323 :JULY 1976) CURRENT TRENDS IN TRANSLATION THEORY DAVID CRYSTAL of a text. For example, adjectives used to describe the desired quality of a SOME CURRENT TRENDS IN TRANSLATION translation, taken from recent correspondence on liturgical matters, included THEORY1 "sober" and "discreet": it is, however, difficult to see in what way such notions as linguistic sobriety might be defined. And in general, one would Prof. David Crystal isProfessor of Linguistic Science at the University of Reading, U.K. like to see a specification of translation variables in linguistic terms as well The aim ofthis paper isto outline those trends in the theory oftranslation in as in psychological or emotive terms. recent yearswhich seemto have beenmost productive inclarifying discussion (iii) Developing theories about the synchronic and diachronic relationships of the priorities and practicalities of everyday work in translating. Such between languages-whether, for example, there is evidence to support the tendencies could derive from many fields. Since 1945,the theory of transla notion ofacommon "deep structure" for alllanguages (cf.eighteenth-century tion has been influenced by ideas stemming from linguistics, literary criti questions for "rational" languages, and of late Chomsky's emphasis on cism, ethnography, communication theory, machine translation, psychology universals; cf. also Firth's criticism of such "naked ideas" theories in and philosophy (cf. the review in Nida 1974)-this contrasting with the Palmer, 1972), or whether there are fundamental psycho-/socio-linguistic pre-war situation, when translation tended to be restricted to the domains of barriers separating different languages (the "deterministic" model of the philologist,languageteacher and professional interpreter. Indeed,attempts Humboldt, Sapir, Whorf; cf. for example, the excessiveclaims made for the at a scientificallyorientated theory oftranslation as such (as opposed to the uniqueness of Hebrew thought soundly criticized by Barr, 1961).The two practice and methodology of text translation) are extremely recent, as the possible models can be diagrammed as follows: contributions of Firth, Catford, Nida and others show. The dominant L1 L2 L1 L2 influencesseemto have comefrom theoretical linguisticsand anthropological linguistics in its broadest sense(i.e. subsuming ethnological and sociological surface surface surface < surface studies of language); at present, there is increasing influence of ideas from structure structure structture ::>structture t t psycholinguistics, but generalizations here would be premature. I shall therefore restrict myself to issues arising out of the former two areas. deep <=> deep deep deep The viewpoint of a general linguist is inevitably limited, by the nature of structure structure structure =f= structure his calling, and certain topics, crucial to any evaluation of the practice of (or perhaps) translation, fall outside of his competence. In the present connection, the ~ ~ following issuesare outside my scope: deep (i) problems of determining which original texts, editions, etc. to use; structure (ii) determining the exact meaning of a source text; (iii) the criteria for deciding on the choice of a text to use in a religious The truth is probably somewhere in between. (e.g. pastoral, liturgical) context; (iv) Clarifying the notion of translation equivalence, or accuracy; see (iv) decisions as to which doctrinal or devotional tradition should be further below; and followed in formulating a translation policy; (v) Clarifying the correlative notion of translation acceptability, the (v) decisions about the aim of the translation, e.g. whether for scholarly, permitted tolerances of variation, along with the explication of such related aesthetic, missionary, etc. purposes. concepts as the "level" of a translation; seefurther below. The insights ofa general linguistic approach, bycontrast, willrelate to the The central task isto investigate the latter two issues,viz.the concept ofan following issues: "acceptable translation". The dominant development in this respect seems (i) Emphasizing the need for a comprehensive account of the translation to reflect very closely the movement of ideas within linguistics as a whole, process: see further below, but cf. the relativity of much of the recent dis namely, the replacement of outmoded,2 simplistic and prescriptive concep- cussion of liturgical translation, which focused on certain distinctive 2 Traditional conceptions of translation, whose inadequacies are now generally recog features of the proposed texts (such as thou v. you, vouchsafe v. grant), and nized, would include that of "word-for-word" translation and "literal" translation (the largely ignored the pervasive and more fundamental significanceof syntactic distinction is essentially that the latter is grammatically acceptable, whereas the former construction (seefurther Crystal and Davy, 1969:ch. 6). rarely is; neither, however, need bemeaningful: cf.It's rainingcatsanddogs-Il estpleuvant chats et chiens-Il pleut des chats et deschiens). Most modern conceptions work within the (ii) Replacing the subjective language of description of the qualities of notion of "free" translation, this term not meaning "loose" or "inaccurate", merely that translation by a more objective metalanguage based upon the characteristics the unit oftranslation isa variable-sometimes the word, but more often the sentence, or some other major grammatical unit, as the starting-point. One is thus translating sense for 1This article isbased on a paper presented to aseminar on Christianity inIndependent pense: there is no necessary correspondence between any grammatical or vocabulary Africa, held in February 1975under the auspices of the School of Oriental and African sattern in the two languages-though some general parallels will usually emerge, e.g. it Studies of the University of London. The seminar was part of a research programme pleut averse. involving several African university and ecumenical centres. 324 THE BIBLE TRANSLATOR (VOL. 27 NO. 3 JULY 1976) CURRENT TRENDS IN TRANSLA nON THEORY 325 tions of translational equivalence by a structured, dynamic and humble usually possible only with distortion of syntactic patterns (cf. the use of awareness of the complex reality of language structure and use. These weird word-orders in hymns, in order to get the lines to rhyme). Total adjectives, as they stand, are, however, opaque, and need clarification: one linguistic equivalence-in the sense of preserving equivalence at all levels method of doing this is to illustrate what is meant by "complexity" in is therefore an impossibility. Lexical and semantic factors are usually per relation to the notion of translation equivalence. mitted to outrank the others, but in certain contexts (usually literary or (A) Complexity in language structure. Here the main influence has been to aesthetic) the other factors are regularly considered as having an important think in terms of "levels" (or "components") of language structure, each bearing on the finished work. The notion of levels has therefore more than level having its own formal identity and unique function, conveying its own merely theoretical importance. (In some languages, moreover, more attention "meaning", which is an ingredient of the "total effect". The most widely is paid to the effect of certain levels than others, e.g. the use of onomatopoetic recognized levels derive from the structuralist techniques of the 1930s-50s: effects (phonetic sound symbolism), or the use of morphologically complex phonetics (and, for the study of written language, graphics), phonology lexical items, or the use of certain patterns of word order: preserving the (graphology), morphology, syntax (these last two are sometimes subsumed cultural values of these formal features is a regular problem for the trans- under one heading, grammar), lexis, semantics. A brief example of each lator.) follows: (B) Complexity of language function. The notion of equivalence involves (i) phonetic equivalence: where there is an attempt to create auditory or not only correspondences between formal pattern and cognitive (i.e. referen articulatory equivalence between the sounds of source and target language tial, or denotative meaning), as outlined above, but also correspondence in (Ll and L2 respectively), e.g. preserving the sonority of certain consonants or the situation of use. One has to take into account both the type of extra vowels in a text, the disjointed rhythm of a line, etc.; cf. graphic equivalence, linguistic situation in which an utterance is used, and theji-equency of its use which would depend upon such variables as layout, type-size, colour, etc. in that situation. Thus a style of English may be distinguished by two (ii) phonological equivalence: in terms of the way in which the units of the features (X, Y) shared by no other English style (for example, in one kind of sound system of a language are used to structure the text, e.g. attempting to religious English, thou and 0); but of the two, X may be a more frequent preserve the alliteration or rhyme scheme of Ll in L2. feature than Y, and thus a more important defining characteristic of the (iii) morphological equivalence: preserving equivalence of complexity of style. Work in sociolinguistics and stylistics has shown that the notion of a word structure, e.g. in terms of roots, affixation, compounding, etc., learned homogeneous language is chimerical: there are variables of language, rather, compounds in Ll being preserved as learned compounds in L2, etc. which are restricted to types of social situation, and labelled variously (e.g. (iv) syntactic equivalence: in the use of grammatical categories, sentence "dialects", "registers", "genres", "restricted languages"). An important types, word order, etc., e.g. preserving a set of tense contrasts in a text. index of the acceptability of a use of language, accordingly, is its appropriate (v) lexical equivalence: between the meaning of the lexical items ofL1 and ness to its situation. An utterance may be grammatically, semantically, and L2 (idioms as well as words), this being defined in some precise way, e.g. in phonologically "correct", i.e. permissible, but inappropriate, e.g. formal terms of semantic "components" (as in Nida, 1969) or meaning relations language in an informal situation, and other antinomies, e.g. conservative v. (as Lyons, 1968). radical, polite v.rude, technical v.expository, male v.female, metaphorical v. (vi) semantic equivalence: in terms of the overall meaning of the utterance, literal. The linguistic exponence of the various social and literary purposes not solely in terms of its component lexical items, including, for example, available to a language-user is only now beginning to be studied, using whether there is an antithetical point being made, or a contradiction, or a sociolinguistic techniques, and the true complexity of the situation appre metaphor (see further, Beekman and Callow, 1974; Nida, 1975). ciated. Taking sociolinguistic equivalence into account in translation This kind of analysis, with minor differences, may be found in the work of practice is, however, something which is rarely done systematically and Firth and Halliday, and is explicitly related to translation theory by Catford. accurately, in view of the absence of basic empirical information about the Other scholars have different views as to the optimum number of levels to detail of social variation within and between languages. Fairly general recognize, and how they should be defined and interrelated. A separate level categories are recognized (e.g. myth), but they are often restricted to a of "discourse" equivalence (viz. inter-sentence, or paragraph equivalence) is specific language background (e.g. Hebrew oral tradition) and we lack any currently attracting interest, for example. But the basic insight remains: that general typology of discourse to use as a framework for establishing cross an utterance's meaning is not a single homogeneous phenomenon, but a cultural similarities. But the principle seems clear enough, even if work at synthesis of various elements, the relative importance of which varies from present is inchoate and intuitive. Thus if a text is written in a very formal one situation and language user to another, e.g. for one purpose, preserving (archaic/metaphorical ... ) style, one would maintain a principle of stylistic alliterative patterns may be a major concern; for another, this factor may be equivalence by requiring that the L2 text should also be formal (archaic/ unimportant. The important point to note is that these level equivalences are metaphorical ... ). Note that I am here talking only about equivalence, not always to some degree mutually exclusive, e.g. phonological equivalence is acceptability. If, after translating (say) a formal sixteenth-century liturgical 4' 326 THE BIBLE TRANSLATOR (VOL. 27NO. 3 JULY 1976) CURRENT TRENDS IN TRANSLATION THEORY 327 text into formal twentieth-century English, one does not like the result, then between persons is also an act of translation, in that no two idiolects and one is at liberty to change to a less formal style-but this would then no sense experiences are identical. But here the notion of translation seems to longer be a question of translation, in the usual sense, but of synchronic have been broadened until it is indistinguishable from that of "communica stylistic choice. tion", and its utility is questionable. The viewisoften defended, however. Other aspects of stylisticmeaning have also been in the forefront ofrecent It isevident that it isimpossible to have total translational equivalence, in discussion, e.g. Leech, 1973. Of particular importance is the notion of the full formal and functional senses of above, and this is of course not a institutionalized (as opposed to idiosyncratic) connotation, e.g. the "snarl" novel conclusion. Its corollary is, however, less evident, if recent popular and "purr" words (to use Hayakawa's phrasing), which arouse opposed discussion of the nature oftranslation isanything to goby-namely, the fact emotional reactions, e.g. "productivity" is (at present) a "good" word, that there are many legitimate translations ofa particular text, depending on "provincial" a "bad" one. The existence of such features, usually extremely the emphasis the individual translator (or team) places on the separate culture-restricted, further complicates the notion of equivalence: their variables. Instead of talking about the need for an acceptable translation, pervasiveness in language, moreover, has usually been much under accordingly, one needs to think in terms of kinds of acceptable translations. estimated. There are many possible approximations to the idealized notion of a "best" Within this general area, cultural equivalence is sometimes distinguished translation, and the problem facingthe translator isthus to make the alterna separately from the above, being partly a semantic and partly a socio tivesasclear aspossible, sothat the users ofthe texts willbeenabled to make linguistic matter. This refers to the use of linguistic terms, structures or ajudicious and informed selection. It isthis flexible and pragmatic attitude sounds to evoke a highly specific cultural response, e.g. "bread" implies towards the acceptability of atranslation that informs much recent academic "staple diet" in L1 context as opposed to "rice" in L2, "fish" in L3, etc. thinking in this area. This emphasis is, however, at odds with the puristic This problem is central to any aim of dynamic, behavioural equivalence of normativeness which characterizes popular discussion on this topic, where texts, and it isofcourse a problem which has beenwellinvestigated byNida, prescriptions and proscriptions of the most unyielding (and uncharitable) the Summer Institute ofLinguistic groups, and others. A recent slant on this kind abound. There are many stereotypes in general use relating to what an topic involves the notion ofpresupposition-what may be taken for granted acceptable translation should be like, some of which have been part of our in formulating an utterance, because one can assume knowledge of it on the literary traditions for a very long time. Moreover, each culture has its own part ofthe hearer. In terms oftranslation, theproblem isthat presuppositions stereotypes; each language has its own hieratic varieties which have shaped are not always shared between L1 and L2, and one has to decide (to quote its speakers' intuitions about prestige, propriety and correctness. Some Nida, 1969)"How what is said fitsinto what isnot said", and how much to cultures display more readiness for linguistic change, more tolerance of add to the text for the L2reader. linguistic variability, than others. One of the most fundamental tasks facing As with the question of formal equivalence, total functional equivalence the translator, therefore, is to understand the cultural attitudes which have would also seemto be a theoretical impossibility. From a scientificpoint of givenriseto such stereotypes, and to attack those which have ledto inflexible view,it seemsimpossible to verifythe intuitions about language function and ways of thinking. How this isbest done is obscure, for the very existence of status held by the native bilingual (cf. the speculations of Steiner, 1975).It the problem has been but recently recognized. But now that it has, it isto be seemsone must be satisfiedwith approximations, avoiding the most obvious hoped that more informed discussion willtake place. blunders. What then are these stereotyped phrases and attitudes with which the One point whichemergesfrom the focuson thedetail oflanguage structure notion of acceptability is in conflict? Here are some brief examples. The and function is the parallelism between translation in the sense of LI-L2, Roman Catholic International Committee on English in the Liturgy cites, as and that of Ll-Ll. Ll-Ll translation may be seenboth diachronically (e.g. a criterion for an acceptable translation, the need to remember the tradition rendering Elizabethan English into Modern English) and synchronically of English devotional writing. This means, for example, bearing in mind (e.g. rendering legal English into everyday English). The processes seem when certain linguistic formulae have been sanctioned by generations of essentially the same, and the notion of "translation" has recently been usage so that they have achieved a favourite place, soto say, inthe minds of applied with somecogencyto the study of both-a point made, for example, the Church community. This point is an important one, but it is often by Firth, and now the keynote of Steiner, 1975.It isillustrated by the recent ignored, as in the "thou/you" controversy. It isnormally assumed here that increase in Bible "translations" for restricted groups, such as Carl Burke's the question of whether "thou" or "you" should be used will have a single prison-orientated texts. The logic of this extension has one point in its answer; but this ismost unlikely, as it ignores the issueof appropriateness in favour: diachronically, the continuum Ll-Ll leads ultimately to an LI-L2 context. To argue that "thy" should be replaced by "your" in all contexts relationship (e.g. English into Germanic). There is also one point against: produces different degrees of acceptability in the results: for whereas it is synchronically, the notion of group varieties reduces ultimately to the relatively easyto persuade people to accept a change ofpronoun from "thy" language system of the individual (idiolect). In a sense, communication to "your" in a sentence like "I know thy name", it will be much more 328 THE BIBLE TRANSLATOR (VOL. 27 NO. 3 JULY 1976) CURRENT TRENDS IN TRANSLATION THEORY 329 difficult to effect a change in "thy kingdom come", the latter phrase having the expectations of the majority of the intended audience. The world is full been hallowed by usage, so that it has become, in a sense, proverbial, of excellent, novel translations: what is lacking often is popular awareness of resistant to change. the need for them. Which suggests that, in order to bridge the gap between As a second example, consider the frequent references made to the translation theory and practice, a new educational emphasis should develop, "demands of corporate public worship", which it is said should influence the in much the same way that applied linguistics bridged the gap between choice of liturgical style. But this notion is extremely obscure. It has some linguistic theory and foreign language teaching, by spending many years in times been interpreted to mean that the language should avoid the use of educating teachers to see the point of changing their techniques. Such a field specialist theological terms and archaisms. But what are such terms? Terms -of "applied translation studies", perhaps ?-would integrate psycho like "con substantial" and "incarnate", which have been attacked, look linguistic and sociolinguistic principles and techniques in order to evaluate complex, but it can readily be shown that there is no necessary connection the assumptions and attitudes of the translation consumer. I hope very between word length and conceptual difficulty. This links up with the much that the field will be invented before the end of the 1970s. demand in correspondence columns for "simplicity" of liturgical style, a stereotyped phrase which seems to mean "intelligibility without effort" BIBLIOGRAPHY hardly a desirable precondition for intelligent liturgical participation! As Earr. l., The Semantics ofBiblical Language, Oxford University Press, 1961. Beekman, J., and J. Callow, Translating the Word of God, Zondervan Corporation, 1974. has often been pointed out, liturgical language is not meant to be viewed as a Burke, C., God isfor Real, Man, Fontana, 1967. self-contained set of utterances: the prayers need further explanation, which \ Catford, J.C., A Linguistic Theory of Translation, Oxford University Press, 1965. it isthe function of other liturgical events, such as the homily, to perform. It Crystal, D., and D. Davy, Investigating English Style, Longman, 1969. :-::true that the language should be familiar, but this is not the same as Leech, G. N., Semantics, Penguin, 1973. Lyons, J., An Introduction to Theoretical Linguistics, Cambridge University Press, 1968. saying it should be immediately comprehensible. "Blood of the covenant" Nida, E. A., and C.R. Taber, TheTheory andPractice of Translation, London, UBS, 1969. may be obscure, but that is not because it is technical. Nida, E. A., "Semantic Components inTranslation Theory". In G. E. Perren and J.Trim As a last example, it is often said that a translation of religious subject (eds.) Applications of Linguistics, Cambridge University Press, 1971,pp. 341-8. matter should display "dignity", "consistency", "euphony"-but with little Nida, E. A., "Translation". In T. A.Sebeok (ed.)Linguistics andAdjacent Arts and&iences (Current Trends in Linguistics, vol. XII), Mouton, 1974. effort being made to determine exactly what such terms mean. What features Nida, E. A., Exploring Semantic Structures, Munich: Fink Verlag, 1975. of language constitute the physical correlates of "dignity" or "euphony"? Palmer, F. R. (ed.), Papers ofJ.R. Firth, Longman, 1972. The latter, for example, has been defined as language which is "suitable for Sleiner, G., After Babel, Oxford University Press, 1975. praying aloud", "suitable for singing", or simply "beautiful" and "har monious" (all these glosses are taken from recent introductions to liturgical translations). This last pair of glosses is impossible to turn to practical utility, for obvious reasons to do with the relativity of personal value judgements. But even the first two definitions are difficult to make sense of. How does one decide what is suitable? What does one listen for, and who are the judges? Remember St. Benedict (in chapter 38 of his Rule): "The brethren are not to read or sing each in his turn, but only those who give edification to the hearers"! These last paragraphs are an attempt to illustrate the complexity under lying some of the commonplace evaluative phrases used about the accepta bility of a translation, phrases whose meaning has for too long been simply taken for granted. A careful investigation of their use is long overdue: it is easy to remain at cross-purposes without it, for example by agreeing to the use of a term without sharing its meaning. Readiness to look critically at the metalanguage of translation, I would argue, is a prerequisite for progress in this field; and there are now many indications that this critical attitude is developing among theoreticians and practitioners of translation. But above all, there is a need for a realistic awareness, on the part of translators, of the psycholinguistic and sociolinguistic attitudes towards language (and thus, any translation) which predominate in the minds of his audience. There is little point in producing a fine, "professional" translation, if it clashes with