential level, the discourse level, and the intertextual level. Within the phonic/graphic and grammatical levels are several ranks of smaller (less complex) and larger (more complex) entities.

The genre matrix covers the genre of the text. Dickinset al. (2017) propose five basic genres for traditional Western texts. These are literary, religious, philosophical, empirical and persuasive (p. xviii). They also recognize hybrid genres resulting from mixing two or more genres. Dickins (2007) argues that if two cultures share the same basic genres, it is possible to find a generic equivalent, and the shift between genres in translation will not be problematic. In contrast, if one culture does not have the same set of genres as another, the translator must shift to a similar general to produce a rough target-text generic equivalent (Dickins, 2007, p. 10).

The semantic matrix covers the meaning and comprises denotative and connotative meanings. Denotative meaning is “the conventional range of referential meaning attributed to a linguistics expression” (Dickinset al., 2017, p. 291). Connotative meaning is “the implicit overtones that a linguistics expression carries over and above its denotative meaning” (Dickinset al., 2017, p. 291). There are numerous types (perhaps an unlimited number of types) of connotative meaning. The most important are attitudinal meaning, associative meaning, affective meaning, allusive meaning, collocative meaning and reflected meaning.

The varietal matrix deals with language variety and involves four aspects: tonal register, social register, sociolect and dialect. Dickins comments that equivalence in this matrix may involve either identity or similarity. Thus, the tone of an element in the source text might be identical to its equivalent in the target text. When it comes to dialect, by contrast, there is no possibility of source text/target text identity; all that can be said is that a target-text dialect may, in some sense, be similar to a target-text dialect (Dickinset al., 2017).

The last matrix is the cultural matrix. This does not present different levels but different translation options when dealing with features that are not common to both the source and target cultures. These translation options are exoticism (e.g., wholesale foreignness), calque (e.g., idiom translated literally), cultural borrowing (e.g., borrowing the name of a historical movement), communicative translation (e.g., standard translations of public notices and proverbs) and cultural transplantation (e.g., Romeo recast as ) (Dickinset al., 2017, p. xviii). Dickins et al. believe that cultural differences can present bigger challenges in translation than linguistic ones (Dickinset al., 2017, p. 36). It is impossible to have an identity (sameness) between culture-specific elements in two different cultures; all one can say is that culture-specific elements in different cultures may be significantly similar in various ways (Dickins, 2007, p. 12).

Conclusion

This paper has addressed the concept of equivalence in broad terms according to the prospects of translation studies. One could argue that many translation notions from the 1950s are predicated on two diametrically opposed approaches to translation. For instance, Pym distinguishes between natural and directional equivalence, Nida and Taber distinguish between formal and dynamic equivalence, and Newmark distinguishes between semantic and communicative translation. These divisive equivalence theories rapidly lost favour and were replaced by more appealing translation models such as Dickins’s matrix medal. However, equivalence serves as an indication of the main issues a translator has when translating. Thus, it is important to emphasize that it is one of the essential definitory axes of translation.