The Translation Process and Translation Procedures

14
84 ~ Nevertheless, the hopes pinned on the use of wide-ranging deductive meth- ods, hopes raised in an extreme form by Revzin/Rozencvejg (1964), remained unfulfilled. This was because translati.on, like every form 9f performance, isa semiotic phenomenon which cannot be fully described: in the form of an abstract model and is therefore only partially accessibl~ to a deductiv,e approach. Formalized hypothesis formulation has no doubtibeen important for the development ofthe science of translation, because f?~maliZed models are a much sought-after status symbol ofthe dignityof '~ny science, but it seems that the science of translation has reached,if riot exceeded, a theoretical saturation point. Apparently, at presentthe prer:~quisites for the / development of more finely differentiated models of thetranslation process, which would capture the wealth of concrete phenomen'a in translation with some measure of completeness, are lacking. Because tif the difficulties involved in ,formulating a comprehensive theoretical fram~work for trans- lation, many a contribution to the discussion of translation theory has done little' more than replace, the terminology of earli~r formulations, without notably adding to what,is ,known of translation theory or further developing explanations of the interlingual transf'er conbept. In a way, translation theory has thus confirmed what Friedrich Withelm Nietzsche said"namely, that we should distrust claims that profess to work out a well-rounded whole by means of abstract reasoning". The slowdown of the last few yearsin the formulation of hypotheses in the science of translation has, of course, proved to be an advantage, since it has largely spared the science of translation a Darwinian struggle between various models, but also has played a major role in seeing the science of translation recently turn its attention tothe detailed description of ,concrete translation procedures, Such a description falls within thearea covered by the language-pair-oriented descriptive science of translation. The problems of this field will be discussed in the following three chapters. The specific science of translation has the task of analyzing concrete events in translation. It sees its justification in the investigation of modes of behavior in translating, thus empirically examining the models of the' translation process developed by the general science of translation; its, aim is'to bring about a successful crossover from a concept of the translation process rooted in the science of communication'to one which takes a linguistic approach. Contrary to the general science of translation, it has the function of s~tting \ up language-pair-oriented paradigms of translation (Levenston 1965). It I achieves this complementary target by looking into the procedural and textual factors and regularities operating in individual processes oftrans- lation. The specific science of translation has therefore derived a data-oriented, language-pair-specific program of research. This program is conducted with the aim of formulating methodological principles for the following three areas of research: ' Q the systematic description, classification, and explanation of lan~uage-( , pair-specific translation procedures. Within this context the language, r---./ pair-oriented science of translation is concerned primarily with those \ " syntactic, lexical, and sociocultural phenomena in a given language I IV') which do not have one-to-one correspondences in the system of another U language and must therefore be translated by means of compensatory, non-literal transfer procedures; 2. the development of a text-typology relevant to the science oftrans-- lation. The twoprincipal research areas here are, first, the defining I rl of textual segments and structures on the microcontextual level (withi n \ the clause/sentence) and the macrocontextual level (going beyond Q the clause/sentence rank) and, second, the testing oftextually adequate transfer strategies, whilegiving consideration to text-syntactic; textJ semantic, and text-pragmatic factors; " 3. the development ,of objective - or at least intersubjective -' yardstick;l for the assessment of the degree of TE with the specific goal of replacing I? the predominantly intuitive judgement of translation qualityby e~Plici~ a textually based evaluations of TE. . i All three areas of research are presently in statu nascendi, with1. sPghtly ahead of 2. and 3. slightly ahead of 3.

description

for translation students

Transcript of The Translation Process and Translation Procedures

Page 1: The Translation Process and Translation Procedures

84 ~Nevertheless, the hopes pinned on the use of wide-ranging deductive meth-ods, hopes raised in an extreme form by Revzin/Rozencvejg (1964), remainedunfulfilled. This was because translati.on, like every form 9f performance,is a semiotic phenomenon which cannot be fully described: in the form ofan abstract model and is therefore only partially accessibl~ to a deductiv,eapproach. Formalized hypothesis formulation has no doubtibeen importantfor the development of the science of translation, because f?~maliZed modelsare a much sought-after status symbol of the dignity of '~ny science, butit seems that the science of translation has reached, if riot exceeded, atheoretical saturation point. Apparently, at present the prer:~quisites for the

/development of more finely differentiated models of the translation process,which would capture the wealth of concrete phenomen'a in translationwith some measure of completeness, are lacking. Because tif the difficultiesinvolved in ,formulating a comprehensive theoretical fram~work for trans-lation, many a contribution to the discussion of translation theory hasdone little' more than replace, the terminology of earli~r formulations,without notably adding to what,is ,known of translation theory or furtherdeveloping explanations of the interlingual transf'er conbept. In a way,translation theory has thus confirmed what Friedrich Withelm Nietzschesaid "namely, that we should distrust claims that profess to work outa well-rounded whole by means of abstract reasoning".

The slowdown of the last few years in the formulation of hypotheses inthe science of translation has, of course, proved to be an advantage, since ithas largely spared the science of translation a Darwinian struggle betweenvarious models, but also has played a major role in seeing the science oftranslation recently turn its attention to the detailed description of ,concretetranslation procedures, Such a description falls within the area covered bythe language-pair-oriented descriptive science of translation. The problemsof this field will be discussed in the following three chapters.

The specific science of translation has the task of analyzing concrete eventsin translation. It sees its justification in the investigation of modes ofbehavior in translating, thus empirically examining the models of the'translation process developed by the general science of translation; its,aim is'to bring about a successful crossover from a concept of the translationprocess rooted in the science of communication'to one which takes alinguistic approach.

Contrary to the general science of translation, it has the function of s~tting \up language-pair-oriented paradigms of translation (Levenston 1965). It I

achieves this complementary target by looking into the procedural andtextual factors and regularities operating in individual processes of trans-lation.

The specific science of translation has therefore derived a data-oriented,language-pair-specific program of research. This program is conducted withthe aim of formulating methodological principles for the following threeareas of research: '

Q the systematic description, classification, and explanation of lan~uage- ( ,pair-specific translation procedures. Within this context the language, r---./pair-oriented science of translation is concerned primarily with those \ "syntactic, lexical, and sociocultural phenomena in a given language I IV')which do not have one-to-one correspondences in the system of another Ulanguage and must therefore be translated by means of compensatory,non-literal transfer procedures;

2. the development of a text-typology relevant to the science of trans--lation. The two principal research areas here are, first, the defining I

rl

of textual segments and structures on the microcontextual level (withi n \the clause/sentence) and the macrocontextual level (going beyond Q

the clause/sentence rank) and, second, the testing of textually adequatetransfer strategies, while giving consideration to text-syntactic; textJsemantic, and text-pragmatic factors; "

3. the development ,of objective - or at least intersubjective -' yardstick;lfor the assessment of the degree of TE with the specific goal of replacing I?the predominantly intuitive judgement of translation quality by e~Plici~ atextually based evaluations of TE. . i

All three areas of research are presently in statu nascendi, with 1. sPghtlyahead of 2. and 3. slightly ahead of 3.

Page 2: The Translation Process and Translation Procedures

86Intensified empirical researc'h ~,asproved necess~ry for tW~ reasons: •

[

1. -Extensiv~ corpor.a of v~rious types of texts ~ust b~ studied from thethree POints of view set forth above, if we are to lellrn to what extentit is possible to develop more complex, semiotic translation processmodels, going beyond stating merely the initial co'nditions of trans-lating. ' '

[

the applied .'science of translation is vitally dependent on the resultsof t~e langUage-pair:,oriented descriptive science of translation (just asforeign language pedagogy is dependent o:n the resL,Jlts of contrastivelinguistics) if it is to provide the necessary preconditions for moreefficient TT within the framework of university curricula designedfor future translators. :

af the three above-mentioned areas of research, the linguistic description andp the functional analysis of transfer phenomena, including the' tonditions

constraints, regulariti.es, ra~ific(ltions and ~he habitual el~ments of transfe;o proced~res, must logically precede the other two. Transfer procedures may

occur In any text to be translated ~nd they are always TE-interconnected.

Translation research has devoted considerable energy to the analysis of thetransition from an SL T to a TL T and how observable transfer proceduresc~n be organized in a plausible and lucid way; as a result, it has come upWith a number of differerit classification systems which more or less ex-haustively reflect the diversity of modes of behavior in translating,

The simplest distinction is the one made between lite~al and non-literaltranslation procedures. It differs to some extent from the distinctionfamiliar to us from our study of the history of translatidn theory, betwee~literal (true-to-the-word) translation and free translation; In the history oftranslation, to translate literally or to translate freely amounts to a basicdecision on the methodologY."of translation which com~its the translatoreither to an SL-oriented, rejrospective, or to a TL-oriented, prospectivetranslation approach. No matter which approach he chooses, it is in anycase a decision which determines the translation profile bf the entire text.The highly normative character of the two concepts literal translation andfree translation explains why they have never really been able to gain afoothold in the terminology of modern descriptive tr,anslation researchwhere they have been replaced by the conceptual pair, literal translationand non-literal translation. These terms designate contrete, linguisticallyanalyzable transfer procedures working straight from $L surface to T Lsurface or c.hanging the SL surface stru.cture syntactic.al.ly ~nd/or semanti.cally according to TL needs In a way which cal! be speclf'rd In each particu-lar case. I, \ I

" I

!

Thls.,rough·division between literal and non-literal translation proceduresseems at first glance to be plausible, but a closer look reveals serious prob-lems of definition and fluctuations in concept. This is especially evidentin the area of literal translation, where the lack of sufficient differentiationwithin the field means that there is often not a c,lear enough conceptualdistinction between literal translation (wortliche Obersetzung,' traductionlitterale, traduction directe) and word-for-word/word-by-word translation(Wort-fij;-Wort-Obersetzung, traduction mot a mot;. Thome 1975). Vinay/Darbelnet (11958), Govaert (1971 b), and Vachon-Spilka (1968), for ~xample,use the two concepts word-for-word translatidn and literal tr~nslationsynonymously. The idea that the two are one and the 'same concept ispresumably furthered by the fact that in the practice of translation theborderline between word-for-word translation and literal translation - likethe borderline between literal and non-literal translation - is constantlybeing crossed by the translator:

" ... the 'rough-draft' or 'word-for-word' method ... in fact rarely does prese~nan actual word-for-word e~uivalence, but rather moves up and down the rank-scale,. 'V'translating sometimes word, sometimes group, occasionally a mere morpheme or 0a whole clause at time" (Ure et al. 1969, 73l. .

Taking up the traditional distinction between literal and free translation,Hockett and Chao have the following to say:

"The terms 'literal' and 'free' thus do not really form a clear binarY contrast. A word-by-word rendering is literal as compared with a loose translation of a whole sentence.but free as compared with a morpheme-by-morpheme rendering. I·t may be proposedthat. for any given passage, there are as many degrees of literalness and freedom oftranslation as there are levels of hierarchical structures in the passage" (19,54, 3131.

"A common distinction is often made between literal or word-for-word translationand idiomatic or free translation. Bu~ there are more than just two degrees on thescale of literalness and idiomaticity. If we go below the level of the word, there canalso be morpheme-by-morpheme translation, while if one 'tries to translate proverbby proverb, there is often no corresponding internal structure at all" (196~, 507 f,).

Vachon-Spilka finds that

"Literal translation is the easiest and simplest form of translation, it occurs wheneverword-by-word replacement is possible without breaking rules in the target language;this, however, Is quite rare unless the two languages are very closely related" (1968,18 f.l.

Equating the concept of \f;v'ord-for-word translation with that of literaltranslation is an ill-chosen solution from the standpoint of methodology,primarily. because there is a difference between th¢ two translation proce- /dures which is essentially irresolvable: word-for-word translation, which (must also include the interlinear version familiar from the Middle Ages(see chapter I f I), follows the syntactic structures of the SL, while preserving

Page 3: The Translation Process and Translation Procedures

88semantic TE between the SL and TL ~egmel1ts; literal' trans;ation, on theother hand, follows the system of syntactic rules' '(on the level of systemand norm) found in the Tl, while again preserving semantic TE betweenSl and TL textual segments (Thome 1981). Rendering the English sentence(1) I have read the book.(2) Ich habe das Such gelesen.

is a literal translation, since the translator, while retaining the Sl clauseconstruction, has changed the sequence of the individual Jwords of thesentence in accordance with the TL syntactic rules. The translation(3) "Ich habegelesendasBuch,

on the other hand, would be a word-for-word translation,! each word inthe Tl being the exaxt replica of the corresponding word in;the SLclause.The following remarks by Catford should also be seen in this light:

"The popu!ar terms free, literal, ana wqrd-for'word ,translation, though loosely used,partly correlate with the distinctio'ns dealt with here. A free tran*lation is alwaysunbounded. - equivalences shunt up and down the rank-scale, but tend to be at thehigher rank - sometimes between larger units than the sentence: Word-far-wordtranslation generally means· what it says: I.e. [sic! J is essentially rank'bound at word-rank (but may include some morpheme-morpheme equivalences). Literal translationlies between these extremes; it may start, as it were, from a word-for-word trans-lation, but make changes in conformity with TL grammar ... ; this may make it agroup-group or clause-clause translation" (1965, 25), i

The structural difference .between th~ English sentence (1) a$d the Germansentence (2) can be illustrated in the following way: I

S i

I have readIch habe gelesen

the book.das Buch.

!S = sentence, Aux = auxiliary verb, NP = noun phrase, P2 = 2nd participle. VP = verbphrase, Art = article. of,.

Incidentally, (3) is an utterance representing a syhtactic sttucture whichGerman native speakers can often hear, when an English native speaker ex.presses himself in German. The English native speaker is so us'ed to buildingthe sentences on the subject-predicate-object (SPO) model (see chapter III) _

in other words, he has internalized the SPO sentence-building concept tosuch an extent - that he automatically transfers this model into foreign-language communication, thus producing a syntactic interference fromhis own language in speaking German: the native-tongue syntactic structureis superimposed onto the foreign-language syntactic structure_

Now it is interesting to see, that a German native speaker, in hearirg anutterance such as (3), perfectly understands this clause despite its deviationfrom German syntactic rules. It follows from this that utterance under-standing is not so much dependent on grammati~al' correctness but onother features. What features are here being referred to can be made clearby presenting another couple of sentences introduced by Chomsky: whohas set himself the task of finding out the linguistic factors controllingsentence production and sentence perception:

(4) Colorless green ideas sleep furiOUSly.(5) Healthy young babies sleep soundly.

Both sentences are orgaflized along the same syntactic principle, eachsentence containing a subject phrase, a predicate phrase and an adverbialphrase. Yet, against the background of our language competence and ourexperience in producing and receiving sentences, it is obvious that only(5) makes sense, whereas one would probably hesitate to recognize (4)as a semantically acce'ptable sentence, because one does not see how onecan sensibly combine "colorless" and "grElen", "colorless" and "green"on the one side and "ideas" on the other, and "sleep" and "furiously".This is so because the semantic relations between the individual words in(4) are logically not coherent or plausible. The situation regarding (5) isentirely different. Here the language user is, metaphorically speaking, onhome ground. This is a clause. which is in line with his experience, whichmakes sense, because he knows from his world knowledge that youngbabies normally do sleep soundly. Exceptions confirm the rule., Thus,although both sentences are syntactically acceptable, (4) is rejected forits lack of semantic verifiabil ity.

Both sentences could be translated literally into German:

(6) Farblose grune Ideen schlafen wild.(7) Gesunde kleine Kinder schlafen tief.

logically speaking, it is the same situation here: (6) for a German nativespeaker sounds just as strange as t~e English equivalent for the Englishnative speaker; on the other hand, a German native speaker has no diffi-culty identifying the meaning of (7) and to accept it as a sentence whichmay occur any time in every-day communication. The result of translating(4) and (5) into (6) and (7) respectively shows that the English and the

Page 4: The Translation Process and Translation Procedures

""

German nativ~speaker sf/ate· :~ertain ·se,mantic corcepts while rejectingothers,'a fact which,is very important in the argument about translatabilitya~d untranslatability and settling this" argument, at' least in.!principle, infavor of interlingu.altranslatability. . ,\ '

Occasionally, literal translation and word~for-word t~a,nslation kre identical:

'('8) Where is he? .'Wo ister?, .. ( 9) He is he~e,,· Er ist hier. J, •

(10) He'will probably come: Er wird wahrscheinlic~ kommen,(11) His atJitude is disgusting. " , " Sail) Verhalten 1st wid\lrlich.(12) They were agreeably surprised, "', ' '" Sie wa'ren~ngenehrn uierrascht.

In languages which'a;ewid;ly div~rgent typ~;'ogiC~II~:'for ~x~mple Indo-· Ge:ma~ic .Ianguag~s and. Ame~i,can Indian languages;,' word-f?~-word trans-lation ,'S vlrt~ally Im~sslb~~: .... to 'e~press :t-le has a vit~ite j~cketon', theQagak. says literally he with Jacket, with-white: or 'he Jacke~y whitey' ••(Whitney 1870, 338). -' . " . i'Even in languages which,are closely related typologically, word-for-wor,dtranslation, as a rule, is',an'inad,e9uate transl.ation ~procedur~ w,ith regardto the T L. ..', , . l"The most Hte,ral type oftransla~ion, an inter'line~r cine; cali scarceiy b~ called a trans-

,'.ationin the usual sense ?f·the term.~However"some·productior\s'initendlld a~ fully· aCcfedited, translations ,are alrnost a.~ absurdly Iite.ra I as arlint,erlineari riln.?ering ..., Moreover, they have attemptedto.match gr<lmmatical forms.and even'to employ the· same word order, if, at all possible., The r~sults arelam'entable, for ihe attempt' tobe' literal in the form of the message has'resu)ted'in'grievous distortions bf the messageitself" (Nida 1964, 23). ", _ ,;

Still, the distortion of a' text's content' criticized by'Nida is a factor whichdoes 'not weigh as heavily in sorn(l types' of texts as ,'in others. It is clear, ,for example, that word-for-word translations have no place in biblicaltranslation, since for'T''a1 shortcomings 'could adversely'affect ther~ceiver's

rinterest in the Bible's 'message. In LSP communication on the other hand

, t,he receiver's understanding ,Of. the.· content, W,OUld,n~,t· nece~sarily $uffe;because of formal defects. ThiS led, MT to initially draw the conclusion

, that a translation produced by means: of word-far-word MT proceduresis better than no translation at all. :

If the minImum standard of quality is set on the basis of. such reasoning,no 'objection can be made to using word-for-word translation as a make-shift to expedite international communication among ~xperts. The sameis true when translation serves as a heur(stic methodological aid,'!-for example,in clarifying interlingual structural differences: .. ,

(13LGestern hat der Mann in'der Bibliothek ein Buch gestohlen.(14) "Yesterday has the man in the library a book stolen.

The comparison of.the&e t~o.c1auses makes clear that the English. languagein its inventory of grammatical rules differs from that of;German. In orderto produce a grammatically acceptable sentence in English, the Germantranslator must know the grammatical system of English. If he knows theserules, he 'will probably come for~ard with translations listed un~er (15)

or (16): "

(5) Yesterday the man stole a booki~th~ library.(16) The man stole a. book in the library yesterday.

What has been said just now becomes even c1eare~ if one looks at thefollowing exa~pl~s '(Wilss 1973):, "

(17) Oer Konig zog in'die reich mit'bunten Fahnen geschmu~kte Stadt elin.(18) *Theking entered the richly with colorful flags decorated town.(19) The ""king entered the town (whiSh vilas) richly decorated with colorful flags.(20) Oer Konig zag in die' Stadt ein, die mit buntenFahnen' reich geschmuckt war.

(17)' represents a frequent German syntactic pattern which cannot bereproduced in English on th'e basis of a literal translation.

A version probably acceptable t.0 an English native spe,aker is listed under(19). This sentence in its.turn can be re-translated literally into German.This permits the conclusion thatifl the realm of ce~tain syntqctic con-figurations, German is more flexible than English (just,~s English is moreflexible in others). ThiS is important for the English'German translator,because in order to render' (19) into German, ,he has the choice betweenat least two alternatives (f7)' and (20); the English-German translation'here reaches the dimensionQf a ~'decision process" (Levy 1967).

In contrast to word-for-word translation, literal translation, which is setup bidirectionally (not unidir~ctionally, as is the case with non-literaltranslation) is a,translation procedure which, under certain circumstances,is totally a'cceptable stylistically. 'Assertions to the "contrary derive fromconfusing word-for-word translation with literal translation. Cermak, forexample, is quite mistaken in his reading of the situation:

"La traduction sous-interpn\tee, litterale est Ie genre de traduction dont on dit que sonauteur n'a aucune idee du texte lorsqu'il Ie traduit. Naturellement, il y a dans unetraduction toujours un minimum d'interpretati.on. L'exemple exagere d'une traductionsous-interpretee sera it la traduction servile, glosee, la traduction ou les elements lexicaux50nt implantes isolement" (1970,34 f.).

Every translator knows that 'there may be language contact situationswhere he must translate literally because no genuine alternative to a literaltranslation procedure exists. From this it follows that the traditional dis-tinction between literal, SL-directed translation and idiomatic, TL-directedtranslation cannot be upheld, at least in this undifferentiated form, andthe dichotomy, going back to Nida, between "formal equivalence" (= SL-

Page 5: The Translation Process and Translation Procedures

92oriented translation) and ,"dynamic equivalence" (= TL-oriElOted trans-lation) takes on added importance. Nida characterizes a translation whichaims at formal equivalence as follows:

" ... a formal equivalence (or F-E) translation is basic'ally source-ori~nted; that isit is designed to reveal as much' ,aspossible.of the form and content of the originalmessage.In doing so, an F-E translation attempts to reproduce severalfo~malelements,including: (1) grammatical units, '(2) consistency in word usage,and! (3) meaningsin terms of the 'source context. The reproduction of grammatical uni~smay consistin: (a) translating nouns by nouns, verbs by verbs, etc.: (b) keeping all phrasesandsentencesintact ... ; and (c) preserving'all formal. indicators, e.g. marks of punctu-ation ... " (1964,165). '

Remarks made by Nida and Taber on ~his subject some years later weremuch more pointed:

" ... formaf correspondence [is thel quality of a translation in which the featuresJ of the form of the source text have'been mechanically reproduced in' the receptor

/

,angUage. Typically, formal correspondence distorts the grammatical, and stylisticpatterns of the receptor. language;and hence distorts the message',so as'to' causethereceptor to misunderstandor'to labor unduly hard' ... " (1969, .203). ,

Vinay, too, considers it necessary to distinguish between'traduction littf!ralefautive and traduction litterale corrective (1968); equati ng, traductionlitterale fautive with mot a mot translation procedures, however, doesnot do justice to the complex subject matter of literal translation.

According to Rabin, the real problems involved in literal translation haveto do with semantics:

"The real problem lies in the fact that the individual semantic items .... arenot com-mensurate. The way around this which is open to the dictionary maker; of providingseveral renderings or of adding circumlocutions or definitions, is not feasible in acoherent translation, where, moreover, the words are often not used in their dic-tionary senseat all, but in the metaphorical, ~nlarged, pregnant, and other off-senseswhich the system of oppositions and interplay of meaningswithin the systempermits"(1958,134 f.>'The following pair of examples may serve as an illustration of semanticaspects of literal translation:

(21) Nixon hasno intention to resign.(22) Nixon hat nicht die Absicht zu resignieren.

It is obvious that (22) is a case of a "faux amis"-translation (Wandruszka1918).

This translation is wrong, because what is meant here was,that Nixonhad no intention to withdraw from his office ("zuruckt:eten"ln German),whereas the German translati.on implies that Nixon simply had ho intentionto allow himself to be demoralized by perpetual public pre:ssure. Now,taking up Rabin's line of reasoning, one must say that it is rot entirely

!

II

I II

convincing, since he - obviously in order to support his own point of viewin translation theory - raises the lack of interlingual lexical commensura-bility between SL and TL lexemes to the level of a general translationdifficulty (TO); he thereby overstates the case in a way which aoes notduly take into account the fact that interlingually the.meaning of manywords, above all in the LSP field, is identical. Moreover, he ignores thefact that translation problems stemming from word meaning also occurin non-literal translation, let alone the realm of. paraphrase translation(see below the discussion on Kade's term "new coding" (Neukodierung)).Catford shares the view that the translator working wlthin the frameworkof literal translation procedures tends toward direct lexical 'TE:

"One notable point ... is that literal translation, like word-for-word. te~ds to remainlexicallv word-for-word, i.e. to use the highest (unconditioned) probability lexicalequivalent for each lexical item. Lexical adaptation to TL collocational or 'idiomatic'requirements seemsto. be characteristic of free transiation" (1965, 25). '

As in connection with Rabin's view, here, too, the lumping together of thesyntactic and lexical aspects of literal translation causes one to questionhis findings. Correct as Catford ,may be in considering literal translationa phenomenon of syntax, the general assumption of interdependencebetween syntactic literalness and lexical literalness on the one hand andsyntactic and lexical non-literalness on the other, appears problematic.Whether or not syntactic literalness, as a rule, actually does entail lexicalliteralness (and vice versa?), whether or not lexical adaptations cannot beand are not also carried out in cases of syntactic literalness, and vice versa,and whether or not lexical substitutions also occur in, cases of syntacticnon-literalness - these are all questions which cannot be given quite so flatan answer. Catford seems to have fallen victim to the fascination of a"correlating hypothesis" of translation which does not stand up to empiricaltesting or does so only to a very limited extent.

In addition to its syntactic and semantic aspects, literal translation, likeall translation procedures, also has a stylistic aspect. This is illustrated byan example from Guttinger's book "Zielsprache. Theorie und Technik desObersetzens" (Target Language: Theory and Technique of Translating)(1963, 7 ff.). At issue is a warning sign observed by Guttinger at the mainrailway station in Zurich, which bears the following inscriptions:

(23) Es ist verboten, die Gleisezu uberschreiten.(24) It is forbidden to crossthe lines.

The English formulation is a 'literal translation of the Germanorigi nal,i.e. the translator - in the fashion of an automatic reflex - has carriedthe pattern of syntax followed in the German sentence over into TL,leaving all lexical correspondences unchanged. The result is a translation,

Page 6: The Translation Process and Translation Procedures

which, it is true, correctly conveys' M of SL and would be completelyintelligible to a native speaker of English, but which ignores the naturalstandards of adequacy in the TL - unintentionally, no doubt -.: and thusgives the text a foreign stylistic ring. For, that reason, the translation isadequate only in terms of syntax and semantics; stylistically, it is inade-quate, since no heed was paid during the translation process to the TLconditions of idiomaticity, and so the resulting TL formulation resemblesthe work of a beginner. This must be judged all the more critically' inthat ~e are here dealing with interlingually conventionalized set phraseswhich permit little or no deviation from the standard expressions onceestablished by a language community. Th~ translation could. only be re-garded as idiomatic if the translator hild departed from the syntcictic struc-ture of the SL sentence and produced one of the following TL versions:(25) No crossingthe lines (better: No crossingthe tracks). "(26) Do not crossthe lines (tracks)., .(27) Only authorized personnel is allowed to crossthe lines (tracks). .(28) Passengersare not allowed to' crossthe lines (tracks).

This means that in the present caseian adequate version, allowirg for thestylistic expectations of the English native speaker, is only p6ssibleonthe basis of a non-literal translation. :In order to specify the problem of non-literal translation, it is, advisableto take the following series of examples as a starting point for our discussion:(29) Everybody is talking of reconstruction.(30) Jeder spricht eben vom Wiederaufbau (printed version).(31) Mle reden vom Wiederaufbau (translation proposed by author, using a familiarslogan of German Rail: Aile reden vom Wetter, wir nicht: Everybody talks aboutthe weather, we don't).

The sentences (29) to (31) have a cqmpar~ble syntactic structur~, insofaras the~ .contain :hree syntactic units, a subject ~hrase, ~.predi~te, a~d aprepositional object; (30) and (31) are, accordingly, literal t~anslatlonsof (20) (for a different assessment concerning cases such as (3D),~ee Thome1981). Their classification as Iiteral transfer procedures is una]fected bythe fact that synonyms were used in the predicate of (30) and (31) and bythe clearly unsuccessful addition of "eben", meant to make tibe Englishexpanded form explicit. On the other hand, (30) is undoubtedly "moreliteral" than (31). This is because (31) changes the original construction,written in the singular, to one written in the plural. ,Nonethel~ss. jf syn·tactic isomorphy is held to be a criterion ,of litaral translatio'n Lthe typeand number of speech parts' must be the same in both senteric~s(Thome1975, 1981) -, it cannot be said that the-above-mentioned.chJnge bringswith it a change from the category of literal translation to th~ categoryof non-literal translation. ' , \ '

(32) It is impossible to solve the problem.(33) Es ist unmbglich, das Problem zu losen.(34) DasProblem ist unmoglich zu losen.

The translator in (34), having dropped "es" as a grammatical subject, hadto reorganize the sequence of speech parts. As a result, he producas aversion, clearly different from (33) in terms of surface structure. Therefore(34) cannot really be called a literal translation. On the other hand, onecould speak of a non-literal version only if "es" were te be interpreted asa speech part in its own right. which, in view of its syntactic function asa substitute, anticipating the meaningful subject phrase contained in theinfinitive clause, is unlikely.Even more controversial with regard to their classification as literal trans·lations are the following examples:

(35) He cannot come, for he is ill.(36) Er kann nicht kommen, denn!lr ist krank.(37) Er kann nicht kommen, weil er' krank ist.

Changing the SL parataxis of (36) to a hypotactic structure in (37) indeedcauses the relationship between the clauses to change, but hardly enoughto move (37) into the category of non-literal translation, unless one insists,like Thome (1981), that the transition from a sentence consisting of twoparatactically concatenated clauses to a sentence, consisting of a com-bination of a superordinate and a subordinate clause implies a shift fromliteral translation to non-literal translation.The following examples are again a border-line case, this time becausecontextually there may be a slight difference in semantic perspective in(39) and (40):

(38) I was nearly run over by a bus.(39) Ich bin fast van einem Busiiberfahren worden.(40) lch ware beinahe van einem Bus iiberfahren worden.

Apart from the fact that this distinction is rather speculative, as a testwith German native speakers has shown, it is by no means, at least in myview, so evident as to justify the assignment of (40) to the category ofsemantically motivated non-I iteral translation.Taken together, these examples bear out the need for establishing a trans·itional area between literal and non-literal translation or a hew specificationof the range of literal translation on'the basis of the examination of anextensive corpus of pertinent cases of translation, A t~ntative approachalong the latter line has been made by Thome (19811.

Page 7: The Translation Process and Translation Procedures

96fA clear crossover from literal to· non-literal trans.laiionr cannot ~e. said tohave taken 'place until the translator - for whatever reasons -: divorceshimself so completely from the syntax or' se'mantics of the SL fo~mulationthat the TL textual segment takes on its own unmistakable shape: .

"

(41) If resourcesexist only in limited quantities, we say that resources'are scarce.(42) Wenn Ressourcen nur in begrenztem Umfange vorhanden sind, SPrechenwir

von Ressourcenknappheit. .

Literal back-translation into English:. r I •

(43) If resourcesexist only in limited quantities,·livetalk about resources' sc~rcity.

A syntactic comparison of (41) and (42) shows that an objective clauseintroduced by "that" in the SL changes in the TL to the nominal compound"Ressourcenknappheit". The result of the transfer procedure chosen by thetranslator is what Catford terms an "unbounded translation" (m65, 25),which leaves the semantic perspective of the o~iginal sentence u~touched.Achange in semantic sentence profile occurs in the following exa~ples:(44) The well-being of all nations· dependsi,; .~~ small measureon' th~ Igeneration

and distribution of energy. ~I(45) Die Erzeugung und Verteilung von Energie'bestimmt in nicht geringe;mAusmaB

denWohlstand alll~rNationen. '!

Literal back-translation into English:(46) The production and' distribution of energy determines'in no small measurethe

well-being of all nations. IBe~ause of the change in the predicate, the sentence undergoes not oniya syntactic restructuring, but also a shift in semantic perspectivk from aneffect-cause relationship to a cause-effect relationship (if not enbugh elec·tricity, then no well·being; only if enough electricity, then well-being).The recourse to non-literal translation is 'not obligatory, but"'optional;more about this will be said below.

The first systematic attempt to draw up a well-ordered set of propositionshaving to do With interlirigual transfer processes arid to devise,! compre-hensive taxonomy of translation procedures encompassing both literaland non-literal translation was made by the representatives of stylistiquecomparee (VinaylDarbelnet 11958; Malblanc 11961). Seizing upon thedistinction first 'made by Bally (1944) between stylistique interne andstylistique ex terne , they developed a framework for describing translationprocedures which proceeds from the premise that the translation processmanifests itself as a series of linguistically. comprehensible procM~s techhi-ques. On the basis of an extensive discussion of examples from the twolanguage pairs English-French and German-French, stylistique compareecame to the conclusion that all translation procedures, at least in these two

language pairs, can be subsumed under seven main headings; the first three Ifall into the category of literal translation (traduction directe) and theremaining four into the category of non-literal translatipn (traduction .oblique):

C9 emprunt, i.e_ the carryover of SL lexemes or lexeme combinationsinto the TL normally without formal or semantic modification (Wilss1966). Direct borrowing English-German may entail capitalization:(47) know-how

brain-drainbrinkmanshipsoundtrackjet-setteenager'establishmentmarketingcompetitive coexistenceroll-backRock and Rolloverkill

containmentnative speakershowdownjointgho~t-writercontainerizationworkshoppattern drillpenthousepaperbacktalk show"small talk

poster~ylinecountdownmid life-crisisstylingfixercomputerinputoutputoverdrive

As these words become naturalized, orthographic and phonetic adaptationsin accordance with the TL pattern of writing and speaking give them thestatus of "Germanized" loan words:

(48) escalation Eskalationdiversification Diversifikationstatus symbo,l Statussymbolperformance' Performanzdomino theory Dominotheoriepilot study Pilptstudieinterdependence Interdependenzattractive attraktivmoderation/ moderator Moderation/Moderator(0 calque, i.e. loan translation (linear substitution) of morphologically

analyzable SL syntagms(primarily noun compounds .and adjective-noun-collocations) which, after a ti me, are often accepted. or at leasttolerated, by the TL community:

(49) growth ratemarket researchdeveloping countrydrug scenefamily planningbirth controlsummit conferenceinterceptor fighteriron curtainnearaccidentpart-time work

WachstumrateMarktforschungEntwicklungslandDrogenszeneFamilienplanungGeburtenkontrolleGipfelkonferenzAbfangjagerEiserner VorhangBeinahe-UnfallTeilzeitarbeit

Page 8: The Translation Process and Translation Procedures

population ll,xplosion ~,·paper tiger :' . ,computer-assisted instruction

, reasonable price -show businessfact-finding';

case studylibrary-systempark-and-ride

" radar"trap

J,.

~":fEr hat das,Geld si~h1en lassen I "Ihm istdas Geld gestohlen worden),

'Wieviele'Fische has~ du gefangeri?F' " ,

How many -fIsh have 'youcaught?, " ,

(52) Four ministers and their." . V\er Minister und ihre Stellvertr~terdeputies v.Jere arrested.' ,w~rden,verhaftet.· ", ,'I

(53) If he had been there: he Wenn er da (anwesend) gewesenware,would have had to witness hatte e'r eine ziemlich haiWche Szenea rather ugly scene" mitansehen mussen. '"o transposition, i.e. the rende.r:ing of an SL elem~nt by,';yn:actico-syn-

tagmatic structures whl,ch have the same' meanmg but do not corre-spond formally (because of 9hanges in the class of wordsusecj):(54) There is absolutelyno SeineBeh'auptting ist at:lsolut I'

truth in his ciai'm. unzutreffend.

Superlative begleitenuns aufSchritt und Tritt.

Paco stand die Scham;ote imGesichto

Superlatives accompany uswherever we go,

Paco's face was red withshame.

It,is accordance with' , I'm SinM der ',Ion der,Regieru(1g;the aim~ pursued by the '" '~erfolgten Ziele hat das:government that Parliam~nt "' "Parl~'ment folgentie~'Progta~m f

has d'ecided on"the follow- ",-', bescl1lossen: " ,,' ,,- •ing program, '

,X·,' ,

Whenever the opportunitY':' Bei jedersich bietend'enhas occurred .'. , ' $',i:;e'legenheit ,:, .'

Aher leaving port the vessel, N~ch Verlassen des Hafensunexpectediy turned, " - ,,' drehte(gingl das Schiff.eastward, ' wider Erwarten au(Ostkurs,

~.(60) ,Everyone was,talking abOut <, Das W~rt vomWettlaufzum

the r\lcll to the moon, , "Mond'war in alle~ Mund.

(61) 'The aircraft made a smooth Das Flugzeug ist trotz Motor-, landing ,despite its schadens glatt ge'landet. ' ", engine trouble. , ,", '

~ modulation, equiv:lence, adaPtat;~n, L~. shifts in semantic perspectiveUwhich, while, prese~ving functional equivalence, entail shifts of meaning

of varying 'size between'SL and TL"textual segmen'ts. Modulation, ispecifically, indicates, ~ change in the 'angle from which somethingis seen (,Changem,ent ,d,e p?"int d,e vu,e). Squivalence 'fs th~ rePlacement, ",of anSL situation by a Communicatively comparable TL situation. 'Adaptation 'amounts to textuaL'compensation for s~ciocultural dif-fEirences betwl'le~the SL and the'r( Gommunities:' ,(62) floating voter

(63) the logic of humanengineering "

(64) He has failed to recognizereality,

(55) , It grows on you,

(66) Take onel

(67) Heis barking up theWrong tree.

,

(68) resounding succ:ess

(69) He has learned it bywatching his neighbourdoing it.

(70) This is a good example ofa self-fulfillingprophecy,

;, ;.

"unentschiedener Wahler•(Wechsel wah ler)

cler sin'nvoll gelenkte Einsatzmenschlictier Leistung

Er hat die Wirklichkeitverkannt,

":'Ian gewohnt sich daran.

Gratisprobe.';

Erist auf d~~ falschen, Spur.

Bomi:lenerfolg

Er hat es seinem Nachbarnabgeguckt.

Page 9: The Translation Process and Translation Procedures

"On the other hand, many questions remain unanswered, as is to be expectedin a firstattempt of this sort. This is true of both literal and non-literal trans-

~ lation. Thei nclusion ofenifmJntiri the classification system ~Q~ld seem out ofplace, since a direct borrowing which has been lexically adapted to a TL, atleast in terms of spelling and pronunciation, cannot be called a translationprocedure in the narrower sense of the word, even if more liberal standardsare, applied; calque, likewise,'belongs in this system only if translation istaken to include substitution processes' below the level of the sentence(Thome 1981).Moreover, literal translation, as already indicated, would have to be seenin a mor~ differentiated way and defined more precisely. One cannot helpbut get the impression that the views of the representatives of stylistiquecomparee have been somewhat distorted by an ill-concealed prejudiceagainst literal translation. This aversion to the' phenomenon rof literal trans-

, lation, which is at cross-purposes with the descriptive aim~ of stylistiquecomparee, is reflected n<;>tonly in its rather ,brief treatment of the topic,,but also ioan empirically unjustifiable slighting 6fwhat literal translationis capable of producing. The negative attitude toward literal translationis demonstrated by Darbelnet:"La traduction Iitterale n'est, en effect, qu'un des sept passages proposes pour passerd'une lilngue a I'autre, ... II faut montrer que .. , fa traduction litterale expose a descontresens, sans compter les deformations qU'elle fait subir a la langue d'arrivee ...la traduction litterale, souvent peu idiomatique, constitue parfois Jri c~tresen~. End'autres termes, il est possible de trahir la pensee en calquant la structure, .. Ainsidonc, la traduction litterale·est trop rigide pour tenir compte de ce :qui se cache der·

. riere les mots. Mais sqn principal defaut'est d'aller souvent a,l'encQntre de ce qu'onappelle parfoisle genie de la langue" (1970,89 ff.).

It is clear from the preceding remarks that Darbelnet does 'not adequatelydistinguish between grammatically and stylistically acceptable and un-acceptable literal translation procedures. In any case, Iit~ral translationseems to be rather powerful, at least in certain sorts of texts, for example

fLSP texts (Wilss 1981d). In addition, it relativizes the c1~im of the irre-versibility of the translation process (Toury 1980a). 1

• The protagonists of stylistique comparee have devoted much: more attentionto non:' iterai translation procedures than they, have to .the three categoriesof literal translation procedures. Its concentration on the ar~a of traductionoblique is an inevitable consequence of the fact that dev\s~ng a system ofclassification for procedures in-the sphere of non-literal tra1slation presentscomplicated problems of definition and subcategorizatioq, owing to thealmost incomprehensible profusion of concrete translJiion procedurephenomena. This explains why Clas, who critically examin,ed the typologyof translation procedures in stylistique comparee, conclucjes that actually

101

only modulation, equivalence, and adaptation belong to the true procedestechniques de traduction (or operations linguistiques, as he also calls them;1971, 608), because .these three relate to semantics., In his opinion, thethree literal translation procedures are to be regarded as special cases oftranslation. His decision to exclude transposition,. the first of the fournon-literal translation procedures, from'consideration as well,' since it dealsonly with formal shifts of expression, can be countered by pointing outthat transposition, like a'nytransfer procedure, requires semantic TE betweenSLT and TLT.

'<: ,"La traduction Iitterale a part, Ie procede.le plus employe est.la transposition quenous dBfinissons ainsi: Ie transfert de la valeur semantique d'une partie du discoursdans la langue de, depart a une autre partiedu discours dans la langue d'arrivee" (Barth1962, 17).

In spite of this clear definition of transpositio';, many inconsistencies con-tinued to exist, until K.-R. Bausch eliminated them in his attempt at a newclassification (1968). The question of whether a clear distinction can bemade between traduction. litterale and transposition has not yet been. answered (see, however, Thome 1981). The problems involved are evidentin, for example1. the rendering of English participial constructions by means of a finite

sentence in German (Wilss 1971, 1978b):(71) Seeing a policeman coming

round the corner, herapidly disappeared fromthe scene .

Er sah einen Polizisten umdie Ecke biegen und verschvvandblitzschnell van der Bildfliiche.

Wir gingen vor Anker.

Weil du uber seinen Hut eine Bemerkunggemacht hast .

Es hat brutto eingebracht.

England hat sich auf das Dezimalsystemumgestellt .

3. the rendering of simple predicate adjectives by adjective-noun-syntagms:(76) Elsie's rejoinder to Elsies Erwiderungen auf die angedeutete

the implied moral moralische RUge war asthetischerrebuke was aesthetic. Natur.

(72) We anchored.

(73) Because you remarkedon his hat ...

(74) It grossed ...

(75) England is decimalized.

It also remains to be seen wheth'er or not the decision to divide shifts insemantic expression into three types of translation pr~cedures (categories5.-7.) can be upheld. The use of the term equivalence to designate thesixth translation procedure in stylistique comparee's taxonomy of trans-lation procedures appears problematic (see chapter VIll. In principle,

Page 10: The Translation Process and Translation Procedures

. "",' "'.';' , ~,:,t,': .' ',:{ f ::every ,translation process is aimed<aC'equivalehce, :meant to"be attained'

, ' :., " ,,' "j". . .1.. "

by whatever translation 'strategyis selected, ," ;: ; h~conceptd:equivalencea unevalidite, extrt3mement 'generale etil tend.~desrgner toute operationde traduction:: (LadmiraI1978, 20f,). Simila~ reservations ~~nbe voicedin regard to the term adaptatfoJ1"1~ls term, too, deriotes,a general state ofaffairs guiding, every, non:'itera'liranslati:on process and estaplishing thecorresporloeQce,betWeenSb·:and Tl: according to ,the'given: conte><tualand situational conditions ofcomniunicationineach case. ,"J , I'adapta-tion n'est deja plus une traduction"'(Ladmiral 1979,20).

Presumably" the best way to do~justice to shifts in~em~ntici:expressjon, isto think ofmodulatibrras a rela.tively·uridifferentiated'generi~term which. ',can'bebroken dOwn into ~ hljmber of subclassesortranslation'procedures .. _"Support for 'th,e hierarchization 'of modulation, equival£jnce,{and adapta;'don cal,be fOlmdin the following r~.marks: '. . .j

: ,,\"',,<, _,i-:J., fl''',,' t_ f:·'" :';'"~appeloils que la modulation est,le .•ter'me que nousproposons. . 'pour designerun 'certain nombre de variat,ions qui devIennent ne~essai"e_s;,!uand Ie pa~sage de LD aLA [Iang'ue de depart, , I<ingue 'd'arriliee l ne "eut, se faire' directement. Nous avonsmontre que ~es variati'ons tienn'ilnta unchl!ngement· de point de vue. Tandis quela transposition opere sur les especes grammaticales, la modulation ,s'¢~erce sur lescategories de la pensee" (Vinay/Darbelriet 11958,66). . .'-.. , ' 'I,. '"'. • _ -tr - ~-, ~,' " ':,:$, ,~.;.,.". '

"La modulation se situe sur Ie plan data pefls~ei' la '~ranspo,siti~n<s~r! Ie: plan de lalangue, Pour,effectuer une modulation 'Iibre',no,n suggeree par Ie (.dictionnaire,il' 'faut .vraiment' se met1fe.a)a plal?~ de I'auteur dl.!,texte(LD at s'interro~~r sur sa visiondumonde"{Vinay1968, 74,5). "". ~ . , ,. '~; >:'1.When shifts of semantic· expression 'are involved: then modul'fItion, wouldbe the equivalent of transposition ,a case of translation proc~dures\"'hichcan· be subdivided into various tnlnsposition'al, subclasses. (Ik,-R. Bausch1968). Jumpelt, for example, uses ;the modulation conc~Pl ;in this sense(1961). Many Qftbe difficulties of, classifying translation, prbced6res can. '. , '" .., .~ t-,probably be avoitled ,by staying:'a'way ,from the linear arrangement usedin'the taxonomy ofstylistiquecomptm}eand choosing inste~a~hiera~chicalprinciple of arrangement which better illustrates the relat!onsh.ips ,betweencategories (see diagram p. 103t ... , ',' ." J

t A rearrangement of transfer categories· would yield little"gajns in'termsof methodology, however,' if. th~, em,pirical science ot" translation failedto ex~mine-'whether and to what:· extlint the' new 'classification scheme,

exhaustiv'ely record, describe,' and assess the degree of TE 'achieved inconcrete transl!ltion procedures. This i~ a subject-matter 'which is plainly'difficult to handle, It is waiting ,to be examined descriptively and 'appli-catively by the language-pair-oriented science of translation. '(

i:non.literal translation

loantrans- .lation(calque) .'

word-for-word

literal..trans-lation

One of the reasonsfonhe intricacies 'in the il;westigationof transfer eventsis ,the limited possibility of textually isolating translation proced~res,. dueto the frequent occurrence, of transfer overlaps',and transfer comb,natl~ns:

Die 'jungen Leute engagieren, sich fUr die Werte, die ihre

Eltern freilich mehr verbal aus-gedrUckt als praktisch verwirklichthaben,

On the other hand, 'it is only by projecting this classification model onto Ia representativ~ text corpus ttiat we will be able to deci~e. to ~hat ext~nt \1

the approach of stylistique,comparee is capable of provldmg mfor~atlOnon the method, technique, and critical review of translation. Its effectivenessin this respect was described as'follows by Malblancin his foreword toVinay /[)arbelnet (11958) andby..the lcittertwo authors :themselves:

, ' 1 ,

" ... la stylistique compareeoffre une technique n~uvelle pour aborder les problemesde la traduotion, quelles 'que soient les lansues cO,nsiderees: il n;e s'agit pas e~ e:fetd'une collection de recettes a appliquer automat,quemer)t, ma's bien de pr!nclpesfondamentaux grace' auxquels peut etredressee la carte des cheminements qUI per-mettent de fake passer tous les elements d'un texte dans une·autre langue" (11958, 1).

(77) The young people are living.out expressed but unim-pie me nteg pare ntal yal ues,

Page 11: The Translation Process and Translation Procedures

y, '. ~, .•

, ",' .~Ie ,passage d'une I~ri~t~~,a unr'~~~U~"Bi'pO~r: el<primer, ~~e me!me r~~lite x,pas,sage que ,,"on de.nomm!l.ha~ltue"er:nent.traductio·n,' rehl~~Jd'une qis,cipline parti.cUI~~re, de nature .c?mpa~atlve, dotiUe.bUt est d'en, ex,pliql!~r Ie, meca~isme etd'en '

: faciliter.la reallsatlon par l!l mise en relief de loi~,valablElspour' les deux! tangues con-siderees.' Nous ram,enons ,ainsi. la,traductidri'l1l un cas particulier,a unk applicationpratique de 'a styii,stique comparef'j1195S,' 20), "', ,; .

<, '''.i'.· o· - ~

An iniportllnt: asp~ct of, stylistlqa(t,c~mparee, i~ the; dist; n'ctio'hbetWeen 'obligat~ry anq opt~6nal translatio;'prote(t~'res,thu~ adding-a~' evaluative

, dimension to their c1assificationmo'del.(Vihay!Darbelnet 11958).' Obliga-tory translation procedures are to be', fouhd 'on the syntactic-s~ntagmatic.orlex,ical-morphematlClevel: They are motivated ';by surface structure'divergences or semantico,sodiocultunil .differences', betvveen"SL and TL, and 'belong ~y their very n~ture to the category of non:llteral'translationprocedures (see, h~wever, Thome ,1981 ).'Obligatory shifts i,!sYr'ltax, w~ich "the representatives of $tylistiquec{Jmparee ~ubsume~:under the he~dingof "servitude (grammaticale)" (Vinay/Darbel\let·11958,'31) " occur when,for example, the translator, in rendering an SL, textual segment" (sentence,subordinate clause, or lexical grouping), is forced to find new ways of, expressinll it in the TL; these entail varying degrees of strJctural' deviationfrom the. SL formulation and 'include deletion operations (~ne-to.zerocorresponden<:es) and/or insertion steps (2ero-to-one c6rrespj:>ndences).Obligatory leXical-morphemic' restructurin,g '(modulations ~igees; Vinay/Darbelnet 11958, 51), which, by analogy with the concept of servitude[grammaticale],coLJld be subsumed under the heading 'of servitudeIex/cafe,is necessary when, for example; a certain 'concept, a' certain lexeme com-bination, ora derivational pattern is absent/from'the TL inventory, andthe only compensatory way out open to the translator is a lexital by-passstrategy such as paraphrasing' or,explanatory' translation. ' , ! '

~ ••.. ;. !

(78) He is an ardent believer 'in Er, ist ein ,begeisterter'fg,'dhender!progress, Anhiinger (not possible in Germ~n:

Glaubigad des Fortschritts/' i;E r ill'aubt begeistert an den "Fortschritt, .

ein Verbrecheil, bei dem au~er d;emTater selbst niemand tu Schadenkommt-, ,. ~(not possible in German:, ''opferloses Verbrechen)

ioss ofqne'; identity :(individuilli~y)in soCiety !

"- j

at~emptHo make something <l new.,~c!entific ~academic) aiscipline 'I, haphazara (uncohtroiledJ bUilqirygactivity in, rural areas'!, ,I

>1housing space which is now (bel'1g)used for other purposes, , i

! ~."

105In 'contrast tQ system,bound' obligatory s~l1tactic '1md'lexical' shifts ofexpression, option,al shifts~. ofexpressi6n ,(transpositions/modulationslibres) (Vihay/Darbelnet 1"958Vare f stylistically motivated. An optionstylistique(Vinay/Darbelnet 11958) occurs when the translator can' choosefrom among several more or less equally acceptableTLversions ..The'trans-lator's singling out of a specif.ic variant may depend on v~rious factors~'_., ·;d:·"

1. the type of text to be translated, ~ ,2., the extent tO,which the SLTbearsstylistic markings,3, the intended TL audience,' '(4. the extent to which the tran'slator can comprehend the SLT and iden-

tify himself with it, ,!

5. the translator's stylisticprefererlces al)d his ability to recognize a!ldhandle stylistic registers.

In contrast to this classification riiodel, 'developed instylistique comparee, on an empirical basis" Kad~ 'pre'sentsa theoretical' model of modes ofbenavior in translation procedures.'He distinguishes 'three basic translatingoperations; failing, however",to include a discussion of examples (1968b):1. substitution (translation on the pasis of a sign-sign-,relationship),2. interpretation (translation on the basis of significative meaning),3. paraphra,se (translation: on the basis of denotative/referential meaning).Substitution is a formal 'translation procedure presupposingsemaritic TEof content ~f SLIT L signs' ahd sequences, of signs: "Where semantico-func-tional equivalence exists, translation can be treated linguistically as inter-lingual substitution" (Kade1968,'100l'. Kilde does not provide any furtherexplanation as' to when semantico-functional equivalence can be said toexist, nor doe~ he 'say whether by this he means relationships ofstringentsyntactic and semantic identity, or whether less' exact relationships of,convergence. can al~o' be accepted as conditions for substitution. In viewof the difficulty' the 'science of 'transi at ion has in objectitying relationshipsof interlingual TE, it ,is understandable, that Kade does not commit himselfin this respect anymore than necessary, Wishing to assure himself of acertain discretional latitude in his methods of classifying transfer proce-dures.

Such caution' is not coincidental, but rather has its roots in the hypotheticalnature of Kade's line of reasoning; proof of this is to befo~nd in his lump-ing together of thesyntactic~spects of speech parts, clause, sentence, andtext in his description of substitution processes:

.' ~ ... _. .. --e

"'n ',Substitu,tion;L2, .u,n"its are 'matchea with ,L,,-units by using P,urely for,ri1al criter, ia.} '"The translation taKes place on the basis of- grammar; i.e. on the,basis of a linguistic Oidescription of equivale.nce relationships' between L 1 ' and L2-formatives" (1968.'15),

Page 12: The Translation Process and Translation Procedures

, 106'T~o points becom~ c;l~a~by,tne above quotation:'~

-f .' < ":,, I,

1. Kade places substitution in"the'realrrl' of liferal translation, his conceptof substitution thus differs clearly from that found in stylistiquecomparee, which' uses' the term substitution ..:todenote certain. phenom-, ena in the category of transposition. .;

2. Substitution is a restrictive transfer procedure which can be used onlyto the extent that clearly discernible relationships. of co- and con-

. ,.textual.is.pmorphY·ex~s~ b~tween.two languages. ,I •

In order.to understand Kade's position, the following iemarks are important:"Substitution ... can be connected with transformations (in the grammatical sphere)and modulations (In the lexical sphere), Transformations and modulations 'are neces,sary. f'irst. in' identifying L1 in order to transfer certain Ll'units into thll realm ofpotential one"to-one correspondences between L1 and LZ, 'this being'a preconditionfor substitution, Transformational operations' also occur., secondly, in the synthesisof LZ, their purpose being to change the LZ' sequence of signs resulting from thesubStitution 'process into an LZ tex:t Waiting the necessary surface ,structure, since thetranSfer proper· takes place partly on' the I~velof" deep' structure" ,(1968b,:13 fJ,

It would appear' th'at Kade is t~ying to ,pla~e substitution in a generative-transformational context. Since the function of transformation in GTGis the transfer from an abstract 'deep structure (the language of ~asic con-cepts and categories) to concrete surface structures (texts), the d,istinctionKade makes between surface structure 'and deep structure translationprocedures strikes one as 'being somewhat unclear. His comment that thetransfer pr,oper is taking place partly on, the level, of deep stru?ture. also, seems to be more confusing than illuminating" since/carried to i~s I!atu~alconclusion, it lends the surface structure ,an air of unimportarce. Kadecould not possibly have meant this, as is shown by his remarkthpt substitu-tion does not as a ru,le appear in, isolation but rather in conn~dtion withsupplementary transformational and modulatory operations. 'W~ican thusconclude that Kade does not use the term transformation rigorously, inthe sense of a coherent system where substitution refers to a' generative-transformational phenomenon of. trans'IatI91i, but rather allows jhirriself tobe guided'by considerations ieading in the directioD of. two co;ncepts de-veloped correlatively by Nida (1964) and Voegelin(1954), thos~ of "back-transformation" (see chapter Ixl and "multiple stage translation" (seechapter X), Back-transformation and multiple stage translation ~re proce-dures in which the text is, transformed, through' paraphrase ~equences.Nida considers back-transformation a means of creating optimal conditions,for .the transfer' b/ r~9ucin!J ,:th'esuXjiace~stru,cture of the SLT \~osimple. structures I~aving the clause/sentence m~al)ing essentially"u~djstorted(1964). Vo~geljn sees m'ultiple stage translation as a procedure for arrivingat an optimally completed translation by means of a' series of ' method-

, ologically objectifiable steps of translation via translational approximation ortranslational interim stages (1954). '

107Whether- or not ..the attempt at. a clarification of Kad~'s 'concept of. trans-formation vis-a-vis the two concepts back-transformation and multiplestage translation does justice to his line of reasoning is another matter.In any case, it can be said that this approach is one way of ilJustratin~ th?sefunctional cross-connections between transformation, and substitutIOnwhich show the translation process to be a procedure for producing a TLTon the basis of many interwoven layers of transfer.VoegelinlVoegelin state the' fact that substitution has a legitimate. function,if it has one at all, only in assodation with:intralingU{lI and interlingualliteral and non-literal transfer operations:, "Translation which goes directlyfrom a natural SL text to a natural, TL text is so loaded with non-equiva,lences as to make it inadvisable to generalize for any pair of SL and TL ... "(1967,2187).Interpretation and paraphrase presen.t just as many problems of classifi-cation and definition' as, substitution. Kade defines these two transferprocedures as follows:. . ,.

"In interpretation, L2-,units are 'matched with L1-units on the, basis of,semanti,co-functional features determined by referring to the image association. The translatIontakes place on the level of content, i.e. on the basis of empirically e,stablished.semant~co.functional equivalence between L1- and L2-units', Since man. IS a conscIous be,lOg,linguistic signs trigger associations with images, and translation by human be~ngstakes place primarily, in the form of interpretation .(1968b, 15), In pa~aphrasrng.L2-units are not matched with Ll-units. The. proc~ss, is no: one of rec?dlOg, butrather one. o~ developing a new .code in L2; startlOg WIth the object. In certain types oftexts (for: example, in Iyri~al,texts. plays on words. and texts' which cannot be re-duced t~ a single. meaning), ,paraphrase is the only possible form in which the trans,lation can be effected" (1968b, 16),

Both translation procedures, Kade says, are determined by content; theyexhibit differing semantic fe,!tures, however, in that interpretation is signi-. fication,oriented, whereas paraphrase is designation:oriented. The focusof interpretation is therefore of a semasiological nature (what does ~ wordmean? L as opposed to the focus of paraphrase. which is of ,~n onomaslolog1c,al nature (what shall I call an object?), In the first instance, the basis ofthe transfer effected during translation is the interpretation of a linguisticsign as a bilateral ~ntity having a form-content relationshi~ dependenton the context;, this is to all appearances determined more or less sub,jectively, In the second instance, the first act of thetran~lator iso?~iou,slya refEnential-semantic (denotative) decoding of the SLT, I.e. a clarificationof the relationship betwee'n the text and extralinguistic, reality; this com-pleted, he proceeds to make use of the TL potential for expression toproduce the TLT. ,In paraphrasing. the translator clearly h.as muchgrea,terlatitude in his selection of TL expressions than he would If he Were uSingsubstitution or interpretation.

Page 13: The Translation Process and Translation Procedures

108

Aside, from the fact that Kade, by faiJing, as indicated, to incl~de a dis-cussion of examples, makes it unnecessarily difficult,to follow his argumentsand leaves himself open to the danger of misinterpretation, the ,followingcan be said about his line of reasoning: . r'

As regards the efficiency of linguistic communication, be it m~nolingualor bilingual in nature, we would be in .a sorry state if signification anddenotation -: or, if you will, 'lexical meaning and referential-semantic(object-oriented) meaning - were '1ot largely interdependent dr at leastconvergent and if each person using a language were allowed to freely con-nect linguistic signs with ad libitum associations. It is precise)ly in the'area of LSP texts, which holds special interest for Kade, in 'vi~w of hispreoccupation with the preservation -of SL!TL semantic inva~iance, thatthe convergences be~ween Iin~uistic, arid textual meaning (Qrl betweenBedeutung and Bezelchnung, In Coseriu's terminology; 1981)[ becomes

lspeciallYrelevant. For that reason, LSP texts are not int!lrpreted h~rmenellti- '~ally,in ind.ividual .me~talsteps cautiou~ly'-approaching the resP~.btivetext;Instead, their meaning IS comprehended and 'retraced cognitivelyl THe text-,pragmatic sign dimension is thus least marked in C:SPtexts, whe~e it tendstoward zero value (Wilss 1979a),. ;" 'I '

, ,." I

The use of interpretation to denote a certain class oftrarislatiorl processesis a rather infelicitous choice of words; the terminological r~servationsvpiced in regard to equivalence and, adaptation (see abdve) apply! here, too.Aside from interlingual set phrases and semantically unambiguous, denotati-vely marked texts, every translation procedure exhibits an interpretativecomponent. Jakobson is thus correct in speaking of translation (regardlessof whether it· is intralingual, interlingual, or intersemiotic in naturel as a"recoding interpretation" (21966,236), :

Interpretation is especially importa~t as an integral component of thetranslation process in the field of artistic translation. Hollander, for ex-ample, considers "interpretative translation" a problem of "literary theory"(1966, 215); Levy notes that

"because of the incongruence in lingutstic material, perfect agreement of mean'ingbetween expressions in the translation and those in the original is not possible .. 'and ... therefore interpretation [is] necessary" (1969, 47l. '

Koller, similarly, calls translation, especi<llly tne translation of belles lettresan "interpretation of the original text" (1972, 66), and Popovic sees th~essence of literary translatioll as consi,sting of a stylistic interpretationwhich he sees as an "analysis of the stylistic m~ans used in the text of atranslation in the light of the stylistic pro~erties of the original" (1971,153).

Kade does not deal with the questions of where the transition betweeninterpretation ,and paraphrase takes place or what criteria of classificationare avaible to research on transfer procedures. He merely notes ratherlaconically "that in the practice of human translation, 'i nterpretation'and "paraphrase' rarely appear in their pure form but rather, as a rule,overlap" (1968b, 17). This remark can be taken only one way: as saying,namely, that at the present time the science of translation has very littledefinitive knowledge about the possibility, of a precise differentiationbetween "r~coding" and "new coding" (Umkodierung/Neukodierung inKade's terms); this is because the science has been unable thus far to defineexhaustively those conditions under which translation via paraphrase is, a legitimate, necessary translation procedure, covering special needs of theSL author, the translator or the TL 'recipient. The difficulties involvedin resolving this problem are compo"lirided by the fact that, in spite 9f allthe differences in translation', pro,cedures, the two forms of translation

, are similar, since re-coding an'd new 'coding,or translation in the narrower,and translatio~ in the broader sense of the word, call for the ability. tocorrectly inte'rpret the meaning oLi sentence or a senten~e sequence. '

"In view of what, has been said, [textual) intuitions of a bilingual informant inVOlve)the same fundamental operations as dqes paraphrase, since both involve the, abilityto asso'ciate particular sentences with the underlying deep structure or, in othe'r words,ability to interpret the meani~g of the sentenc,es" (Krzeszowski 1971, 47J.' ,

In addition, it should- be noted that paraphrasing is actually ~'monolingualoperation which is based on textual synonymity relations 'and thus, strictlyspeaking, falls into the category of int'ralingual translation. Finally, deter-mining what position paraphrase occupies in the arsenal of possible trans-lation procedures is further complicated by the lack of agreement on thescope of meani':lg of paraphrase translation; some of the definitions foundin the science of translation conflict with one another, Rabin, for example,thinks that the function of paraphrase - as a counterpart to literal trans-lation - is to serve as a collective term for all TL-directed (non-literal)forms of translation."There are, of course, two distinct methods of translation: paraphrase and literaltranslation .. , Naturally, the paraphrase does not work independently of all word-references in statement A. It picks out leading referents, such as nouns and verbs andbuilds up a new context appropriate to the way in which the situation would benormally analysed in language S" (1958, 131 f.J.

Nida and Taber, on ,the other hand, regard semantic equivalence betweenSLT and TLT without lexical correspondences as an essential componentof a paraphrase translation (1969). Tancock, in his turn, sees a paraphrasetranslation as one in which the TLT deviates in meaning from the SLT(1958), Catford hands down a Salomon-like verdict:

Page 14: The Translation Process and Translation Procedures

iI

110, ". \1."It is clear ~'hat'these commonly used terms '~td-for.word', 'literal', anb 'free', infer~ry largely t~ types of translation in Which .equivalences are set Up at d.ifferent ranksIn a grammatical hierarchy. Equivalences l"l1ay be setUp at ranks still higher than thesentence - say the paragraph. When this occurs there.is some change of the 'free trans.lation' becoming more losely related to the source text, so that the result might moreproperly be called a 'paraphrase' .• (1967, 1351.- ' .

The normative implications of translation via paraphrase hinted at here byCatford take on their full importa.nce, from thE{socroc;:ulturalpciint of view:, ." ,', . it is possible, to make a bad translation .•. by 'paraphrasing loosely Jnd distorting~he messag~ 'to conform to alien cultural patterns ... \his is the bad sense of the wordparaphrase. But ... a good translatlot:l focuses on the meaning or content as suchand aims to preserve that intact; and in the process it may quite radically restructurethe form: this is.paraphrase in the proper sense" (Nida/Taber 1969, '173). \ .

S611 demontrates yet another perspective ofparaphr~se translation with hisreference t.oJhe function of..paraphrase'as related to a specific tYpe of text:"Une theor~e stric~e",ent ,linguistique de la' trad~~iion d~it~oflsid~r~rl~ paraphrasecomme ~qUlva.lente (Ia. paraph;ase e.st con~titutive de toutce qui concerm! ,Ie langage),~a the~>rte de .'a traduction IItteral;e par ,contre,.ne ,peut la considerer';que comme. ersa~z et de ce fait comme non equlvalente" (1971,29 f.). ." , :

In view of the uncertainty surrounding the definition .and cl~ssificationof the par~phrase concept, a strong case ·can be made for urging the. scienceof translation to te~porarily put aside the aspect of paraphrase and to tryother means of solving the problems of classifying translation procedures.One conceivable approach is a psycnolinguistically oriented classificationsystem which divides translation proce.dures into three 9roups:

1. habitualized translation procedures (translation mechanisms basedon the principle of association):(84) How could you take that

into your head?Was haben Sie sich (was hastdu dirt dabei eigentlich ge·dacht7/Wie sind Sie (wie bist du)nur auf diese (verruckte)Idee gekommen? .

Er' hat' zwei Fliegen mit einer. Klappe geschlagen.

Er kommt.wahrScheinlich nicht;

2. partlvhabitualized translati6~ procedures which can be re~alled semi·internalized: .

(85) He has killed two birdswith one stone.

(86) He is.unlikely to came.

(87) Having finished his job hewent home,

. ,

E~ b8!l~dete seine Arbeit 'und .:'ging dann nach H'llUse/ . ; .,Nachdem er seine Arbeit been'dethatte, ging er nach Hause/ . iNach Beendigung seiner Arbeit 'ging er nach Hause. . ,

Die'Methode, die Steuern zu be·rechnen ...Die Methode der Berechnung van

. Steuern .. ,Die Methode der Steuerberechnung '.' .

3. non-habitualized translation procedures which force the translatorto muster up all his creative tesources and come up with his own formu·lations on the basis of cognitive psychology" (not associative psycho-logy):

(88) The method of ca.lculatingthe levy, ..

(89) In these private theatricals Bei diesen LiebhabeNluffuhrungenIt was.her own ·many-facetted .... spielte sie ihre eigene, schillerndenature that she put on exhibit, Personlichkeit aus, und die Zuschauer-and the audience .. , could· ... zollten sowohl ihrer Rol/enbe·applaud both her skil/ herrschung wie ihrer Wandlungsfahig·of projection and her 'keit Beifall.i,ntri nsic vari ety. .

Of course; this attempt; too, at classifying translation procedures .will veryquickly run into limits of pbjectification, a problem it will 'not be ableto overcome, since a classification 'of'translation procedures accordingto degrees of habitualization might: end up varying from translator totranslator. This is partly because' translation, like any" act of linguisticcommunication, is a mental activity in which internalized transfer strategiesand non·internalized transfer strategies work together ina way controlledlargely by the 'translator himself (Wilss 1981 d). Richards, although some·what overstating the case, is tendentially correct in calling translation "veryprobably ... the most complex type of event yet produced in the evolutionof the cosmos" (1953, 250), F6r'that reason, it is questionable whetherresearch in the field of translation procedures will, in the foreseeable future,be able to come up with any exhaustive classification system giving sufficientconsideration to all types of texts relevant to translation. .