SOURCE: BOOK TITLE: READINGS IN TRANSLATION THEORIES
98 10. Equivalence in translation theory W. Koller
"Equivalence" is obviously a central concept in translation theory, and much has been written on it. The general consensus nowadays seems to be that it is not helpful to think notion notion as a uniform uniform one, nondifferent nondifferentiate iated. d. Rather, Rather, there there are several several types more important than others, some applicable to one type
of the
of equivalence, some
of text and others to other types. Each text
needs its own hierarchy ofpriorities. Much
of the
modern modern discussio discussion n starts starts with Nida's Nida's advocati advocation on
of dynamic equivalence
against formal equivalence (see the introduction to chapter 9 above). Koller, in this extract from his book (1979), differentiates more types and sets out to clarity some
of the confusion surrounding the
term. The increasing use the study
of computers in linguistic research has an obvious application in, for instance,
of "connotations of frequency" (cf. 2(f)
below), with respect to the both items and
structures in different text types. We can surely expect more studies of this "statistical equivalence" in future. "Text normative-equivalence" normative-equivalence" (cf. section 3) is obviously linked to the study
of text types;
see chapter chapter 11 below, which also discusses discusses Bühler's analysis analysis referred referred to by Koller here. Koller's Koller's "pragmatic equivalence" (section 4), with its stress on translating for a particular readership, has close points
of contact with Vermeer's chapter 16, below.
To Koller's references to studies focusing on literary translation one might also add at least these in English: English: Brower Brower (1959), (1959), Brislin Brislin (1976), (1976), Holmes Holmes (1970), (1970), Steiner Steiner (1975), (1975), Bassnett-M Bassnett-McGui cGuire re (1980), Rose (1981), Frawley (1984), Hermans (1985). 0. The conce concepts pts "equi "equival valenc ence" e",, "equi "equival valent ent to", to", "the "the equiva equivalen lent" t" appear appear in defini definiti tions ons and and descriptions of the translation process, particularly in studies with a linguistic or communication approach. Examples are: equivalent elements (Oettinge (Oettingerr 1960: 1960: 110); 110); equivalent textual material (Catford 1965: 20); as equivalent as possible (Winter 1961: 68); the closest natural equivalent (Nida equivalent (Nida
and Taber 1969: 12); a maximally equivalent target language text (Wilss 1977: 72); communicatively equivalent (Jäger 1975: 36). Such definitions reveal quite different concepts of equivalence; and the picture becomes even more confusing when one looks at the various categories of equivalence that have been 99
.
proposed in the literature on translation theory: content equivalence (often also: content invariance), stylistic
equivalence,
formal
equivalence,
functional
equivalence,
textual
equivalence,
communicative equivalence, pragmatic equivalence, equivalence of effect. The following analysis is an attempt to specify the concept of equivalence more precisely, bearing in mind these various categories. (a) The concept of equivalence postulates a relation between SL text (or text element) and TL text (or text element). The concept as such does not say anything about the kind of relation: this must be additionally defined. The mere requirement that a translation should be "equivalent" to a given original is vacuous. (b) The kind of equivalence relation is defined in terms of the frame and the conditions to which one refers when using the concept of equivalence. In other words,
a normative
statement is made: there
exists equivalence between a given source text and a given target text if the target text fulfils certain requirements with respect to these frame conditions. The relevant conditions are those having to do with such aspects as content, style, function, etc. The requirement of equivalence thus has the following form: quality (qualities) X in the SL text must be preserved. This means that the SL content, form, style, function, etc. must be preserved, or at least that the translation must seek to preserve them as far as possible. Five factors can be argued to play a relevant role in the specification of equivalence types:
1. The extralinguistic content transmitted by a text; the kind of equivalence oriented towards this factor I call denotative equivalence (terms commonly found in the literature are "invariance of content" or "invariance at the content level"). 2. The connotations transmitted by means of the word choice (especially where there is a specific choice between synonymous expressions), with respect to level of style (register), the social and geographical dimension, frequency, etc; this is connotative equivalence (cf. "stylistic equivalence"). 3. The text and language norms (usage norms) for given text types: this kind of equivalence, having to do with text-type specific features, I call textnormative equivalence (cf. "stylistic equivalence", again).
4. The receiver (reader) to whom the translation is directed (who is supposed to be able to understand the text), and to whom the translation is "tuned" in order e.g. to achieve a given effect; this is pragmatic equivalence (cf. the commonly used term "communicative equivalence"). 100 5. Certain formal-aesthetic features of the SL text, including word play, metalinguistic aspects, individual stylistic features; the kind of equivalence that relates to these textual characteristics I call formal equivalence, although this is admittedly a heterogeneous concept (the literature also refers in this sense to "artistic-aesthetic equivalence", "expressive equivalence" etc, especially with respect to the translation of poetry). 1. Translation as the achievement of denotative equivalence sets translation theory the task of describing the potential equivalence relations between any two languages, together with the textual factors that determine the choice of a given equivalent in any specific case. Correspondences of different types (one to many, many to one, one to zero, one to part) need to be analysed in order that the translation process can achieve referential identity between SL and TL units. The central area of concern here is the lexicon (the words and syntagma of a language), since it is here that languages are (or should be) at their most productive (particularly regarding the use of existing or new methods of word formation), in order to account for ever-changing and expanding communication needs and aims. From the translation point of view, it follows that denotative equivalence is in principle attainable, even though the language may not always be very economically used in attaining it. "In principle" means disregarding the other factors which play a role in translation (readability and comprehensibility, the receiver, the connotative and formal value of the text, etc). 2. With respect to translation equivalence, the term connotative indicates that individual expressions in the textual context, and also complex texts themselves, do not only have a denotative meaning; according to the specific means of linguistic expression of the denotatum, additional values are also transmitted, particulary those with what Bühler (1934) called a symptom function. A single denotative meaning can be expressed in various ("synonymous") ways:
eat : dine : nosh die : pass away : kick the bucketcomplete : bring to completion we are the guilty ones : the guilty ones are we.
The following connotative dimensions are thus relevant for translation (see e.g. Rossipal 1973,
Baldinger 1968): 101 (a) connotations of speech level (connotative values such as + elevated, + poetic, + normal, + colloquial, + slang, + vulgar); (b) connotations of socially determined usage (+ student language, + military usage, + working-class language, + educated class, etc); (c) connotations of geographical relation or origin (+ non-regional, + American English, + dialect X, etc); (d) connotations of medium (+ spoken language, + written); (e) connotations of stylistic effect (+ archaic, + pompous, + artifical, +. fashionable, + euphemistic, + plain, + descriptive, etc); (f) connotations of frequency (+ common, + uncommon); -(g) connotations of register (+ normal usage, + technical, + medical); (h) connotations of evaluation (+ positively evaluative, + pejorative, + ironic,, etc); (i) connotations of emotion (+ emotive [i.e. using emotive language to describe a given topic], + neutral). 1 A major task of translation theory is to characterize the connotative dimensions of individual languages (e.g. with the support of stylistic studies), to analyse their features and structural elements, and then relate these to the connotative dimensions of a given target language. Further research could examine problematic cases in translations of particular texts, and also the translation procedures involved in the area of connotation. The achievement of connotative equivalence is one of the hardest problems of translation, and can seldom be absolute; this makes it all the more important to set up corpus-oriented studies of individual languages and texts, focusing on particular lexical and syntactic areas that are connotatively "loaded" (cf. e.g. Boecker 1973). 3. Legal contracts, instructions for use, business letters, scientific texts and the like all follow lexical and syntactic norms of both selection and usage (i.e. norms of style); to translate in accordance with these norms is to aim at textnormative equivalence. In a similar sense, in his discussion of translation criticism, Wilss (1974:37) speaks of "usage norms" because both SL and TL have certain pre-established schemata of linguistic expression, accepted
forms of linguistic behaviour and restrictive rules, where the communicative effect of the translation therefore lies in the TL realization of quite specific performance norms; these norms are basically intralingual, and hence also interlingually conventionalized to some extent, and they must be correlatable. 102 The description and correlation of these patterns of speech usage in particular text types is an important goal of a translation theory oriented towards two given languages. This kind of research can make good use of the methods and results of functional text analysis, which looks at the functionally differentiated, obligatory patterns of language usage in a variety of actual communication situations. 4. The concepts of "usage norm" and "functional norm" introduce a pragmatic point of view: in observing the usage norms for particular texts one takes account of the linguistic/textual expectation norm, the expectations that the reader brings to a given type of text. Now, it often happens that a text - e.g., a legal text - needs to be translated in such a way that the TL text does not follow the usage norms, because the TL readership is not restricted to a narrow circle of legal experts. In such a case the text must be translated, and in fact edited, into a form that will reach the intended receivers, i.e. be comprehensible to them. The achievement of pragmatic equivalence, then, means translating the text for a particular readership (see e.g. Heger 1976). This may - or even must - result in deviating from the requirements of text-normative, connotative or even denotative equivalence. A translation of a political commentary which sought to persuade the original readers to a particular political action usually has a different function in the target language, and addresses its readers with different presuppositions. From this point of view, translation theory should analyse the communicative conditions appropriate for different receiver groups in different language-pairs and texts, and establish the principles and procedures whereby pragmatic equivalence can be achieved. 5. To achieve formal equivalence in a TL text is to produce an "analogy of form" in the translation, by exploiting the formal possibilities of the TL or even by creating new forms if necessary. Reiss (1976: 21) describes this kind of equivalence as follows: It [the translation] orients itself towards the particular character of the work of art, taking as its guiding principle the author's creative will. Lexis, syntax, style and structure are manipulated in such a way that they bring about in the target language an aesthetic effect which is analogous to the expressive individual character of the source text.
Here, translation theory needs to analyse the possibilities of formal equivalence with the respect to categories such as rhyme, verse forms, rhythm. 103 special stylistic forms of expression in syntax and lexis, word play, metaphor and so on. There already exist a large number of individual literary studies of different texts and authors, and there are also some general discussions of literary translation such as those of Kloepfer (1967), Levy (1969) and Savory (1968).
6. In his study of the translation of scientific and technical literature, Jumpelt (1961: 46) refers with good reason to the empirical fact that "translation carinot guarantee a global, undifferentiated preservation of all values; translation always involves the necessity of a choice." With every text as a whole, and also with ;every segment of a text, the translator who conciously makes such a choice must set up a hierarchy of values to be preserved in the translation; from this he can derive a hierarchy of equivalence requirements for the text or segment in question. This in turn must be preceded by a translationally relevant text analysis. It is an urgent task for translation theory - and one on which no more than some preliminary work has so far been done - to develop a methodology and conceptual apparatus for this kind of text analysis, and to bring together and systematize such analyses in terms of translationally relevant typologies of textual features. 104
