Equivalence Revisited: A Key Concept in Modern Translation Theory
By Sergio Bolaños Cuellar Universidad Nacional de Colombia
This paper attempts to discuss the importance, relevance and validity of the concept of equivalence as a constitutive notion in translation theory. Equivalence is defined as a relation that holds between a Source Langue (SL) text and a Target Language (TL) text. Our conception of equivalence is supported by a modern text-linguistic theory that considers that the text is the unit of analysis of the communicative event and, as translation itself is seen as a communicative event, then it turns out logical that it should be studied from a modern text-linguistic approach. A brief background is provided to support this perspective. However, this is a point of view not necessarily shared by all modern translation theorists and, therefore, two antagonistic positions in translation theory as well as the criticism against Linguistics/Text-oriented Theories (TOT) are analyzed. Then arguments for and against the notion of equivalence within TOT and COT (Non-Linguistics/Context-oriented Theories) are discussed in detail and new perspectives reviewed. Finally, equivalence within the framework of a Dynamic Translation Model (DTM) is discussed and a brief illustration of its application in translation criticism is provided.
Key terms: Translation Equivalence, Linguistics/Text-oriented Theories (TOT), Non-Linguistics/Context-oriented Theories (COT), Dynamic Translation Model (DTM)
Since then we have been working in this direction trying to prove our initial hypothesis was wrong (or right). This communicative approach has widened our perspectives, especially with regard to the way language actually works. Linguistics, the scientific study of language, had traditionally been concerned with the study of what Saussure called 'la langue', that is, the historically and socially determined linguistic systems (e.g. English, French, Spanish, etc). It is not until the 70's that a new orientation in linguistic studies takes place (cf. G.Helbig 1986). Within this new approach much emphasis is placed on 'la parole' (speech), that is, on language use. New language-related disciplines emerged and consolidated, among others, Textlinguistics, Pragmatics, Discourse Analysis, Sociolinguistics, and Psycholinguistics.
Of all these new disciplines it is Textlinguistics which we deem to be of special relevance for explaining the nature of translation both as a process and as a product. Textlinguistics plays a crucial and decisive role in the study of language use and obviously in the case of translation it does have a close relation with other disciplines such as Pragmatics, Sociolinguistics and Psycholinguistics. The central role played by the text in language use is explained by Hartmann (1971:15):
"Contrary to the current abstract and reductionist limitation to the linguistic system, text-oriented linguistics corresponds more strongly to the linguistic reality, to the fact that language only occurs and functions in texts, that the so-far-prevailing linguistic units (such as Phoneme, Morpheme, Word, and Sentence) do not occur or have meaning by themselves or as such but depend on previously arranged intentions and goals, on basic units, which provide them with sense, that is, with the capacity to function"
On the other hand, in our initial proposal of a Model for the Analysis of a Text as a Communicative Event (Bolaños 1995), we also pointed out that there were three core components for analyzing language use: Sender, Text, and Receiver. Other components which we labeled Conditions and Determinants are also important for understanding language use from a communicative point of view: Competences (grammatical, communicative, textual, and cognitive), Socio-Psychological Characterization of Participants (social variables such as gender, age, role; psychological variables such as motivation, attention, interest, memory), and Context (time and place of communication; historical, economic and social circumstances). At first this proposal was devised for analyzing monolingual communicative situations. Later we expanded it to include a bilingual situation similar to the one that takes places in translation (cf. infra Dynamic Translation Model). We underlined then that the text seemed to be a more suitable linguistic unit of analysis than the sentence, among other things, because: 1) It is a part of a communicative event that is contextualized and located in specific time and space coordinates with real participants and under real social, historical, and economic conditions, and therefore corresponds to the true nature of language that, by definition, is 'contextualized'; 2) It is not an immanent linguistic unit; its meaning is determined both by its inner structure and by the external factors that surround it; 3) its extension cannot be established beforehand; it may be as short as a word or as long as a whole book as far as it satisfies the participants' communicative needs (Ibid, p.57).
Now it is clear that the communicative approach for the analysis of language use holds both in the case of monolingual and bilingual situations. By definition translation is a bilingual situation that differs from bilingualism, among other things, because in translation speakers do not master the same language and need the help of a third party, i.e. the translator, that will re-establish the interrupted communicative act. Another important difference in relation to what occurs in bilingual situations is that in translation the translator has to reproduce the same or at least a very similar content the sender has uttered in L1 (semantically and pragmatically determined) but this time in L2. There is, so to say, a conscious effort on the part of the translator to be faithful to the message initially expressed in L1 in a process we can call of semantic and pragmatic 'duplication', which rarely occurs in traditional bilingual situations where the flow of information continues even if there is exchange of languages in the interaction, in which case phenomena such as 'code mixing' or 'code switching' may emerge.
Once translation has been recognized as a peculiar bilingual communicative situation, it becomes evident that the linguistic unit of such communicative interaction is the text. There is a source text (ST) written in L1 and a target text (TT) in L2. As this is an interaction that, linguistically speaking, takes place between participants through texts, then it should be textlinguistics the discipline that would provide a better foundation for understanding the translation process. This was foreseen by E. Coseriu (1977) more than 20 years ago when textlinguistics was just establishing itself as a new discipline:
Y en cuanto al estado de la investigación, la tarea no es fácil porque la teoría de la traducción debería, en rigor, ser una sección de la lingüística del texto y ésta, a pesar de los progresos de los últimos años, se encuentra todavía en sus comienzos (p.215).
When we attempt to describe and explain the relation that holds between a Source Language Text and a Target Language Text in translation we necessarily come across the concept of 'equivalence'. Our view is that translation does not exist as such if no clear link between ST and TT can be established. J. House (1997) expresses this point of view as follows:
"The notion of equivalence is the conceptual basis of translation and, to quote Catford, 'the central problem of translation practice is that of finding TL (target language) equivalents. A central task of translation theory is therefore that of defining the nature and conditions of translation equivalence' (1965:21)" (p.25)
G. Jäger (1989), from the Leipzig school of translation, presents his view about the importance of dealing scientifically with the concept of translation equivalence, more specifically in relation to the possibility or the need of using (=operationalizing stricto sensu) this concept for the practical goals of the so-called automatic (machine) translation:
"Against the background of modern conceptions of translation theory which attempt to understand globally the linguistic exchange, there arises inevitably the question about the general meaningfulness of research on the discovery and description of equivalence relations. Undoubtedly we would give an affirmative answer to this question and here we bear in mind a specially demanding test case for the science of translation: automatic [machine] translation" (p.33).
However there is still much controversy as regards the nature of this notion known as equivalence, to the extent that some authors even discard it as an unnecessary concept in translation theory. R. Rabadán (1991) summarizes adequately this bipolar situation:
"Como vemos, lejos de estar cerrado, el debate entre defensores y detractores de la equivalencia, se recrudece con cada nueva publicación (cf. Pedersen, 1988:11-29; Snell-Hornby 1988:13-21; Albrecht, 1990). Esta noción compleja e indeterminada parece estar condenada a seguir siendo la cuestión clave, y sin duda la más polémica, de los estudios contemporáneos de traducción. La raíz del problema está en que la equivalencia constituye la esencia misma de la traducción, y de ahí el gran reto con que nos enfrentamos: cómo conseguir que el texto original y su traducción sean 'el mismo texto' cuando todos los factores que intervienen en el proceso son, por definición, distintos (cf. Steiner, 1975:261)" (p.31).
In the remaining part of this paper these two antagonistic positions in translation theory as well as the criticism against Linguistics/Text-oriented Theories (TOT) will be analyzed. Then arguments for and against the notion of equivalence within TOT and COT (Non-Linguistics/Context-oriented Theories) will be discussed in detail and new perspectives will be reviewed. Finally, equivalence within the framework of a Dynamic Translation Model (DTM) will be discussed and some brief illustration of its application in translation criticism will be provided.
2. Linguistics/Text-oriented versus Non-Linguistics/Context-oriented Translation Theories
"The prevalent approaches can be divided -loosely but without too much conceptual violence- into a linguistics based orientation, aiming to construct an empirical science, and an aesthetics-based orientation that emphasizes the cultural and political values in forming translation practice and research" (p.8).
We propose to distinguish these two approaches in translation studies as follows:
Non-Linguistics/Context-oriented theories strive to understand translation as a culture-bound phenomenon, where the most important aspect is the translator's role and the effect the Source Language Text (SLT) will have on the Target Language audience (readers) (cf. infra). Within this approach the role played by ST in the translation process is rather ancillary, i.e. it is the point of departure for the translation process but once the process has begun there exists no compelling reason whatsoever to come back to it. One may say that it has already served its purpose. What actually matters in this contextual approach is how translation can be accounted for by the cultural and historical factors surrounding it.
We can briefly describe some of the most important COT by using J. House's (1997) characterization .
Anecdotal, Biographical and Neo-hermeneutic Approaches. They correspond to the typical anecdotal (=subjective) reflections made by generations of professional translators, poets, writers, philologists and philosophers (Iibid.p.2). In this approach the understanding and interpretation of the original and the translation are individual, creative acts that cannot be generalized or systematized (ibid). The translator's subjective interpretation and transfer decisions are based on his linguistic and cultural intuitive knowledge and experience. The relation between original and translation and the expectations of the target text readers are not given the attention they deserve, and the problem of distinguishing between a translation and various types of versions and adaptations is not even recognized (ibid). Some authors within this approach are: Paepcke 1986; Stolze 1992; Kupsch-Losereit 1994.
Response-oriented, Behavioral Approaches. For House "assuming that it is true that a translation should produce equivalent responses, the question remains, however, whether the degree to which this requirement is met, can be empirically tested. If it cannot be tested, it seems fruitless to postulate the requirement, and the appeal to 'equivalence of response' is really of no more value than the philologists' and hermeneuticists' criterion of 'capturing the spirit of the original'". This approach "all but ignores the raison d'être of any translation which undeniably lies in the existence of an original text, and the need to present that text in other words. [They] have nothing to say about the relationship between original and translated text, nor can they shed light on whether a translation is in fact a translation and not a version, an adaptation or another secondary textual product derived from an original text" (p.6). Some authors within this approach are: Nida 1964; Nida and Taber 1969; Carroll 1966; MacNamara 1967.
Literature-oriented Approaches: Descriptive Translation Studies. "In this approach the existence of a source text that served as a basis for the translated text is thus played down to a considerable extent" (p.7). For House, the basic problem in this approach is "how one is to determine when a text is a translation and what criteria one is to use for evaluating a translation -but these are questions which a descriptive translation researcher would probably never ask, since he would typically start from the hypothesis that a translation belongs exclusively to the literary system of the target linguaculture" (ibid). The major problem with taking this approach is summarized by House in one question: "On which criteria are we to legitimately say that one text is a translation, another one not, and what exactly are the criteria for judging the merits and weaknesses of a given 'translation'?" (p.8). Some authors within this approach are: Even-Zohar and Toury 1981; Hermans 1985; Toury 1985; van den Broek 1985/1986; Holmes 1988; Bassnett and Lefevere 1990.
Post-Modernist and Deconstructionist Approaches. Theorists in this approach "undertake to unmask the unequal power relations that are reflected in the translation directions from and into English, and the promotion of further English language hegemony through one-sided translations from English and an ever decreasing number of foreign texts being translated into English" (p.9). "No answers are sought for the question of when a text is a translation, and when a text belongs to a different textual operation" (p.11). Some authors within this approach are: Graham 1985; Man 1986; Spivak 1988, Benjamin 1989; Cheyfitz 1991; Venuti 1992; Niranjana 1992; Gentzler 1993; Bhabha 1994; Arrojo 1994.
Functionalistic and Action and Reception-theory Related Approaches. "In their functionalistic or 'Skopos-theory' of translation, Reiss and Vermeer (1984) claim that it is the 'skopos' , i.e., the purpose of a translation, which is overridingly important. Given the primacy of the purpose of translation, it is the way target culture norms are heeded that is the most important yardstick for assessing the quality of a translation" (p.12). More relevant for House's discussion "is the failure of the authors to spell out exactly how one is to determine whether a given translation is either adequate or equivalent let alone how to linguistically realize the global 'skopos' of a translation text" (ibid). Some authors within this approach are: Vermeer 1978/1996; Reiss and Vermeer 1984; Holz-Mänttäri 1984; Hönig and Kussmaul 1982.
On the other hand, Linguistics/Text-oriented theories do take into account the SLT and the way an equivalent TLT has been produced and consider it to be of crucial importance for determining how the translation process has occurred and to what extent it has been successful. Within this approach translation itself per definitionem has to do with the actual existence of a SLT to which one can refer. TOT are characterized by House as follows:
Linguistically-oriented approaches "In these approaches the source text, its linguistic and textual structure and its meaning potential at various levels (including the level of context of situation in a systemic framework), is seen as the most important, indeed constitutive factor in translation" (p.16) House's approach can be located within these linguistically-oriented approaches. Reiss' (1971) has been one of the most influential linguistic textual approach. She suggested that "the most important invariant in translation is the text type to which the source text belongs, as it determines all subsequent choices a translator has to make. She claims that different types of texts can be differentiated on the basis of Bühler's three functions of language: content-oriented texts, e.g., news, scientific, technical texts, form-oriented texts, such as poems and literary journal, and conative texts, e.g., advertisements and texts of persuasive bent" (p.17). Some proposals are revised (Koller 1972; Wilss 1974/77; Neubert 1994). Other authors' approaches that integrate recent research on sociolinguistics, speech act theory, discourse analysis and pragmatics (Hatim and Mason 1990; Bell 1991; Gutt 1991; Baker1992, Schreiber 1993; Steiner 1995; Gerzymisch-Arbogast 1994) are also discussed. House presents Koller's (1992) five 'frames of reference' for determining the type of equivalence: denotative, connotative, text normative, pragmatic, and formal-aesthetic equivalence".(ibid).
As pointed out above, COT have in common their concern for partial / (internal and external) contextual aspects of translation: the subjective interpretation of the original by the translator (hermeneutic approach), the maintenance of an allegedly similar response on the target audience (response-oriented approach), the peculiarity of receptors' polysystemic culture (descriptive translation studies), the question of power and manipulation in translation (post-modernist / deconstructivist approaches), and the purpose of translation according target culture norms (functionalistic / skopos theory). They also have in common their reluctance to recognize that translation is characterized as such and differentiated from other linguistic products because there exists a close link between ST and TT: equivalence. This is precisely the point of departure of TOT: the constitutive role played by the ST and its contextual factors in the fabrication of the translation into the target language is widely recognized.
3. Groundless criticism on Linguistics/Text-oriented Theories (TOT)
"Translation research and translator training have been impeded by the prevalence of linguistics-oriented approaches that offer a truncated view of the empirical data they collect. Because such approaches promote scientific models for research, they remain reluctant to take into account the social values that enter into translating a well as the study of it. Research then becomes scientistic, claiming to be objective or value-free, ignoring the fact that translation, like any other cultural practice, entails the creative reproduction of values. As a result, translation studies get reduced to the formulation of general theories and the description of textual features and strategies. These lines of research are not only limited in their explanatory power, but directed primarily to other academic specialists in linguistics, instead of translators or readers of translators or even specialists in other humanistic disciplines" (p.1).
Even though we have already discussed this aspect elsewhere (Bolaños 1997 / 2001), we would like to review once again the main points of our counterargument. First of all, it is not clear what Venuti means by "the prevalence of linguistics-oriented approaches that offer a truncated view of the empirical data they collect". It is evident that linguistics as such has not shown much interest in the problematic of translation. This is aptly highlighted in the following quotations from three different authors which cover a three-decade period: G. Mounin, J. Albrecht, and M. Doherty.
"Cosa todavía más singular referente al estudio científico de la actividad traductora: mientras que todo tratado empírico de filosofía debe incluir una teoría del lenguaje, ésta última jamás ofrece un estudio sobre la traducción, considerada como una operación lingüística, específica y sin embargo corriente, reveladora quizás en lo que se refiere al lenguaje y sin duda al pensamiento. La traducción, como fenómeno y como problema especial del lenguaje, ha sido silenciada. En Ferdinand de Saussure, en Jespersen, en Sapir y en Bloomfield, es difícil observar más de cuatro o cinco menciones episódicas, en las que el hecho de la traducción interviene de manera marginal, en apoyo de un punto de vista no relacionado con él, casi nunca por sí mismo; y en el cual el total de estas indicaciones apenas si llenaría una página" (G. Mounin 1963 / 1971: 25-26).
"It is one of the commonplaces in the literature on the science of translation to regret or to verify in astonishment that so far linguistics has barely dealt with translation problems" (J. Albrecht 1973: 1)
"This is a special issue of Linguistics in more than one way. Problems of translation do not belong to the regular topics of this journal. Nor do they belong to any of the other renowned linguistic journals and series. The specific aspects of translations are traditionally localized beyond linguistics proper, if not as extralinguistic altogether. And any linguistic aspect of translation should be identical to one of these dealt with in linguistics anyway ad therefore deserve special attention. It is one of the goals of this special issue of Linguistics to prove that translational problems are, to a large extent, genuine linguistic problems of a special type that has not yet been dealt with within linguistics systematically" (M. Doherty 1996:441).
In the previous quotations it is evident that there is no such a thing as "prevalence of linguistics-oriented approaches to translation" because linguistics itself has not been much concerned with this bilingual process. Furthermore we wonder what linguistic approach(es) Venuti bears in mind when he refers to the alleged denial of these linguistic approaches to recognize the cultural nature of translation and the "creative reproduction of values" in this process. It remains a guessing game.
Another statement by Venuti (1998) seems to single out specific shortcomings of the linguistics-oriented approaches:
"The key assumption of the linguistics-oriented approaches is that language is an instrument of communication employed by an individual according to a system of rules. Translation is then theorized on the model of Gricean conversation, in which the translator communicates the foreign text by cooperating with the domestic reader according to four 'maxims': 'quantity of information', 'quality' or truthfulness, 'relevance' or consistency of context, and 'manner' or clarity (Grice 1989:26-27; cf. Hatim and Mason 1990:62-65,95-100; Baker 1992:225-254; Neubert and Shreve 1992:75-84)" (p.21).
Based on the above quotation one might think that the linguistics-oriented approaches are only concerned with the way Gricean conversational maxims can be applied to explaining translation. This is not the case. Our reading of Hatim and Mason's, Baker's, and Neubert and Shreve's approaches clearly indicates that these authors do not exhaust at all their linguistic understanding and explanation of translation by mentioning Grice's maxims. This is evident from the outset when one reads, for instance, M. Baker's (1992) conception of linguistics:
"Linguistics is a discipline which studies language both in its right and as a tool of generating meanings It should therefore have a great deal to offer to the budding discipline of translation studies; it can certainly offer translators valuable insights into the nature and function of language. This is particularly true of modern linguistics, which no longer restricts itself to the study of language per se but embraces such sub-disciplines as textlinguistics (the study of text as a communicative event rather than as a shapeless string of words and structures) and pragmatics (the study of language in use rather than language as an abstract system" (p.5).
It seems that some of this critique against linguistics-oriented approaches is based on the lack of information about recent developments in the field of modern linguistics. If in fact linguistics devoted much attention to the study of language systems in the past, nowadays it is evident that this approach has been widened by the inclusion of the analysis of language use by disciplines such as sociolinguistics, psycholinguistics and textlinguistics.
4. Equivalence: a controversial notion
However, the denial of the need for the concept of 'equivalence' within the context-oriented approach has not been grounded convincingly as we will see below. Instead of clarifying that their rather (text-free) context-bound approach by its very nature does not include the concept of equivalence, these theories attempt to deny the nature of translation itself by portraying it as non-equivalence related phenomenon/activity. This has been pointed out (Kelletat, Wilss, House) as a major pitfall in this approach. The result of their non-equivalence-based process is not a translation, but another type of language outcome: a clearly culture-bound adaptation, an imitation, or some other text-type whose inspiration derives from an already wiped off (=discarded, omitted, forgotten) SLT. This has been somehow acknowledged by some authors like H.Vermeer (1990) himself, who prefers to talk about a culure-bound 'Translat', and not a translation stricto sensu.
One of the most remarkable critics of the notion of equivalence in translation theory is Mary Snell-Hornby (1986). After having compared the meaning of 'equivalence' in English and German, which the author considers as 'non-equivalent', she refers to the illusion of equivalence in the following terms:
"It means that the lexemes equivalent or equivalence exist nowadays in English above all as strictly delimited specialized terms, but at they same time they oscillate in the fuzziness of common language, that is, quantitatively relative in the sense of 'of similar significance', 'virtually the same thing'. And they entered the English specialized literature on translation with the latter, blurred, common meaning"(p.14). […]"Altogether, one should ask oneself whether Äquivalenz or equivalence are suitable terms in the science of translation: on the one hand Äquivalenz -as a scientifically fixed constant for a given goal- is too static and one-dimensional, and on the other hand equivalence has been watered down up to the loss of its meaning. Equivalence itself is not equivalent, although the similarity fakes: the borrowing from the exact sciences has turned out to be an illusion" (p.15)
Juliane House (1997) comments on the argument presented by Snell-Hornby for turning down the notion of equivalence:
"Given the relative nature of 'equivalence' and the fact that it has clearly nothing to do with 'identity' it is more than surprising that a polemic attack should have been directed against the concept of equivalence, in the course of which an analysis of the English and German dictionary meanings of the term 'equivalence' was presented (see Snell-Hornby 1986:12ff). Snell-Hornby singles out one particular dictionary entry, which supports her claim that equivalence basically equals identity and promptly proceeds to dismiss equivalence as 'an illusion' in translation studies. She writes that equivalence means 'virtually the same thing'. By contrast, I found the following dictionary entries for 'equivalent' and 'equivalence' in my own dictionary searches: 'having the same value, purpose… etc. as a person or thing of a different kind (Longman Dictionary of Contemporary English 1995), and 'having the same relative position or function; corresponding…' (Shorter Oxford English Dictionary 1993), as well as 'equivalence is something that has the same use or function as something else' (Collins Cobuild 1987). And in German, too, 'Äquivalenz' is not only a term in the 'exakte Wissenschaften' as Snell-Hornby claims: in my Brockhaus I read: 'das was in gewissen Fällen gleiche Wirkung hervorzubringen vermag'. Such entries were not mentioned by Snell-Hornby as they would clearly not serve her purpose of discrediting the concept of equivalence in translation studies" (p.26).
It is evident that the objection presented by Snell-Hornby in relation to the use of the notion of equivalence in translation theory is not totally convincing. In this respect we would say that 'equivalence' in translation studies, as pointed out by House, does not mean 'identity' or 'virtually the same thing'. The fact that this is the meaning this concept has traditionally been given in the so-called exact sciences does not entail that this very same meaning should be maintained when using the notion of equivalence in translation theory. In this case equivalence has to do more with 'having the same use or function as something else'. To think of equivalence in terms of complete 'identity' (e.g. as in mathematics) would be linguistically rather naïve, to say the least, because languages are by definition different and complex linguistic systems and, moreover, translation takes place not even between languages but texts embedded in complex communicative situations.
On his part, Wolfram Wilss (1992), one of the most remarkable translation theorists, expresses his astonishment in relation to some authors' refusal to acknowledge the usefulness of the notion of equivalence in translation studies as follows:
"Thus I find it incomprehensible that recently the concept of translation equivalence has been so discredited (Snell-Hornby 1986; 1988). It is true that in the 70's the science of translation considered the concept of equivalence partially in a rigid and static way. But the reason why precisely the representatives of the functional-sociocultural translation approaches want to wipe out this notion puzzles me. And this because the modern science of translation, as repeatedly pointed out, presents arguments by referring to texts and one cannot actually relate ST and TT but through a form of some micro- or macro-textual equivalent relation, which also holds when a text is not actually translated but e.g. 'transplanted' into a literary milieu" (p.197).
We would like to close this part by quoting J. House (1997), who summarizes her position in relation to the notion of equivalence and the way it has been criticized:
"The attack against the concept of 'equivalence' in the field of translation studies has a slightly dated touch: definitions of equivalence based on formal, syntactic and lexical similarities alone have actually been criticized for a long time, and it has long been recognized that such narrow views of equivalence fail to recognize that two linguistic units in two different languages may be ambiguous in multiple ways. Formal definitions of equivalence have further been revealed as deficient in that they cannot explain appropriate use in communication. This is why functional, communicative or pragmatic equivalence have been accredited concepts in contrastive linguistics for a very long time, focusing as they do on language use rather than structure. It is these types of equivalence which have become particularly relevant for translation, and this is nothing new (cf. Catford 1965)" (p.26).
5. New perspectives
"What is translated are utterances and texts; the translator establishes equivalence between SL-utterances/texts and TL-utterances/texts (SL=Source Language, TL=Target Language), not between structures and sentences of two languages"(1978:76) .
Recently the author (Koller 2000) has proposed to distinguish two concepts of equivalence:
"As a theoretic-descriptive concept equivalence designates the relation between a B text in language L2 (target language text, TL-text) and an A text in language L1 (source language text, SL-text) which allows to speak of B as a translation of A. Equivalence is then understood as a basic, constitutive, translation concept. It is suitable for distinguishing translations from other forms of secondary text products (texts related to a primary or source text)" (p.11) "As a translation normative critical concept equivalence is used in the sense of sameness of value between a target text (translation) and a source text (original text). Target language correspondences -from word to text level- are assessed. The optimal correspondence will be designated as equivalent, in contrast to non- or less-equivalent correspondences. This second use of the concept of equivalence belongs in the field of translation criticism and assessment. In the scientific discussion -as well as in the discussion between translation theorists and practitioners- the descriptive-theoretic and the normative-evaluative concepts of equivalence are often mixed up". (Ibid)
We do not agree fully with the above distinction of equivalence into two different concepts. The theoretic-descriptive concept is essential for distinguishing a translation from other secondary products (e.g. adaptations, paraphrases, imitations, etc). However it is the same concept of equivalence what accounts for the relations between ST and TT in case we are describing and assessing the way these relations have been established, that is, in translation criticism. Then our assumption is that as the concept of equivalence is dynamic it can be used for analyzing both the process and the product of translation. Nonetheless, this should not mean that two concepts are being used; we would say instead that the same concept is being employed for two different purposes.
On the other hand, the usefulness of this constitutive concept of equivalence for distinguishing actual translations from other secondary linguistic products (such as those resulting from a functionalistic-skopos approach) is stressed by the author:
"Equivalence as a disciplinary constitutive concept for the science of translation means selection, that is, reduction and abstraction: not all (secondary) texts which hold some relation (thematic and/or structural) with a (primary) text can be considered as translations and thus as subject-matter of the science of translation. Reduction and abstraction are characteristics of all empirical sciences that attempt to pursue substantial theoretical and empirical knowledge"(p.14)
Furthermore, a very enlightening proposal is made by Koller (2000) with regard to the double relationship of translation equivalence:
"Therefore the concept of equivalence should be made dynamic starting from the fact that translation is characterized essentially by a double-bound relationship: on the one hand by its specific relation with the source text and on the other hand by its relation with the communicative conditions on the part of the receiver. Translation concepts that focus exclusively on the special relation of translation with regard to the SL-text appear in this respect as problematic as conceptions that make absolute the relationship with the receiver, as is the case in action theories (Holz-Mänttäri) or functional theoretical approaches (Reiss/Veermer)"(p.21) "Translations that make absolute the relationship with regard to the source text run the risk of becoming illegible and incomprehensible; the borderline case of this type [of translation]is represented by word-for-word translation. On the contrary, translations that make absolute the relationship as regards the receiver's side run the risk of breaking the autonomy of the original text, to the extent that they overlook the specific translation relationship with respect to the source language text. The borderline case here has to do with target language original texts that relate more or less closely (thematically or structurally) with the source language text" (ibid)
And the author further clarifies:
In an attempt to determine the nature of the relationship that underlies the concept of equivalence in order for this notion to be useful in the analysis (description, classification and explanation) of translation cases, Koller (2000) proposes to specify and differentiate this concept as follows:
"There are different attempts to systematize the magnitude of equivalence. I have proposed myself (Koller 1992:214ff) to apply the following five frames of equivalence: (1) the extralinguistic situation, that is mediated in a text (=denotative equivalence), (2) the connotations mediated in the text through the type of verbalization (specially through the specific selection among synonymous or quasi-synonymous possibilities of expression) in relation to stylistic level, sociolectal and geographical dimension, frequency, etc. (=connotative equivalence), (3) the text and language norms (norms of use), that are valid for certain texts (=text-normative equivalence), (4) the receiver (reader) to whom the translation is addressed and who could receive the text, in which the translation is 'placed', based on his/her conditions for comprehension, in order for the translation to fulfill its communicative function (=pragmatic equivalence), (5) certain aesthetic, formal and individual characteristics of SL-text (=formal-aesthetic equivalence)"(p.24) .
Some aspects should be highlighted in Koller's proposal in relation to the concept of equivalence. First of all, it is clear that equivalence is the key constitutive concept that allows one to recognize the relationship that holds between a SL-text and a TL-text as a translation. Thus a distinction can be drawn between translations proper and other secondary linguistic products such as paraphrases, adaptations, summaries, etc., in which case there exists a link between a SL-text and a TL-text but of a different nature, i.e., the TL-text can also be an original text which is only distantly related (thematically or structurally) to the SL-text. A second aspect stressed by the author has to do with the existence of a double-bound relationship of equivalence with regard to the SL-text and the TL receiver, which prevents us to reach borderline cases such as word-for-word translations and original (not translation) TL texts. Finally Koller's proposal of frames of equivalence to distinguish five types of equivalence is very interesting. However, we consider that this initial point of reference needs further elaboration. In Koller's proposal it is not clear what the underlying conception of translation is and how it helps to give coherence to the proposed differentiation of frames of equivalence. In other words, one may think that these frames of equivalence are independent and, to some extent, unconnected. Neither do we find any exemplification of how these frames could be actually put into practice, that is, make operational. We attempt to deal with this issue in the next section.
6. Equivalence within the framework of a Dynamic Translation Model (DTM)
Three main basic components are distinguished in our DTM: 1) participants, 2) conditions and determinants, and 3) text. The participants in the translation process are: client (whose participation is facultative), sender (who utters a text in L1), translator (who is in charge of producing a target language text from a source language text the sender wrote in L1), receiver (audience/addressee of SL-text in L2). The conditions and determinants of the translation process are those factors which are present in any communicative event: participants' competences (linguistic, communicative, textual, cognitive, and with regard to the translator, translation competence), participants' socio-psychological characterization (gender, age, role, status; motivation, interest, memory, etc.), and context (historic, economic, social, cultural). The text corresponds to the linguistic realization of the communicative purpose of the sender in L1). It is structured in the following interwoven levels: (stylisic) syntacic (cohesion mechanisms such as references, connectors, substitutions, etc), (stylistic) lexical (linguistic variety: sociolect, dialect, jargon, etc), semantic (coherence mechanisms: time sequence, topic sequence, argument; presupposition, inference; denotation, connotation), pragmatic (speech acts: performative, constative; direct/ indirect; sender's intention (illocution) and intended effect (perlocution), and semiotic (combination of verbal and non-verbal signs in SL-Text: diagrams, tables, charts, etc.).
Let's see briefly how the model works. The potential initiator of the translation process is a client. If there is no client, the translator himself can start the process of translating a SL-text into a TL. The SL-text's sender and/or author wishes to fulfill some communicative purpose in his speech community. The translator should be aware of the communicative value of SL-text and its closest equivalent in L2. According to the different conditions and determinants, the sender fulfills some specific role when he produces the source language text. He can be a father, a club's member, a scientist, etc. Let's assume he decides to write a scientific text. In this case his most relevant social characterization is the role he plays as a scientist. There will be a potential community of receivers who belong to the same scientific field, to whom the text is basically addressed. They would be the first addressees of the text, that is, the text has been articulated in such a way that it is these readers who can work out more profitably the conveyed content of the message.
The translator in charge of translating this text should possess the linguistic and communicative, and textual competences for the initial reading of the text. Afterwards, as the text deals with a specialized topic, the translator activates his cognitive competence in order to update and contrast his previous knowledge with the knowledge that is being presented in the text. (cf. scheme-script-frame theories, Heinemann/Viehweger 1981:71) Thus he can fully understand the text's meaning. Now he activates his translation competence that will allow him to carry out a reading we have called 'surgical', which consists in reading the text once again in order to determine the way it has been constructed in relation to the stylistic syntactic and lexical characteristics, as well as the semantic, pragmatic and semiotic peculiarities, which may prove potentially problematic for the translation process. At this point a first draft of the translation is prepared taking into account the peculiarities and communicative potentiality of the target language, that is, equivalences are being established. Subsequent translation drafts are reviewed in order to verify that all translation problems encountered (=equivalence problems) have been adequately accounted for. Furthermore, here translation is seen as a problem-solving activity (cf. Wilss 1992:51ff). Then it should be pointed out that the translator carries out a dynamic task in search of equivalences, which takes him from the source language text to the target language text, and back to the SL-text, in a movement that can be traced by using our DTM.
Therefore, the main task the translator faces in his work is the establishment of equivalences in a continuous and dynamic problem-solving process. Instead of the five frames of equivalence -to a certain extent disarticulated- devised by Koller, we propose that equivalence is the relationship that holds between a SL-text and a TL-text and is activated (=textualized) in the translation process as a communicative event in the five text-levels we identified in DTM: (stylistic) syntactic, (stylistic) lexical, semantic, pragmatic, and semiotic, based on the SL-text verbalization and taking into account the conditions and determinants of the process, that is, participants' socio-psychological characterization and competences, and context. It is clear that equivalence is carried out at the different text-levels. We would speak then of equivalence at the stylistic-lexical, stylistic-syntactic, semantic, pragmatic, and semiotic text-levels. It is important to bear in mind that one cannot know beforehand which text-levels will be activated as problematic in the translation process, however one can say that equivalence-problem activation will take place at one or more of the described text-types of the DTM.
At this point two aspects should be clarified. First, the fact that equivalences are established at text-levels does not mean that they are isolated in and restricted to each of those levels. The text is to be understood as a complex, interwoven network of linguistic relations, and, for instance, the use of a lexical entry in a text does not simply affect the lexical level but may have semantic or pragmatic implications (denotations, connotations, special effects on receivers, etc.). Second, linguistic choices at the different text-levels are the product of a communicative event in which a SL-sender and a potential TL-receiver are involved; therefore conditions and determinants (participants' competences and socio-psychological characterization, and context) have necessarily been taken into account for fulfilling a communicative purpose. Thus they are somehow 'visible' in the different language choices present in the SL-text. The translator is also embedded in this communicative process and when he translates, i.e. establishes equivalences, he does so by taking into account again not only the linguistic material of the SL-text itself but also the conditions and determinants of the SL-text production and those of the TL-text reception. Our working hypothesis for analyzing (describing, classifying, explaining) translation equivalence is that it is a text-bound relation which is linguistically realized in texts and whose only tangible, empirically apprehensible form is the SL and TL texts. Therefore our point of departure and arrival is always a text. Obviously conditions and determinants of the translation communicative process should be taken into account as a resource to resort to when necessary but not as a perennial limbo of speculation which drives one away from the source language text or highlights excessively the role played by the translator.
Translation equivalence as envisaged in our DTM could have immediate impact on the work of the professional translator by helping him solve (=define, describe, analyze, explain) translation problems, on teaching translation as an inter-subjective activity where clear parameters can be established as to what may count as a translation, and on the field of translation criticism as it would allow to avoid simplistic, impressionistic (=biased) critiques of translations. In this respect we have already used our equivalence-related DTM (Bolaños 1998) for analyzing the translation of a literary text, El coronel no tiene quien le escriba, a novella by Gabriel García Márquez. This is a translation case, as are literary texts in general, where supposedly no textual equivalences could be actually established and, therefore, analyzed. Our initial premise is that no sound translation critique can be carried out if the whole text is not analyzed. The final judgment as to the translation quality as a whole should be made in terms of weaknesses and strengths of the analyzed translated text. By using a metaphor one would say that the translation product is like a wave: it has peaks and troughs. The ideal case would be a straight line, but it is ideal precisely because there are no absolute equivalences, that is, 1:1 inter-textual (SL-text and TL-text) equivalences. However, one should be able to determine when a peak or a trough has just gone off boundaries, i.e. when the translation equivalence is off the limits of the linguistically possible and textually realizable equivalence range. As a brief and partial illustration of how one can use DMT to assess a translation, one can analyze the initial paragraphs of García Marquez' novella with the corresponding English and German translations:
El coronel no tiene quien le escriba (Bogotá, Norma 1961/1996) El coronel destapó el tarro del café y comprobó que no había más de una cucharadita. Retiró la olla del fogón, vertió la mitad del agua en el piso de tierra, y con un cuchillo raspó el interior del tarro sobre la olla hasta cuando se desprendieron las últimas raspaduras del polvo de café revueltas con óxido de lata. Mientras esperaba a que hirviera la infusión, sentado junto a la hornilla de barro cocido en una actitud de confiada e inocente expectativa, el coronel experimentó la sensación de que nacían hongos y lirios venenosos en sus tripas. Era octubre. Una mañana difícil de sortear, aun para un hombre como él que había sobrevivido a tantas mañanas como ésa. Durante cincuenta y seis años -desde cuando terminó la última guerra civil- el coronel no había hecho nada distinto de esperar. Octubre era una de las pocas cosas que llegaban.
Der Oberst hat niemand, der ihm schreibt
No one writes to the colonel
In the initial sentence it is said that there was just 'una cucharadita' of coffee left. This was translated into German as 'ein Löffel voll übrig war', that is, as 'cucharada', the opposite of 'cucharadita'. In the English version this was rendered as 'one little spoonful left', which is actually equivalent to the original 'una cucharadita'. The adequate (=equivalent) rendering of the original denotation (semantic text-level) 'cucharadita' is very important also at the pragmatic text-level, specifically in relation to the author's intention and the effect on the reader. García Márquez could have chosen 'cucharada' instead of 'cucharadita', but he did not because as a writer he wanted to create from the outset an atmosphere of scarcity surrounding the colonel's life to be perceived by the reader as a crucial feature in the development of the novella's plot. This image of scarcity is reinforced by the semantic association between 'cucharadita' and 'últimas raspaduras de polvo de café' (G. letzten Kaffeepulver; E. last scrapings of the ground coffee).
The second paragraph of the Spanish original begins with a very interesting phenomenon at the syntactic text-level: a cataphoric reference, that is, some predicates about the subject (the colonel) are introduced without identifying it at the very beginning (Mientras esperaba a que hirviera la infusion, sentado junto a la hornillade barro cocido en una actitud de confiada e inocente expectative, el coronel…). The use of this textual syntactic mechanism has a direct bearing on the reader's visualization of the colonel's patience and also creates some expectation as to who is the subject of such actions. In other words this syntactic structure requires additional attention on the part of the reader, that is, it also has some implication at the pragmatic text-level. This cataphoric reference was recovered in the English version, but lost in the German translation, with an immediate consequence on the corresponding target readers. However, the English translation fails to reproduce the Spanish denotation 'hornilla de barro cocido' ('stone fireplace'), which is aptly reproduced in German as 'Herd am gebranntem Ton'. In English a humble 'hornilla' becomes an elaborate 'stone fireplace', thereby weakening the initial image of scarcity we referred to above.
Based on the discussion between Non-linguistics/context-oriented theories (COT) and Linguistics/text-oriented (TOT) theories, it is clear that these are two complementary perspectives to deal with translation. A refusal of TOT is not justified to the extent that modern (text)linguistics does take into account the text as the unit o analysis within the framework of a communicative event. If, as we state, translation is also a communicative event, then it is reasonable to expect modern linguistics to strive to shed light on the complexity of translation as a process and as a product. In line with this argument we have proposed to analyze the key concept of equivalence from this perspective. In order to accomplish this goal we have devised a Dynamic Translation Model (DTM) which has proved useful both theoretically and practically. In our approach the point of departure and arrival in the translation process of establishing equivalences between a SL-text and a TL-text is always the source language text. This dynamic process of the establishment of equivalences is seen as a problem-solving activity. Whenever ones translates one tries to produce a TL-text equivalent to the SL-text as a whole, but due to the text complexity, i.e. to the fact that the text is a network of interwoven relations, we have proposed to analyze (describe, explain) the text into different text-levels: (stylistic) syntactic, (stylistic) lexical, semantic, pragmatic, and semiotic. In the translation process the translator has necessarily to take into account what we have called the conditions and determinants of the translation communicative event, i.e. participants' socio-psychological characterization and competences, and the context (communicative event's time and place and social, economic, historical and cultural factors surrounding it). Therefore, contrary to the common belief held by COT, within our linguistics/text-oriented approach, the translator does resort to internal and external, non-linguistic/contextual factors, in order carry out his work.
Our linguistics/text-oriented approach, DTM, has also proved useful for criticizing translation products. This is a very demanding task which requires that the critic basically duplicates the translator's decision-making, problem-solving activity when he was establishing originally the equivalence between SL-text and TL-text levels. This task requires that both texts be fully compared in order to be able to state an objectively balanced judgment as to the quality of the translation as a whole in terms of its main (frequently repeated and semantically and pragmatically outstanding and relevant cases) strengths and weaknesses. Translation quality should tend to be homogeneous, i.e. with smooth peaks and troughs in the development of the translation wave, thereby keeping the TL-text within the equivalence range.
ALBRECHT, J. (1973). Linguistik und Übersetzung. Tübingen, Niemeyer.
|download als PDF-Dokument ca.150 KB|