The translation shift approach

Key concepts

Translation shifts =small linguistic changes occurring in translation of ST t o TT.
Vinay and Darbelnet (1958): classical taxonomy of linguistic changes in translation.
Catford's (1965) term translation 'shift' in his linguistic approach t o translation.
Theoretical work by Czech scholars Levy,Popovir and Miko(1960s-1970s) who adopt stylistic and
aesthetic parameters of language.
Most detailed model of translation shifts:van Leuven-Zwart's,an attempt to match shifts t o discourse and
narratological function.
The problem of the subjectivity of the invariant that is used t o compare ST and TT.

Taxonomy- the practice and science (study) of classification of things or concepts, including the
principles that underlie such classification

Introduction
Since the 1950s, there has been a variety of linguistic approaches to the analysis of translation that have
proposed detailed lists or taxonomies in an effort to categorize the translation process. Thus, the focus of
this lesson is on the following three models:

1.

Vinay and Darbelnet's taxonomy in Styiistiqwcornpuree d i ~fran~aiset de l'anglais (1958195)' which is
the classical model and one which has had a very wide impact;

2.

Catford's (1965) linguistic approach, which included the introduction of the term 'shift' of translation;

3.

van Leuven-Zwart's (1989, 1990) very detailed model, designed for the analysis of the key concept of
small 'microlevel' translation shifts and the gauging of their effect on the more general 'macrolevel'.

I. VinayandDarbelnet'smodel
Vinay and Darbelnet carried out a comparative stylistic analysis of French and English. They looked at
texts in both languages, noting differences between the languages and identifying different translation
strategies and 'procedures'. Although their Stylistique compare dii francair et de I'anglais (1958) is based
solely on French and English, its influence has been much wider. Amongst others it has formed the basis
for a work in the same series on French-German translation and two similar books on English-Spanish
translation.
The two general translation strategies identified by Vinay and Darbelnet (2000: 84-93) are direct
translation and oblique translation, which go back to the 'literal vs. free' division. Indeed, 'literal' is
given by the authors as a synonym for direct translation (p. 84). The two strategies comprise seven
procedures, of which direct translation covers three:

a). direct translation
1. Borrowing: The SL word is transferred directly to the TL. This grouping covers words such as the
Russian rouble, datcha or, more recently, glasnost and perestroika, that are used in English and other
languages to fill a semantic gap in the TL. Sometimes borrowings are employed to add local colour
(pitancpe, armagrlac and bastide in a tourist brochure about south west France, for instance).

2. Calque: This is 'a special kind of borrowing' where the SL expression or structure is transferred in a
literal translation. For example, the French calque 'Compliments de la Saison' for the English
'Compliments of the Season'. Vinay and Darbelnet note that both borrowings and calques often
become fully integrated into the TL, although sometimes with some semantic change, which can turn
them into false friends.

3. Literal translation: This is 'word-for-word' translation, which Vinay and Darbelnet describe as being
most common between languages of the same family and culture. Their example is 'I left my spectacles
on the table downstairs' which becomes 'J'ai laissC mes lunettes sur la table en bas.'
Literal translation is the authors' prescription for good translation: 'literalness should only be sacrificed
because of structural and meta- linguistic requirements and only after!

checking!

that

!

the! meaning is fully preserved.' But, say Vinay and Darbelnet, the translator may judge literal

translation to be 'unacceptable' because it:

(a)

gives a different meaning;

(b) has no meaning;

(i)

i s impossible for structural reasons;

(d) 'does not have a corresponding expression within the metalinguistic

experience of the TL';

(a)

corresponds to something at a different level of language.

In those cases where literal translation is not possible, Vinay and Darbelnet say that the strategy of
oblique translation must be used. This covers a further four procedures:

4. Transposition: This is a change of one part of speech for another without changing the sense.
Transposition can be:

obligatory: 'des son lever' in a particular past context would be translated as 'as soon as she got up';
optional: in the reverse direction 'as soon as she got up' could be translated literally as 'des qu'elle s'est
levee' or as a transposition in 'des son lever'.

Vinay and Darbelnet see transposition as 'probably the most common structural change undertaken by
translators'. They list at least ten different categories, such as:

verb -> noun: 'as soon as she got up' -> 'des son lever';
adverb -> verb: 'He will soon be back' -> 'I1 ne tardera pas A rentrer' [lit. 'He will not tarry in returning'].

5. Modulation: This changes the semantics and point of view of the SL. It can be obligatory or optional.
Modulation is a procedure that is justified, 'when, although a literal, or even transposed, translation
results in a grammatically correct utterance, it is considered unsuitable, unidiomatic or awkward in the
TL'.
Vinay and Darbelnet call modulation 'the touch- stone of a good translator', whereas transposition 'simply
shows a very good command of the target language'. Modulation at the level of message is subdivided
along the following lines:

abstract for concrete !

!

!

!

!

change of symbol (including fixed and new metaphors).

cause-effect !

!

!

!

!

!

rethinking of intervals and limits (in space and time)

part-whole ! !

!

!

!

!

!

space for time

part-another part ! !

!

!

!

!

active to passive (and vice versa)

reversal of terms ! !

!

!

!

!

negation of opposite

This category therefore covers a wide range of phenomena. There is also often a process of originally
free modulations becoming fixed expressions. One example given by the authors is 'Vous l'avez
kchappi. belle' [lit. 'You have!

escaped! beautifully']-> 'You've had a narrow escape'.

6. Equivalence: Vinay and Darbelnet use this term to refer to cases where languages describe the same
situation by different stylistic o r structural means. Equivalence is particularly useful in translating idioms
and proverbs (the sense, though not the image, of 'comme u n chien dans un jeu de quilles' [lit. 'like a
dog in a set of skittles'] can b e rendered as 'like a bull in a china shop').

7. Adaptation: This involves changing the cultural reference when a situation in the source culture does
not exist in the target culture. For example, Vinay and Darbelnet suggest that the cultural connotation of
a reference in an English text to the game of cricket might be best translated into French by a reference
to the Tour de France. The authors claim that a refusal to use such adaptation in an otherwise 'perfectly
correct' TT 'may still be noticeable by an undefinable tone, something that does not sound quite right'.
However, whereas their solution may work for some restricted metaphorical uses, it would make little
sense to change cricket to cycling in phrases such as 'that isn't cricket' o r
'a sleepy Wednesday morning county match at Lords'.

The seven main translation categories are described as operating on three levels; these three levels
reflect the main structural elements of the book. They are:
1 the lexicon;
2 syntactic structures;
3 the message.
The key question the 'unit of translation' is concerned, the authors reject the individual word. They
consider the unit of translation to be a combination of a 'lexicological unit' and a 'unit of thought' and
define it as: 'the smallest segment of the utterance whose signs are linked in such a way that they
should not be translated individually'.
Two further terms are introduced which look above word level; these are:
1 word order and thematic structure

2 connectors which are cohesive links, discourse markers (however,first, etc.), deixis (pronouns and
demonstrative pronouns such as this, that) and punctuation.

A further important parameter taken into account by Vinay and Darbelnet is that of servitude and option:

- servitude refers to obligatory transpositions and modulations due to a difference between the two

language systems;

- option refers to non-obligatory changes that are due to the translator's own style and preferences.

This is a crucial difference. Authors stress that it is option, the realm of stylistics, that should be the
translator's main concern. The role of the translator is then 'to choose from among the available options
to express the nuances of the message'. The authors give a list of five steps for the translator to follow
in moving from ST to TT; these are:
1.Identify the units of translation.

1.

Examine the SL text, evaluating the descriptive, affective and intellectual content of the units.

2.

Reconstruct the metalinguistic context of the message.

3.

Evaluate the stylistic effects.

4.

Produce and revise the TT.

These procedures are applied on three levels of language:

i. the lexicon
ii. the grammatical structures and
iii. the ‘message’, which is used to refer to the situational utterance and some of the higher text

elements such as sentence and paragraphs.
At the level of message, Vinay and Darbelnet discuss such strategies as compensation,
an important term in translation which is linked to the notion of loss and gain.

Compensation, loss and gain

A translation technique used to compensate for translation loss. The translator offsets an inevitable

loss at one point in the text by adding a suitable element at another point, achieving a

compensatory translation gain.
Compensation in an interpretive sense, restoring life to the TT, is the fourth ‘movement’ of

Steiner’s hermeneutic process.
II. Catford and translation 'shifts'

The term “translation shift” was first introduced by Catford. His definition of this concept relies on his
distinction between formal correspondence and textual equivalence:

formal correspondence is a relationship that holds between two linguistic categories that occupy
approximately the same place in the organisation of their respective languages,
textual equivalence holds between two portions of text that are actual translations of each other. When
a textual equivalent is not formally correspondent with its source, this is called a translation shift, of
which there are two major types: level shifts and category shifts.

A formal correspondent is 'any TL category (unit, class, element of structure, etc.) which can be said to
occupy, as nearly as ~ossible,the "same" place in the "economy" of the TL as the given SL category
occupies in the SL' (Catford 1965: 27).
A textual equivalent is 'any TL text or portion of text which is observed on a particular occasion ...to be
the equivalent of a given SL text or portion of text'.

Textual equivalence is thus tied to a particular ST-TT pair, while formal equivalence is a more general
system-based concept between a pair of languages. When the two concepts diverge, a translation shift
is deemed to have occurred.
Catford considers two kinds of shift: (1) shift of level and (2) shift of category:

1 A level shift would be something which is expressed by grammar in one language and lexis in another;
this could, for example, be:
- by aspect in Russian being translated by a lexical verb in English: e.g. igrat' (to play) and sigrat' (to
finish playing);

- or cases where the French conditional corresponds to a lexical item in English: e.g. 'trois touristes

auraient ete tues' [lit. 'three tourists would have been killed'] = 'three tourists have been reported killed'.

He speaks of a level shift when a source text item has a textual equivalent on a different linguistic level.
He distinguishes the four linguistic levels – phonology (the medium-form of spoken language),
graphology (the medium-form of written language), grammar (closed systems), and lexis (open sets) –,
which are related in language-specific ways to extra-linguistic levels of substance: phonology to phonic
substance, graphology to graphic substance, and both grammar and lexis to situation substance.
Level shifts, however, can only occur between the levels of grammar and lexis. This restriction is due to
Catford’s understanding of translation equivalence, which, from his structuralist point of view, is not
based on a sameness of meaning, for meaning is defined as “the total network of relations entered into
by any linguistic form” and consequently cannot be the same across languages.

2. Most of Catford's analysis is given over to category shifts. These are subdivided into four kinds:
(a)! Structural shifts: These are said to be the most common form of shift and to involve mostly a shift
in grammatical structure. For example, the subject pronoun + verb + direct object structures of I like jazz
and j'aime le jazz in English and French are translated by an indirect object pronoun + verb + subject
noun structure in Spanish (me gusts el jazz) and in Italian (mi piace il jazz).
(b) Class shifts: These comprise shifts from one part of speech to another. An example is the English a
medical student and the French un itudiant en midecine, where the English pre- modifying adjective
medical is translated by the adverbial qualifying phrase en me'decine.
(c) Unit shifts or rank shifts: These are shifts where the translation equivalent in the TL is at a different
rank to the SL. 'Rank' here refers to the hierarchical linguistic units of sentence, clause, group, word and
morpheme.

(i)

Intra-system shifts:These are shifts that take place when the SL and TL possess approximately
corresponding systems but where 'the translation involves selection of a non-corresponding term in
the TL system'.Examples given between French and English are number and article systems, where,
although similar systems operate in the two languages, they do not always correspond. Thus, advice
(singular) in English becomes des conseils (plural) in French.

In Catford’s theory, the concept of system has a more restricted meaning that in other structuralist
approaches, where it usually refers to the relations that hold within a language in its entirety. Here, the
term is used for “a finite set of alternants, among which a choice must be made” for example the
system of pronouns or of number. An intra-system shift occurs when the two languages have a formally
correspondent system, but choose a non-corresponding item as translation equivalents. For instance,
English and French both have a two-place number system, but the English plural “trousers” is
translated as the French singular “le pantalon”.

Despite the steps taken by Catford to consider the communicative function of the SL item and despite
the factt that his terminology was based on a functional approach to language, the main criticism of
Catford's theory is that his examples are almost all idealized (i.e. invented and not taken from actual
translations) and decontextualized. He never looks at whole texts, nor even above the level of the
sentence.

III. Van Leuven-Zwart's comparative-descriptive model of translation shifts

The approach by van Leuven-Zwart (1989, 1990a) differs from the previous three in that it is designed
and used for the description of actual translations rather than of the relationship between two linguistic
systems. A further difference lies in van Leuven-Zwart’s general attitude towards shifts. In the
approaches described so far, the overall tendency was to see shifts as inevitable but somewhat
undesirable. They were accepted as ways of coping with the systemic differences that exist between
any two languages, but at the same time, the need to overcome these systemic differences was
generally seen as the only justifiable cause for the occurrence of shifts. Van Leuven-Zwart’s view of
shifts, however, is more neutral. She does not describe what translators could and should do or should
not do, but simply observes and describes what they actually have done, she analyses “the translator’s
interpretation of the original text and the strategy adopted during the process of translation”.

The most detailed attempt to produce and apply a model of shift analysis has been carried out by Kitty
van Leuven-Zwart of Amsterdam. Her model takes as its point of departure some of the categories
proposed by Vinay and Darbelnet and applies them to the descriptive analysis of a translation,
attempting both to systematize comparison and to build in a discourse framework above the sentence
level. The model is 'intended for the description of integral translations of fictional texts' and comprises
(1) a comparative model and (2) a descriptive model.

1 The comparative model involves a detailed comparison of ST and TT and a classification of all the
microstructural shifts (within sentences, clauses and phrases).

There are three main categories: modulation, modification and mutation.

- syntactic-semantic modification: the noun + Saxon genitive mother's becomes the possessive

pronoun suyo;

- syntactic-pragmatic modification: the selection of mother's rather than hers meaning that more

pragmatic information is supplied to the reader in the English ST than in the Spanish TT, where the
reader has to understand the link to mudre.

Once all the shifts are identified and categorized on this low 'micro- structural' level, the number of
occurrences in each category is totalled and their cumulative effect is then calculated by using a
descriptive model, as follows:

2 The descriptive model is a macrostructural model, designed for the analysis of translated literature. It
is based on concepts borrowed from narratology and stylistics. It attempts to interweave the concepts of
'discourse level' (the linguistic expression of the fictional world) and 'story level' (the narration of the text,
including narratorial point of view) with three linguistic 'metafunctions' (interpersonal, ideational and
textual).

There are, however, drawbacks to this model, and these drawbacks relate to taxonomies in general.
First, as van Leuven-Zwart herself partly recognizes, the comparative model is extremely complex.
There are practical implications in allocating the different kinds of shift since there are eight different
categories and thirty-seven subcategories, not all clearly differentiated. Second, keeping track of all the
shifts throughout a long text is also difficult.

The preponderantly linguistic approaches to translation were effectively supplanted by the culturally
oriented branch of descriptive translation studies, which still continues to be one of the most prominent
areas of research in translation studies today. However, in recent years, one can observe a renaissance
of linguistic approaches, certainly encouraged by Baker’s influential suggestions as to how linguistic
corpora could be applied to the study of translation.