Subjectivity in translation as interlingual re-instantiation

1. Introduction

Regarding theories of translation, Nida (1993, 155) observes, “As yet there is no theory of translation in the technical sense of ‘a coherent set of general propositions used as principles to explain a class of phenomenon’, but there are quite a few ‘theories’ in the broad sense of ‘a set of principles which are helpful in understanding the nature of translation or in establishing criteria for evaluating a translated text’”.

Indeed, translation theorizing lags behind practice. For instance, articles on translation theories constitute only 12% of the total published from 1949 to 1978 in China, while those on translation techniques make up 55.7% (Xu and Mu 2009). The existent theories of translation are more product-oriented than process-oriented. In other words, the theories are expressed in the form of criteria or principles for evaluating the effectiveness of translation. The process of translation is comparatively under-theorized.

The state of the art mentioned above is not surprising if we take into consideration the fact that translation as an interlingual and intercultural activity cannot be comprehended without insights from a number of relevant disciplines, namely linguistics, psychology, sociology, literary criticism, semiotics, cultural anthropology, communication theory, information theory and aesthetics. These disciplines are developing, and there are a flux of theories in each of them. As observed by Nida (1993, 157), translation theories are based on insights from these, and “basic points of reference for some primary insights” are from philology, linguistics and semiotics.

Linguistics as a discipline studying the nature of language naturally has more to contribute to theorizing the process of translation. But linguistics itself is diversified. Halliday and Matthiessen (1999) have differentiated two orientations of linguistics: (a) linguistics oriented towards logic and philosophy, with language seen as rules; and (b) linguistics oriented towards rhetoric and ethnography, with language seen as resources. First, the two orientations differ with respect to location of meaning: in (a), meaning is seen as transcendent – something that lies outside the limits of language; in (b), meaning is seen as immanent – something that is constructed in, and so is part of language itself. Secondly, the two differ with respect to the basic unit of meaning: in (a), the basic unit of semantics is proposition; in (b), the basic unit of semantics is text since language is functioning in context. Thirdly, the two differ in the metafunctional scope of their models of semantics: in (a), the metafunctional scope is restricted to the ideational function, with meaning seen as closely related to representation, reference, denotation, extension or “aboutness”; in (b), the metafunctional scope involves the ideational function, interpersonal function and textual function, with meaning seen as closely linked with context and rhetoric.

This paper will, following the rhetorical–ethnographical orientation, examine the concept of equivalence, and then elucidate the process of translation and subjectivity of the translator from a functional linguistic perspective. The specific questions to be explored are

1. On what parameters is equivalence idealized or created?

2. If translation is envisaged as interlingual re-instantiation, what kinds of information are processed by the translator in order to create maximal equivalence?

3. What phases of translation as interlingual re-instantiation particularly hinge on subjectivity of the translator?

2. Equivalence as ideal quality of translation

As pointed out in the introduction, existent translation theories are more product-oriented, expressed in terms of criteria and principles. In China, the most influential theory of translation was put forward by Yan (1896) in the preface to the Chinese translation of Evolution and Ethics by T. H. Huxley. Yan’s three principles, namely faithfulness, smoothness and elegance, have been extensively discussed as a traditional theory of translation. His contemporary Ma Jian Zhong also explored qualities of an ideal translation in his proposal to establish the institute of translation (Ma 1894). Addressing how to overcome the difficulty of translation, he first emphasizes the preparation, including the translator’s command of the two languages and his/her thorough understanding of the source text (ST). Then he explicitly states what an ideal translation should be like:

The translated text should be exactly equivalent to the original text without anything added or lost, so that the readers of the translated text can be benefited as if reading the original. That is true and ideal translation. (译成之文适如其所译而止，而曾无豪发出入其间，夫而后能使阅者所得之益与观原文无异，是则为善译也。) (Ma 1894, 2)

In the above citation, we find the concept of equivalence expressed clearly although no Chinese term is used that corresponds to “equivalence” at the time. “Equivalence” as a theory of translation is put forward by Nida (1993, 116):

The adequacy of translation has traditionally been judged on the basis of the correspondence in lexicon and grammar between the source and target language. The correspondence has frequently been stated in terms of “equivalence”, even though the term equivalence is often not used. There is, however, a serious problem involved in discussing the adequacy of a translated text primarily in terms of lexical and grammatical features, or even in terms of discourse structures. Translating means communicating, and this process depends on what is received by persons hearing or reading a translation. […] What is important is the extent to which receptors correctly understand and appreciate the translated text. Accordingly, it is essential that functional¹ equivalence be stated primarily in terms of a comparison of the way in which the original receptors understood and appreciated the text and the way in which receptors of the translated text understand and appreciate the translated text.

From these remarks, we can see that (1) Nida’s theory of translation is more product-oriented; (2) Nida’s interpretation of equivalence is restricted to perlocutionary effects on the part of the readers,² while the term can be understood in many other senses, that is, equivalence on different dimensions of language and text; and (3) equivalence is a gradient concept.

These points have been addressed by other translation theorists. Regarding equivalent effect of translation, Venuti (2009, 159) writes:

[…]the notion of an equivalent effect – that a translation can produce for its readers an effect that is similar to or the same as the effect produced by the foreign text for the foreign language reader – describes an impossibility: it ignores the manifold loss of contexts in any translation.

Hickey (1998) reminds us that a translator is not concerned with real effects produced on readers, only with potential effects. To validate his point, he cites the opinion of Gu (1993) that the agent of perlocutionary effects is after all the hearer, and that different individuals may respond differently in different contexts to the same locutionary act as trigger. Hickey (1998, 219) prefers the term “perlocutionary analogy”.

A translator […] must examine all potential perlocutions (the effects and responses reasonably predictable – on the basis of the common trigger or cause – in or on the mind, imagination, feelings or actions of a reader of the TT as a check to ensure that perlocutionary equivalence – “perlocutionary analogy” might be a more accurate term – has been achieved. […]Any TT, must, in turn, be capable of producing “analogous” perlocutionary effects on its readers. The use of the adjective “analogous” rather than “similar”, allows for the fact that, just as not all ST readers or their contexts are the same, so also the TT readers may be quite different, and situated in different contexts, from any of the ST readers.

Halliday (2009, 17) discusses six dimensions of equivalence, namely metafunction, stratification, instantiation, rank, delicacy and axis. He points out that “all translators know from their own experience that there are kinds of equivalence whose demands very often conflict”. Steiner (2001, 186) views translation as “preservation – or maximally close preservation – of experiential, logical, interpersonal and textual meanings in the relationship of translation between texts, or in the process of translation by the translator”.

Matthiessen (2001, 78) writes, “I shall assume that translation equivalence and translation shift are two opposite poles on a cline of difference between languages – from maximal congruence to maximal incongruence”. He proposes the principle of contextualization, “the wider the environment of translation, the higher the degree of translation equivalence; and the narrower the environment, the higher the degree of translation shift”.

In summary, the concept of equivalence has been long and widely accepted as ideal quality of translation. Nevertheless, equivalence can be interpreted in different senses because there are many parameters of equivalence. In addition, equivalence is a gradient concept. The degree of equivalence is closely related to context. It is the consensus of functional linguists that the translator endeavours to achieve maximal equivalence between the ST and the target text (TT) in ideational, interpersonal and textual meanings.

Although equivalence is seen as an ideal quality of the translated text, the attainment of that quality hinges on understanding of the translation process, the reading of ST and the reconstruction of meaning in TT. Since translation is intercultural communication through language and text, it is essential that the translator understands the relation between language and text, the relation between language and cultural context, and the relation between text and situational context. These relations have been recently elucidated by systemic functional linguists (SFL), a point to which we turn in the following section.

3. Translation as interlingual re-instantiation

It is pointed out in the introduction that the translation process is under-theorized, and in Section 2 that equivalence as an ideal quality of a translated text still needs clarifying. Recent developments in SFL have led to new insights into linguistic communication and given novel interpretation of the process of translation. In this section, a brief review will be made of the three-dimensional model of linguistic communication and the view of translation as re-instantiation will be discussed, a process model of translation as re-instantiation will be then proposed to account for the sorts and sources of information processed by the translator in order to achieve maximal equivalence.

3.1 Three dimensions of linguistic communication

In Halliday and Matthiessen (1999, 2009) and Martin (2012), a model of language as social semiotic can be seen. They model language and communication as composed of three complementary dimensions: realization, instantiation and individuation.

SFL sees language as a multistratum system, a resource for making meaning. The speaker’s/writer’s communicative purpose in context is realized by discourse semantics, which is realized by lexicogrammar, which is, in turn, realized by phonology/graphology. Realization is a scale of abstraction.

In SFL, instantiation is a term used to refer to the cline relating language as a system and text as an instance of language use. Language as a system evolves in its context of culture, while a text is generated in a context of situation. A type of situations produce a certain kind of text named genre. Language as a system provides resources for the generation of a text. A text is constructed with system resources deployed, and in the meantime it contributes to the development of the system. The relation between language, text and context is shown in Figure 1.

Figure 1.

A model of instantiation (adapted from Halliday and Matthiessen 2009).

Individuation specializes meaning potential according to users rather than uses of language. Martin (2012) draws from Bernstein’s terms to iterate this dimension of linguistic communication. A person’s meaning potential is named repertoire, and the meaning potential of the community is termed reservoir. Martin holds that Berstein’s interpretation involves both recognition rules and realization rules. The former allows speakers “to identify the specificity or similarity of contexts and thus orient to what is expected or legitimate in that context; realization rules enable speakers to produce culturally specific texts” (Martin 2012, 31).

The three dimensions represent simultaneous commitment of language in use. Realization accounts for a scale of abstraction of meaning potential as resource. Instantiation accounts for specialization of meaning potential in context. Individuation accounts for specialization of meaning potential according to users. This new model of SFL has important implications for understanding the process of translation.

3.3 A process model of translation as re-instantiation

Based on the above insights, a process model of translation as re-instantiation (Figure 2) is constructed to account for the sources and sorts of information processed by the translator in order to achieve maximal equivalence, which is understood as analogous intertextual relation between ST and TT. As SFL scholars hold that ST and TT should be equivalent in interpersonal, ideational and textual meanings, in the figure the analogous intertextual relation between ST and TT is symbolized by three horizontal dotted lines.

Figure 2.

A process model of translation as re-instantiation.

As indicated in Figure 2, the sorts of information dealt with by the translator can be put into two categories, namely encoded information (indicated by solid lines) and subliminal information (by dotted lines). Encoded information is provided by ST as discourse meaning. It must be processed by the translator through careful readings and then conveyed in another language to the readers of TT. The translator, both as a reader of ST and the writer of TT, serves as a mediator between the author of ST and the readers of TT.

Subliminal information includes knowledge of the source language system and its context of culture, knowledge of the target language system and its context of culture, the context of situation in which ST is produced, the context of situation in which the TT is written, information about the community of readers of ST, and anticipation of the community of readers of TT. Although ST and TT are the start and the end of the translation process, all the subliminal messages contribute to the success of the task of translation. As elucidated in Section 3.1, a text as an instance of the language in use is generated in a situational context to fulfil the purpose of communication. It is created with resources of the system and the system evolves in its context of culture. Therefore, the translator must take into consideration relevant subliminal information in reading and creating although he/she works on the two texts. The model reflects the complexity of the process of translation as interlingual re-instantiation. The complex process of interlingual re-instantiation requires expertise and competency of the translator, which is exemplified in the following section.

4. Subjectivity in the process of translation as re-instantiation

Based on the above delineation of the process of translation as re-instantiation, this section discusses subjectivity of the translator. As indicated in Figure 2, in reading ST and writing TT, the translator needs to think independently and creatively. Pragmatics has shown that understanding an utterance often requires inference on the part of the hearer (Leech 1983) because what is meant is not always what is said. Cognitive linguistics has revealed that subjective attention is necessarily involved in constructing a sentence. “The lamp is over the table” and “The table is below the lamp” describe the same scene, but which of them is actually uttered depends on the speaker’s attention of what is given information (Langacker 1987). The instantiation model put forward by Halliday and Matthiessen (2009) integrates language, text and context in linguistic communication. Language as a semiotic system is conventional in nature, but a text as language in use is constructed by the writer to fulfil his/her purpose in a situational context. It naturally contains the writer’s subjective consciousness. In this connection, the author holds that both conventionality and subjectivity are inherent in linguistic communication. The higher the rank scale of language in use, the more subjective; the lower the rank scale, the more conventional. If a text and a morpheme are taken for the two ends of the cline of rank scale, the degree of subjectivity in a text is definitely higher than that in other rank scales. As an ST contains subjective consciousness on the part of the writer, the task of the translator is first and foremost to work out the writer’s intention.

In the process of translation, making inferences in understanding ST and reconstructing meaning in writing TT clearly require the translator’s independent and creative thinking. Here is an advertisement posted near our campus, which may illustrate the point:

不小心， 有了。

(bu xiaoxin, you le)

(“not”“take care” “have got” “particle”)

到协和，没了。

(dao Xiehe, mei le)

(“go to” “Xiehe” “not have” “particle”)

If one is asked to put it into English, he/she must first figure out what is got, who has got it, then get to know who are the intended receptors of the ad. The next step is to reconstruct the discourse meaning in English, striving to produce a maximally equivalent English text. One possible translation is this:

Carelessly you have got a problem, lady?

Easily it will be resolved here in Xiehe (Hospital).

As getting pregnant and undergoing abortion are highly private, in the original text these are implicitly expressed. So it is absolutely necessary for the translator to infer correctly the implicit meaning before putting it into English. In the English text, the word “lady” is added so as to explicate the receptors and to specify the meaning of problem. The two texts correspond in interpersonal, ideational, and textual meaning and context of situation. The idea of getting pregnant unintentionally is expressed implicitly in English through “carelessly”, “problem” and “lady”. This example illustrates the necessity of the translator’s subjectivity, particularly making inference and reconstructing meaning in the process of re-instantiation.

Lack of subjective consciousness in making inference and reconstructing meaning leads to translation shift. Here is a sentence from a book entitled Why Photography Matters as Art as Ever Before:

What sets diCorcia’s street photographs apart from earlier work in that tradition is the dramatic to-be-seenness.

The writer, a reviewer of photography, coined the word to-be-seenness to highly praise the special feature of the works of the photographer diCorcia. A translation student translated the word into待看到性, which is coined by him as an equivalent. But the coined Chinese term does not make sense. The translation shift is due to the inability to recreate meaning in the TT. The translator fails to re-contextualize the original meaning of highly positive evaluation of the photographer’s style. Actually the keyword to-be-seenness can be translated into a Chinese idiom 耐人寻味。

These two examples suffice to illustrate the necessity of the translator’s subjectivity, particularly making inference and reconstructing meaning in the process of re-instantiation. In the rest of this section, data drawn from two English translated texts of Lao Zi’ s Book of Tao and Teh will be discussed to further reveal the subjectivity of the translator.

5. Conclusion

The paper starts with the observation that the process of translation is under-theorized compared with the product of translation. The concept of equivalence, which is widely accepted as the core of criteria of translation, is then examined. Recent development of SFL provides new insights into the concept of equivalence. The interpretation of equivalence as similar perlocutionary effects of St and TT seems to be waning and the idea of equivalence in terms of interpersonal meaning, ideational meaning and textual meaning is waxing. Based on this assumption, the nature of translation is discussed from the perspective of SFL, and a process model of translation as interlingual re-instantiation is proposed to reveal the kinks of information processed by the translator. The translator has to work out the discourse meaning of ST and then to reconstruct meaning in TT. Doing both of these is a subjective conscious process. The subjectivity of the translator, specifically making inferences and selecting rhetorical strategies, is discussed with data from two versions of English translation of Lao Zi’s Book of Tao and Teh by a Chinese and a British expert translator, respectively.

The paper is written out of an attempt to deepen understanding of the process of translation and the competency of the translator. Given the situation of translator preparation programmes in China and the rest of the world, it is an urgent task to do a better job in translation theorizing. More than 150 universities in China are running Masters of Translation and Interpretation (MTI) programmes. It is definitely inadequate to teach the students what an ideal translated text is like without showing them how the product is produced, that is, what the complex process involves and how the translator copes with the complexity. The author’s experience in working with MTI students reinforces his conviction of the value of translation theorizing. Admittedly, translation theorizing can be undertaken in relevant disciplines other than linguistics and other schools of linguistics. It is a daunting but rewarding task to gain more new insights into the process of translation that contribute to the preparation of competent translators in the context of globalization.

Disclosure statement

No potential conflict of interest was reported by the author.