Vous êtes sur la page 1sur 16

ERRORS IN TRANSLATION AND HOW TO AVOID THEM

JASMINA DJORDJEVIC
Faculty of Legal and Business Studies Novi Sad, Serbia
Abstract
Translation can be quite an ordeal especially if the translator does not
know what kinds of traps they might encounter. However, when aware of the
possible dangers and mistakes, translation may be improved considerably. The
first step towards better translation is the identification of possible errors in
written and oral translation. The second and more demanding step is the
application of appropriate techniques so as to avoid inaccurate renderings in
any target language two of them being of substantial relevance - the contrastive
and the component analysis.
Key words: translation, errors, analysis, contrastive, component
Introduction
On translation
A somewhat paradoxical standpoint among scholars and practitioners
engaged in translation is that there is no such thing as translation. This almost
absurd point of view is based on the fact that the real world, that is the
extralinguistic reality1, that is everything outside language, can be categorised 2
in many different ways. In so far as this categorisation is concerned, the
problem is twofold. First, it may be difficult to determine the relationship
between language, the tool for the mentioned categorisation, and the results of

the categorisation, that is the different categories. Second, this relationship has
to be translated.
A possible illustration of categorisation in two languages:
Category
I
II
III
IV

English
German
animal
Tier
dog (N)
Hund (N)
canine (Adj)
hunde- (Adj)
dog (for compound N)
Hunde- (for compound N)
Table 1: Categorisation in translation

When regarding this very simple example, we can see that I and II
category have their direct equivalents, that is translations, whereas the third
category does not. The term canine has to be translated by the adjective form
hunde- derived from the noun Hund in German because there is no direct
equivalent for the term canine, which by the way is of Latin origin. As far as
the meaning of the term itself is concerned, there might not seem to be a
problem, because the term hunde- covers the meaning of the term canine since
it refers to concepts and items connected with dogs. However, on a deeper level
there might be a problem as some phrases might not be easy to translate. For
instance, canine trainer can be translated Hundetrainer causing uncertainty as
this is not a Noun Phrase as it is in English, consisting of a pre-modifier,
canine, and the Head, trainer. The German equivalent is a compund Noun!
(Table 2)
In the case of the phrase dog-tired, when translated, it undoubtedly
refers to hundemde (Table 2), which might cause confusion since Hunde, the
plural form of the noun and hunde-, the derived form used for adjective
compounds, look just the same and are yet different the capital letter in the
Noun is dropped in the Adjective compound. Obviously the spelling rule asking

for all Nouns in German be written with a capital letter is of crucial importance.
On the other hand, the specific term genus canis is translated as Gattung canis,
obviously a combination of a German and a Latin term.
English
German
dog breed
Hunderasse
canine trainer (NP) Hundetrainer (comp. N)
dog-tired (Adj)
hundemde (Adj)
genus canis
Gattung canis
Table 2: Illustration of equivalence in English and German based on categories

Furthermore, there are two main, but mutually exclusive points of view
regarding the above problem of there being no true translation. The first is
based on the idea that the structure of language is universal and that because of
that human beings are able to communicate among themselves regardless of the
language they are speaking exactly because translations make it possible to
transfer concepts from one language system to another. The second point of
view, by contrast, is based on the idea that every single language has its own
system of categorising reality including thus an individual system of seeing the
reality being categorised as based on the particular language. It is this second
point that leads us to the absurd statement that there are no true translations.
Among other things we are now facing a quite complicated riddle: What then
are we doing when translating if not translating?

Translation being decoding and encoding


A very convenient point of view that may be assumed regarding
translation is that it is a constant process of coding, on one hand, and
modification on the other.
In the seventh edition of the Oxford Advanced Learners Dictionary, the
term coding is explained in reference to the language of computer technology where
the process of coding refers to the process of writing a computer program by putting
one system of numbers, words and symbols into another system. The second entry
refers to translation of languages; however, it has to be taken into account that this
process, both in computer technology and in translation as well, consists of two steps.
The first is decode, that is analyse and understand the meaning of something in the
source language and then, the second step, rewrite or express that meaning in the target
language. Thus the coding process enables us, the translators, to realise

intentions uttered in one code (the source language - SL) in another code (the
target language - TL). A further step is check the meaning of the term encoded
in the SL by decoding it once again. If the decoded meaning in the SL is exactly
the same as the meaning in the TL, the process of coding has been realised
successfully.
The process of modification is applied if the process of coding is
insufficient or otherwise impossible to realise which enables the translator to
alter intentions expressed in the source language so as to make them fit the
target language. In both cases, translators have to face the fact that they may
either undertranslate or overtranslate concepts!

1. Term in the
English code:
cat

4. Decoded again:
kleines Haustier
mit weichem Fell
und Schnurrhaaren

2. Decoded:
small animal with
soft fur and
whisker, kept as a
pet

3. Encoded in
German:
Katze

Figure 1: Illustration of coding in translation

Errors in translation
Accurate translation is hard to achieve but it is not impossible. What a
good translator must bear in mind at all times is that translation is a synthesis of
different types of knowledge the two basic ones being linguistic and extralinguistic knowledge. Apart from that, the translator has to develop linguistic
and the extra-linguistic skills so as to make the best of their formal knowledge
acquired in long and exhausting grammar and vocabulary classes while
studying language including:
1. Competence on every linguistic level (phonology, morphology, syntax,
semantics and pragmatics),
2. Ability to identify linguistic contrasts between the SL and the TL and

3. Skill to transfer meaning from the SL into the TL.


In comparison to that, the extra-linguistic skills should be the appropriate
context for the linguistic skills including the ability to:
1. Understand the sociolinguistic and socio-cultural context of both the SL
and the TL (the artefacts of a particular culture),
2. Identify the extra-linguistic contrasts between the SL and the TL and
3. Transfer the meaning of the context from the SL into the TL
In reference to that, errors are easy to define both in written and oral
translations including:

Morphological (tenses, articles, modality, gender, number, inflections,


etc.)
e.g. there is no present perfect tense in the TL

Syntactical (word order, phrases, clauses, etc.)


e.g. the problem of finite and non-finite phrases

Semantic (meaning of course!)


e.g. idioms, collocations, fixed expressions, etc.

When referring to oral translations we must add phonological mistakes


which are a result of bad pronunciation and articulation and can lead to
confusing situations.
e.g.

ship / p/ sheep / p/
chip /

p/ cheap /

bad /b d/ bed /bed/

p/

Analyses in translation
The following pages shall be an attempt to illustrate two possible
techniques, the contrastive and the component analysis, to be applied in
translation so as to achieve the most accurate renderings in a target language.
The problems and the solutions covered here are supposed to direct the
translator into the world of translation with proper ammunition.
Contrastive analysis
One possible way of translating is to find the closest equivalent in the
(target language) TL for a certain concept in the (source language) SL. Some
authors might not be satisfied with the decoding and encoding procedure as a possible
explanation of translation which is why they may refer to transcoding when discussing
translation of some accepted terms. Modern science refers to transcoding as being a
character encoding converter and it is also used in reference to computer technology. In
reference to translation, Danica Seleskovic (2007: 141) says that in transcoding we
take only that meaning that a particular word has in both the target and the source
language. But she also claims that transcoding has its special importance in
several cases, especially in the case of numbers. Based on both meanings of
transcoding, it may easily be concluded that it is a process of transcoding of one lexical
unit uttered in the source language into an equivalent lexical unit in the target
language. Bottom line is that the translator has to find an appropriate unit in the target
language to express exactly the same meaning expressed in the source language.

However, a new problem will arise and that is that the degree of
correspondence will not always be the same. What is more, it may range from
absolute correspondence to complete lack of correspondence. We may
distinguish between three levels of correspondence:

1. Absolute correspondence where there is a perfect equivalent for a


term in the SL. The area of absolute correspondence includes
internationalisms and terms that are of one and the same origin and
are easily recognised in several different languages.
2. Partial correspondence where the meaning of a particular lexeme in
the SL has to be adjusted to the context of the TL.
3. Absolute lack of correspondence where there is no equivalent term
so that the translation has to be paraphrased or a note explaining the
concept has to be introduced.
Absolute correspondence
The area where there is an absolute correspondence is the one where
translation is realised completely. It is generally accepted that the units that may
be regarded as being in a relationship of absolute correspondence are
internationalisms, such as names for chemical elements, trade marks, technical
terms, names of international organisations, units of measures. It goes without
saying that these terms still suffer some alteration as they have to be adjusted to
the morphological and phonological system of the TL.
Yet, there might be problems arising from the fact that even in
the case of absolute correspondence some lexical items might cause confusion:
1. Chemical elements: there might be names in the TL that have been
derived from popular names, such as olovo for lead, zlato for gold,
srebro for silver, etc.
2. Trade marks: a BMW /bi:emdabelju/ in English is a BMV
/beemve/ in Serbian

3. Technical terms: the English computer may be either kompjuter or


raunar in Serbian
4. International organisations: transliteration causes that UNESCO
/juneskou/ is UNESKO /unesko/, and translations of meaning cause
that OSCE (Organisation for Security and Cooperation of Europe) is
OEBS (Organizacija za evropsku bezbednost i saradnju) in Serbian.
Absolute lack of correspondence
Conversely, the area of absolute impossibility of correspondence
covers terms and concepts that cannot be translated at all. In such cases, the
translator has a legitimate right to explain a certain concept in an appropriately
inserted footnote in a written text whereby the term is usually put into the TLtext in its original form, even between quotation marks, and the meaning is then
offered in the reference note. If a term should be mentioned in an oral
rendering, the translator will have to insert a short but skilfully derived
explanation of the term:
1. English: a drive-in-movie in Serbian would have to be film koji se
gleda na velikom platnu i to tako to se sedi u automobilu koji se parkira
na mestu predvienom za to; a take-out meal in Serbian would be
obrok koji se kupi u restoranu a nosi se kui ili za poneti, the
second being a rather modern equivalent.
2. Serbian: ljivovica, kaamak, ranji, svekrva in English would have to
be a type of alcohol drink made of plums, a kind of porridge made of
corn flour, the mother-in-law that is the mother of the husband.

Particularly problematic are lexical units, that is words and phrases,


referring to concepts used in a regional dialect because they will most likely not
have an equivalent in the TL. For instance, the word merak, belonging to the
regional dialect spoken in the south of Serbia, will have to be explained in a
footnote.
e.g. Merak is a certain kind of feeling referring to the pleasure and
indulgence in soulful music, alcohol and beautiful women as well as the
atmosphere provoked by it. This term is used in the south of Serbia and
it often appears in the works of Bora Stankovic, an author whose regular
theme was to write about lost youth and the wish of the aged person to
feel the merak once again.
In an oral translation the explanation might be less elaborate.
e.g. Merak is special kind of feeling of pleasure and indulgence.
Sometimes the translator can find a cultural equivalent if there is one in
the TL, such as with ice-cream parlour which may be poslastiarnica. In this
example the translator knows that both venues mean similar things to the
people going there: school children like to drop in after school, mothers meet
there for a cup of coffee when taking a walk with their babies or younger
children, ice cream is served in cones, the cakes are sold per piece, the music is
not as loud as in some modern cafes, etc.
Partial correspondence
It should be noted that the translator is most likely going to deal with
partial correspondence which is the most frequent form of relationship between
two units belonging to two different language codes whereby the attention of

the translator is facing the most difficult challenges because he will have to
make a choice based on knowledge referring to more than just the languages he
is dealing with.
e.g. the term administration.
According to the Oxford Advanced Learners Dictionary, the 7th
edition, the term administration is explained as follows:
1. the activities that are done in order to plan, organize and run a business,
school or other institution: Administration costs are passed on to the
customer. - the day-to-day administration of a company
I work in the Sales Administration department.
2. the process or act of organizing the way that sth is done: the
administration of justice
3. the people who plan, organize and run a business, an institution, etc.:
university administrations
4. (often Administration) the government of a country, especially the US:
the Bush Administration - Successive administrations have failed to
solve the countrys economic problems.
5. the act of giving a drug to sb: the administration of antibiotics
In Serbian, the first adiminstration would probably be uprava, the second
sprovoenje, the third administracija, the fourth vlada and the last one would
be izdavanje. But this is not the end; for what about management, head office,
department, government, rule, reign? These are all words standing for the
Serbian uprava. And what about the term paperwork? In Serbian it would have
to be administracija since there is no other term! The answer is simple: the
correspondent term has to be adjusted to the context of the whole text.

To conclude, the area of correspondence depends upon both the


linguistic and the extra-linguistic knowledge of the translators because they are
the ones to identify the term to be coded and eventually they are the ones to
think of the best possible equivalent. A theory might not be helpful especially
because there is no such thing as a unifying theory of translation which might
either accept or dismiss the statement from the beginning that there is no such
thing as translation.
Component analysis
Linguistic theories trying to help translators find the best possible
theoretical standpoint also mention lexical decomposition and atomisation of
meaning or the component analysis. However, the term itself is not helpful at all
because in any case, what a translator has to do is invest time and effort in order
to find the best possible and the most corresponding term in the TL for a term in
the SL which is why the process of decomposition will be highly dependent
upon correspondence.
The component analysis is a technique that is used to describe the
meaning of the word in the SL so as to find the best equivalent in the TL by
reconstructing the meaning in the TL. This partially complicated definition is
standing for a partially complicated procedure which is based on the
assumption that certain structural characteristics appearing on different levels of
a term in the SL may appear on different levels in the TL as well. What the
translator has to do is analyse the meaning of the SL term by decomposing it,
that is breaking it down to its component or constituent parts and then find

direct or indirect translations of each component/ constituent in the TL in order


to reconstruct the meaning of the whole term in the TL.
e.g. the word boy may be broken down to its component meanings
including: [+HUMAN], [+YOUNG], [+MALE]. If now translated into
Serbian, the component meanings [+LJUDSKO BIE], [+MLADO],
[+MUKO] will inevitably lead to the word deak.
This technique has been praised and criticised at the same time because
it can be regarded as being economical but subjective as well. Yet many
translators use this technique when trying to find corresponding equivalents
belonging to one semantic field especially when related to other words in the
form of collocation, that is fixed phrases of words belonging together. The
translation of collocations is probably one of the most difficult areas in
translation because word for word translations are mostly incorrect. Different
synonyms of one and the same term in the SL do not always correspond to the
same synonyms in the TL. Also a very important fact is that the basic structure
of every semantic field contains one conceptual nucleus which may be
identified in every lexeme belonging to the semantic field. It is this conceptual
nucleus that is bearing the essential component meaning of each and every
member of a certain semantic field.
e.g. the conceptual nucleus [+HUMAN] of the above mentioned
example boy is at the same time a component by itself. When adding
new components, such as [+YOUNG], [+ADULT] or [+FEMALE]
and by combining them, we will arrive at lexemes such as fellow,
young man, girl, young woman etc. which in Serbian would be
momak, mladi, devojka, ena.

The technique of component analysis is most useful when faced with the
necessity to understand the relationship between similar words which otherwise
might be difficult. We might argue that this technique is equivalent to defining
words in the way a dictionary does it which also tries to stress the similarities
and differences between lexemes belonging to the same semantic field. Of
course, extra-linguistic knowledge must accompany the component analysis
because otherwise it would be impossible to find appropriate translations.
e.g. living room [+room in the house], [+family members sit there
together, talk or watch TV], [+visitors are invited to that room]
A literal translation based on absolute correspondence would render ivea
soba. Since we know there is no such thing, it is much better to resort to the
component analysis so as to find an appropriate equivalent. When the
components are translated into Serbian, we will get:
[+soba/ prostorija u kui], [+lanovi porodice sede u njoj,
razgovaraju ili gledaju TV], [+obino se posetioci pozivaju u tu sobu/
prostoriju]
Our extra-linguistic knowledge tells us that the Serbian word boravak is used
to denote staying in one place and that staying in one place in the house or at
home usually refers to daytime, that is dnevni in Serbian. Thus we arrive at
the term dnevni boravak which we know is used for the term living room. Of
course this is a somewhat simple example but even in the case of more complex
ones, the procedure will be the same.

Conclusion
Obviously the greatest area of difficulty a translator is faced with is to
avoid inaccuracy. Unfortunately, in the absence of an appropriate word,
translators, probably desparate because of their inability to find a proper term,
decide to simply substitute one lexical item for another resorting thus to the
most dangerous techniques of all: the word-for-word translation. What we may
end up with, is not only an incorrect and inaccurate translation, but also serious
confusion.
Instead of a conclusion, let us take a look at an example of a a
somewhat hilarious situation found in the area of the film and television
business where translators have to deliver translations of a film, series or some
other programme in a very short time. Be it either on grounds of lack of time or
competence, the result is well, lets just say funny.
e.g. In one episode of Las Vegas, Danny, one of the main characters, is
talking to Derek, who is about to get married to De Linda. Derek is very
nervous and in the attempt to clear his doubts referring to his fiancs
determination to marry him, he asks Danny:
De Linda is not getting cold feet, is she?
The translator missed the fact that this is an idiom. So what he
suggested in the translation was that she needed somebody to keep her
feet warm!

Notes
1. According to Prado (C. G. Prado, 2006) extralinguistic reality shows

how things are in the world. In other words, the way we perceive things in the
world determines the way we create and use language.
2. The system of categorisation generally refers to the activities that we
apply to a certain context within which information can be apprehended which,
in turn, influences the semantic information - the information about reality available to the individual (Jacob, E. K., 2004). In other words, we look at
different things in the world around us, identify characteristics according to
which they may be defined as similar and put them into categories. For
instance, cats, dogs, cows, horses, etc. are similar in the sense they have four
legs and a tail. In comparison to human beings, who have only two legs and no
tail, we learn about the category of the animal. When children start acquiring
language, one of the most important process is categorisation whereby category
is like a container into which items are put according to similarity. In short,
categorisation is one way of structuring reality.
References
orevi, J. 2008 Semantika uslovljenost srpskog poslovnog jezika in Prevodilac asopis
udruenja naunih i strunih prevodilaca Srbije, 3-4, str. 15-28
Jacob, E. K. 2004 Classification and categorization: a difference that makes a difference in
Library Trends, Wntr 2004
Prado, C. G. (2006) Searle and Foucault on Truth in Foucault Studies, Cambridge: Cambridge
University Press, No 5, pp. 118122, January 2008, p. 118
Oxford Advanced Learners Dictionary, seventh edition (2005), Oxford University Press
Seleskovic, Danica 2007 Obrazlozeni pedagoski prilaz nastavi konferencijskog prevodjenja,
Beograd: Udruenje naunih i strunih prevodilaca

Vous aimerez peut-être aussi