Article by Gunnar Persson on semantic nuance in translation in English  
HomeDictionaries by languageDictionaries by subjectDictionary searchLexicool shopTranslation online
Home > Resources > Articles > Article by Gunnar Persson on semantic nuance in translation
Share: Share this page on Facebook Share this page on Twitter Share this page on LinkedIn
What is 'Shift of Meaning'? An Attempt at a Linguistic Analysis

A tricky problem that assessors of translations are often faced with is how to draw the line between serious mistranslations and so called 'shifts of meaning'. For want of safe criteria it is often difficult to determine how serious a shift of meaning is and whether it should rather be regarded as a serious error. I will try here to find some linguistic criteria for such judgements and will begin by discussing some cases of erroneous translations from English into Swedish (the examples are fictitious).

1. The weapon used by the perpetrator was a sawn-off shotgun.
2. Det vapen som användes av förövaren var ett avsågat hagelgevär.(CORRECT)
3. Det vapen som användes av förövaren var ett avsågat gevär.(WRONG)

The translator of 3 has ended up on the wrong hierarchical level, since a shotgun is a special type of gun. In other words gevär is a so called hyperonym or superordinate term and hagelgevär is a hyponym or subordinate term, as shown in Figure 1 below, in which some other hyponyms are also shown:

Hyponymy system for 'gevär'
The designation 'gevär' in Figure 1 indicates that this is a superordinate term that is included in the meanings of all the hyponymous terms. The translation 'gevär' is therefore partially correct, but not precise enough, since the component hagel ('shot') is missing. The two words belong, however, to the same lexical field or domain, which may be defined as "small arms handled by both hands". Should this translation then be judged as a serious error or as a shift of meaning? In my view it should be regarded as a serious error, in particular if it occurs in a legal context where it may be of the utmost importance for all details pertaining to a law case to be correctly rendered. It may for example be of great importance in a case of murder to be given exact information about what kind of weapon was used when searching for and identifying the murderer.

This shows that it is not sufficient - at least not when translating a legal text - to use terms that are too general. It may be concluded that using the wrong term belonging to the same semantic domain as the correct term cannot automatically be judged as a shift of meaning. The next question is whether there may be similar cases in which the faulty translation can nevertheless be judged more leniently. There follows a case that might be taken from a general text:

4. He put a new record on the record player.
5. Han lade på en ny skiva på skivspelaren.(CORRECT)
6. Han lade på en ny skiva på grammofonen.(WRONG)

CED defines record player as "a machine on which you play records" and gramophone as "an old-fashioned type of record player". The Nationalencyk-lopedin ('The National Swedish Encyclopaedia' (NE)) defines skivspelare as "another name for gramophone" and grammofon as "a device for mechanical or mechano-electrical reproduction of sounds by means of a disc-shaped medium". This indicates an interesting difference between the two languages. In English record player is the superordinate term and gramophone the hyponymous 'a kind of record player', while skivspelare and grammofon are synonyms in Swedish, if one is to believe NE (whose definitions are sometimes disputed, however). According to BCL a record player is an "apparatus for playback of gramophone records", and among various types of gramophones are mentioned crank gramophones, electrical record players and CD players. BCL's definition, according to which record player is a hyperonym, is thus more in accord with CEED's. In the Swedish folk lexicon there is also likely to be a lot of confusion in this area, partly due to so called apparent time. Elderly people tend to call all types of record players gramophones, while younger people make a clear distinction between gramophone, 'playback apparatus with a pickup arm', and record player, 'digital playback apparatus'.

Another interesting fact that this case shows is that the source language and the target language may have different hierarchical relations between two terms that seem lexically to be direct equivalents. Such pairs might be regarded as a special type of "false friends".

One conclusion to be drawn from this case is that none of the possible translations can be judged as a serious mistake. Another more general conclusion is that a dubious translation cannot automatically be regarded as a shift of meaning, if it happens to be in the same semantic domain as the correct translation. On the other hand this conclusion cannot be drawn across the board. All translations are very much dependent on factors such as genre, target group, context, level of style, etc. Let us look at some other examples that may lead to other conclusions.

The following example might be taken from a general text:

7. He put the saucepan on the hot plate.
8. Han satte kastrullen på kokplattan.(CORRECT)
9. Han satte kastrullen på spisen ('the stove').(WRONG)

This example resembles no. 1-3, but is nevertheless of a different nature. This is not a case of the wrong choice of translation within a semantic domain. The relation between kokplatta ('hot plate') and spis ('stove') is not hyponymous (hierarchical) but rather a relation between the whole and one of its parts, that is, so called "meronymy". Everybody who knows what a modern stove looks like and how it works also knows that it has a number of hot plates. In this context it is possible to use truth-conditional arguments to determine how much deviant the erroneous translation is. The arguments are presented in 10-11 below:

10. If a stove has at least one hot plate, then 8 above means that the saucepan is normally put on both the hot plate and the stove.

11. Therefore 9 has the same truth-value as 8 in the sense that a saucepan that is put on a stove will in all likelihood also be put on a hot plate, if the action is to have a meaningful function.

In this case of a part-whole relation it would seem reasonable to judge the erro-neous translation as a shift of meaning, as our knowledge of the real world tells us that the effect of the described action should be the same irrespective of whether the referent is described as a stove or a hot plate. But again, this must not be taken to imply that all erroneous translations of this kind may be regarded as shifts of meaning. A somewhat similar case with the opposite effect is pre-sented in 12-14 below:

12. He put the joint in the oven.
13. Han satte steken i ugnen.(CORRECT)
14. Han satte steken på spisen (on the stove).(WRONG)

If a translator for some strange reason should render 12 as 14, then the error is different from the one discussed in 1-3. This is not an erroneous hierarchical description but a description of an event where the whole replaces one of its parts in a way that, due to the assessor's knowledge of the world, s/he cannot but judge as a serious error. While the relation between stove and hot plate may be regarded as a shift of meaning because of the truth relation between the two lexemes, in the oven and on the stove (the choice of preposition is significant here) has no such relation.

Obscure translations may be made for many other reasons, where both the translator's and the assessor's age may cause complications in the choice of translation and the judgement of it. Let us look at two English lexemes and their Swedish equivalents:

15. Lazily she reached out for her bottle of perfume.

To translate lazily in a stylistically acceptable manner proved impossible for a group of students of translation theory. My translation was:

16. Lojt (Lättejefullt) sträckte hon sig efter parfymflaskan.

None of the students (who were all 18 - 23 years old) knew the lexemes lojt and lättjefullt. Instead they suggested slött ('sluggishly'), which corresponds to neither the style nor the content of the text. While lojt and lättjefullt may have positive senses (just as lazily in the English sentence), slött is definitely a word with a negative meaning. This type of mistranslation may however be judged as a shift of meaning, since all the three words contain the meaning component 'unenergetic'.

This may perhaps serve as a tentative definition of what should regarded as shift of meaning, that is, using a word or an expression that has a basic meaning in common with the word that is correct in the context but in addition has a deviant axiological meaning.

BCL Lidman, S. et al. (eds) (2001) Bonniers compact lexicon. Stockholm: Bonnier lexicon.
CED Hanks, P. et al. (red) (1986) The Collins English Dictionary. 2nd ed. London, Collins.
Nationalencyklopedin (1989). Höganäs: Bokförlaget Bra Böcker AB.

By Gunnar Persson. Copyright © 2003.
Emeritus professor of English, translator, interpreter, language reviser
Published by, November 2003. permits the re-publication of this article on condition that the author agrees and that is clearly shown as being the original publisher. The information should incorporate a hypertext link to, and show the year and month of the original publication. Please contact us for further information.


Copyright © 2000-2020lexicool françaislexicool italianolexicool españolSeptember 2020