The Translation of Hidden Quotations

Brian Mossop


Have you ever suspected, while translating a text, that the sentences you are reading were not composed by the author but rather were lifted from other documents? In this article, we’ll look at these ‘hidden quotations’.


The advent of LAN servers and intranets has made it exceptionally easy for employees of large bureaucracies to write documents by cutting and pasting passages from existing documents available in the organization’s various stores of electronic information. For example, the part of a report which sketches the background can be prepared very quickly in this way. Aside from speed considerations, employees may favour this approach over personal composition if they do not enjoy writing, or know they are not very good at it! They may find it possible to use whole chunks of existing documents, either unchanged or with a few deletions or additions.


Assuming (as we will throughout this article) that you are translating from French to English, here are the situations you may be facing:


·        The passage you are reading has been lifted from a document that was originally written in French. It’s possible that an English translation already exists.

·        The passage you are reading originated in a document that was first written in English. That document was subsequently translated into French, and the passage now before you was lifted from that translation by the author of the text you are working on.

·        The passage you are reading originated in a document first written in English. What you are reading is a translation into French prepared personally by the author of the text you are working on. (This type of quotation might also be called a ‘hidden translation’.)


Normal quotations

The options for translators are well known in the case of normal quotations—those which are identified as such by some means such as quotation marks, an indented paragraph which is introduced as a quote, or a bracketed reference to the source document.[1] The two possibilities are:


1.      find the original English, or the previously prepared English translation

2.      if no English can be tracked down, then prepare an English translation yourself in one of two forms:

·        as indirect speech (according to X…….; Y says that……..);

·        inside quotation marks with a suitable signal of status (“[translation from French]……….”; “[not original English]………….”; “ ………….[my translation]”).


In principle, you will look for the existing English, especially if it is original English rather than a translation. How quickly you give up, and resort to preparing your own translation, will depend on such factors as the use to which the translation will be put, the time available for translation, and any instructions in this regard from the client. With hidden quotations, the situation is different. Unlike normal quotations, hidden ones do not impose any duty to search for the source document. The sentences are being presented as the author’s own, and can therefore be translated as if they were original. Whether you conduct a search at all, or how long you spend searching, will depend strictly on the expected costs and benefits.


The cost in time of searching for hidden quotations

The big difference between normal quotations and hidden quotations is that any sentence or sequence of sentences could be a hidden quotation. It is not uncommon to find texts which turn out, upon investigation, to consist mostly of material lifted from other documents—sometimes a considerable number of other documents. A key consideration, then, is how much time you will devote to tracking down these sources.


If you have access to a database of relevant previous translations, then you may be able to track down previously prepared English translations of original French fairly quickly. This will especially be so if you have Translation Memory software that can automatically compare your entire text to the French texts in a translation database and retrieve the English translation of every passage found in that database. You may also be able to find the original English of documents which have previously been translated into French (and are being cited in French in the text you are working on). However most often the material you are looking for will not be available in a translation database. You can try various Internet/Intranet search methods, for example:

·        guess a set of perhaps five or six English words that might have appeared in the original English you are looking for, and enter these in the search engine at;

·        find a web site that you have reason to believe will contain the document, and enter your keywords in its Search utility.

If such automated methods don’t work in fairly short order , then you need to ask yourself just how important it is to find the source material. You must weigh the cost (in time) against the likely benefits.


Unenlightening finds

Just how much better will your final English be if you hunt down sources rather than prepare your own translation? Often the differences will be trivial; that is, the difference in meaning between your translation and the source document will be trivial, though the vocabulary and syntactic structure may differ considerably. Here’s an example:


[French text to be translated – actually a quotation lifted without change from a French document at an Environment Canada web site]

L'ozone porte atteinte aux sacs alvéolaires des poumons au moment de l'échange entre

l'oxygène et le gaz carbonique. Ces sacs sont faits d'un tissu mou, spongieux, qui se durcit et réduit ainsi la capacité des poumons.


[English produced by translator]

Ozone attacks the alveoli in the lungs when oxygen and carbon dioxide are exchanged. The soft, spongy tissue of the alveoli hardens under the influence of ozone and reduces lung capacity.


[English at Environment Canada web site]

Ozone harms the air sacs in the lungs where oxygen and carbon dioxide are exchanged. This soft, spongy tissue gradually hardens and reduces the capacity of the lungs.


In this particular instance, only these two French sentences were lifted from the web document; the remainder of the paragraph was written by the author of the French text sent for translation. So it might have taken less time to simply translate the two French sentences than to find the web site, locate the two relevant English sentences, and cut-and-paste. With a lengthier hidden quotation, of course, translation might well have taken longer than finding the web document, so the reward of searching might have been greater. On the other hand, since with hidden quotations there is no way of knowing ahead of time whether an English document does in fact exist, the time spent searching the Internet may prove a complete waste. And as the above example illustrates, even if the document does exist, it may be only trivially different from your own translation.[2]


Comprehension benefits

Searches to find original English documents are much more worthwhile when you are having problems understanding the French text that has been sent for translation. If you can locate an English source document, it may clarify the meaning. Here’s an example, from a text on the teaching of mathematics:


[French text to be translated]

…une approche d'enseignement axée sur la réflexion collective en classe portant sur des symbolisations paraît pertinente, en ce sens qu'elle produit un discours qui semble supporter la réification de l'activité mathématique et contribuer ce faisant au développement de la pensée mathématique. Ce discours est caractérisé par des changements de perspective répétés (shift of attention), de telle manière que ce que les élèves et l'enseignant accomplissent en action devient subséquemment l'objet de réflexion.


[English translation produced without benefit of English source document]

..a suitable teaching approach would seem to be one based on group reflection in the classroom concerning symbolization. Such reflection creates discussion which seems to support objectification of mathematical activity and thus contributes to the development of mathematical thinking. The type of discussion in question is characterized by repeated shifts of attention, so that what the students and the teacher do subsequently becomes the object of thought.


Unless you already know about the pedagogy of mathematics, you may find several points here unclear. For example, just what does ‘qui semble supporter’ mean? In other words, jut what is the relationship between ‘discours’ and ‘réification’? A web search using the keywords ‘collective’, ‘reflection’ and ‘symbolization’ produced a hitlist whose first item was the following abstract of an article in a journal on mathematics education. I have italicized the wordings which are of interest.


Reflective Discourse and Collective Reflection


The analysis in this paper focuses on the relationship between classroom discourse and mathematical development. We give particular attention to reflective discourse, in which mathematical activity is objectified and becomes an explicit topic of conversation. We differentiate between students' development of particular mathematical concepts and their

development of a general orientation to mathematical activity. Specific issues addressed include both the teacher's role and the role of symbolization in supporting reflective shifts in the discourse. We subsequently contrast our analysis of reflective discourse with Vygotskian accounts of learning that also stress the importance of social interaction and semiotic mediation. We then relate the discussion to characterizations of classroom discourse derived from Lakatos' philosophical analysis.


The full article was even more enlightening. It contained the following passage:


Our purpose in this article is to suggest possible relationships between classroom discourse and the mathematical development of students who participate in, and contribute to it. To this end, we focus on a particular type of discourse that we call reflective discourse. It is characterized by repeated shifts such that what the students and teachers do in action subsequently becomes an explicit object of discussion.


Clearly the last sentence of the French text is the author’s own translation of the last sentence of the above passage from the full article. This same idea comes up several times in the article, and on one occasion it is expressed as “making what was previously done in action an object of reflection” (cf French ‘réflexion’). It is clear from the article, however, that the reflection in question is not internal (within the mind); it is a matter of discussion in the classroom—what the English article calls ‘collective reflection’. So ‘object of thought’ in the above English translation is perhaps not the best way of expressing this idea.


As regards the problematic ‘qui semble supporter’, it is clear that the mathematical activity is objectified in the course of discussion. That is, the discussion has the effect of making mathematical activity itself the topic of conversation (rather than just a set of procedures the students use). It would thus be better to write ‘a discourse which objectifies’ rather than ‘a discourse which seems to support the objectification’.


A reading of the article also shows that the translation ‘reflection in the classroom concerning symbolization’ may mislead the reader. What is involved here is not the process of creating symbols but the physical symbolic objects themselves (the translator wrongly took ‘des symbolisations’ as a synonym of ‘la symbolisation’). In the example discussed at length in the article, the students give answers to a mathematical problem, and the teacher writes these answers on the board in the form of a table of paired numbers (the symbolization). One might perhaps say that the table becomes the topic of discussion (‘reflection concerning a symbolization’), but it is really more a matter of the teacher’s table helping the students to objectify their mathematical activity. (This might have been a good place for the French writer to use the verb ‘supporter’: ‘réflexion collective supportée par des symbolisations’.)


The English article further suggests that the translator should perhaps not use the expression ‘shift of attention’, which the author has inserted in the French text. This expression occurs nowhere in the English article, the authors of which call it simply a ‘shift’ or else a ‘shift in discourse’. It’s a shift from doing mathematics to talking about it. The French ‘changement de perspective’ captures this nicely, but ‘shift of attention’ could be misleading.


The abstract and article also of course help with the rhetoric and terminology of this field. For example, it seems that French ‘discours’ should be rendered as ‘discourse’. Let’s look more closely at this benefit of finding source documents.


Terminology benefits

Hidden quotations are often revealed in the course of normal terminological research if this is conducted using an Internet search engine or a web site’s Search utility. For example, in a text on numerical weather prediction from the Canadian Meteorological Centre in Montreal, the following passage appeared:


Au même moment, une nouvelle version du modèle GEM en configuration globale sera installée. Les changements principaux dans cette version du modèle sera l’élimination de la diffusion horizontale explicite (sauf près des pôles de calcul et du toit du modêle), et l’utilisation d’un schéma d’interpolation plus précis par splines cubiques afin d’augmenter le niveau d’activité dans le modèle. La diffusion horizontale au toit du modèle sera également augmentée afin d’amélorer l’efficacité de la couche éponge à la limite supérieure du modèle.


I already knew that ‘diffusion horizontale’ is ‘horizontal diffusion’ but wasn’t sure about ‘explicite’. I entered the English phrase ‘explicit horizontal diffusion’ in the Google search engine. The sixth item on the hit list was:



... of the elimination of filters on the topography, elimination of explicit horizontal

                diffusion (except near the computational poles and model top) and the use ...


A trip to the indicated site revealed the following paragraph:


At the same time, a new version of the GEM model in the global configuration will also be implemented. The main modifications, made to increase the level of activity in the model, consists of the elimination of filters on the topography, elimination of explicit horizontal diffusion (except near the computational poles and model top) and the use of the more accurate cubic spline interpolation. In addition, the horizontal diffusion at the top layer has been increased to improve the effectiveness of the sponge layer at the model upper boundary.


The term ‘explicit horizontal diffusion’ is confirmed, and the term ‘computational poles’ is discovered as a bonus. Since the site in question was the client’s site, I felt confident in using ‘computational poles’ without any further checking to ensure this is genuine English rather than a not-so-good translation. (In the days before the Internet, I would have left it at that, but now I could increase my confidence by a quick check in Google—which immediately took me to original English meteorology documents containing ‘computational poles’.)


Stylistic issues

The English document found at the CMC website raises an interesting stylistic question. Notice that the relationship between the English and French is not clear. The content of the English does not exactly match the French (and other paragraphs in the English document are more distantly related to the French, or not related at all). It is possible that the author of my French text was drawing from a translation of the above English, or was translating it personally. But it is also possible that the above English is itself a translation of an originally French text which the author of the text being translated then borrowed, with modifications. Clearly the English text and the French text being translated have a common ancestor, but it is not clear which is the original.[3] In such cases, the decision to use or not use the English document you have found may rest on whether you believe it is original English written by a native speaker of English. If it is a translation, or if it was not written by an Anglophone, and you cut-and-paste it, you may be importing unEnglish phraseology and style features that are not in fact used by Anglophone experts in that field.


Another stylistic question arises when hidden quotations are interspersed with original writing by your French author. If you cut-and-paste English material from the source document, how will you deal with the original parts? Unless you attempt to imitate the style of the source documents, your final translation may be rather bumpy stylistically speaking, with constant shifts back forth between the document’s style (or the multiple styles of several documents) and your own style.


Hidden mistranslations

What if you track down an original English document and discover that your author has either personally mistranslated it into French, or else lifted someone else’s mistranslation? Will you restore the original English or backtranslate the mistranslation? The problem is more difficult than the one which arises when authors try to quote French material from memory; for example, they cite a portion of the organization’s mission statement, but don’t get it quite right. Here the author’s intent is obvious, so you can simply insert the correct English wording of the mission statement. With hidden mistranslations, however, matters are different. Since the material is being presented by your author not as a quotation but as his or her own writing, do you not have to backtranslate the mistranslation? Consider this example:


Your French text contains this passage:

En outre, les effets de la contamination des eaux souterraines ne s'arrêtent pas avec la perte de réserves d'eau de puits. Plusieurs études ont porté sur la migration des contaminants depuis les lieux d'élimination ou de déversements jusqu'aux lacs et cours d'eau voisins puisque les eaux souterraines font partie du cycle hydrologique, processus sur lesquels les connaissances restent fragmentaires.


You find the original English of which the above is a translation, and it contains:

Several studies have documented the migration of contaminants from disposal or spill sites to nearby lakes and rivers as this groundwater passes through the hydrologic cycle, but the processes are not as yet well understood.


Clearly the person who wrote the French misunderstood the English word ‘as’, which means ‘while’ here, not ‘because’. The consequence is that your translation of the French will be nonsensical unless you restore the original English. In other cases, however, you may need to backtranslate the French as it stands, because if you restore the original English, the subsequent sentences won’t make sense. For example, if the sentence following the problematic sentence is the French author’s own, the argument it contains may depend on the author’s misunderstanding of the English passage he translated.


Copyright and plagiarism

If in the course of translating a text you find extensive hidden quoting, you may wonder whether any legal or moral issues arise—copyright infringement or  plagiarism. Obviously when writers quote from their own organization’s store of documents, there is no problem because the legal owner, as well as the source of the ideas being lifted, is the organization itself rather than the individual employee who wrote the quoted text. The quotations are in effect self-quotations. But how far afield does this licence to lift extend? For example, what if someone writing on behalf of a Canadian Government department lifts material from organizations of which Canada is a member, such as the United Nations? What about material lifted from the web site of a private-sector company with which the Government has dealings? What about material from an on-line journal? If your translation is going to be published, you might want to point out to the client any hidden quotations which you have found.



A search for source documents is especially likely to be rewarding if you cannot see what your French text means. You may also find some useful terminology as well as phraseology used by experts in the field. Dangers include wasting time, stylistic unevenness, and potential legal issues.

[1] In one text type (comments on a report or proposal), the existence of a source document is self-evident, and any part of the text before you may turn out to have been lifted from the original report or proposal.

[2] There are two ways in which the original English is better than the translation: (1) Strictly speaking, alveoli are outpouchings in the alveolar sacs, not the sacs themselves. But this is not an anatomy text, so ‘alveoli’ will do. However a reading of the English source document may prompt the translator to think about the future readership. If (as in this case) the readers are non-scientists, ‘air sacs’ may be preferable. (2) ‘where’ is better than ‘when’. On the other hand, the translation is better in two ways: (1) the translator’s addition of ‘under the influence of ozone’ (an idea implicit in the French) may help make the translation clearer to a lay audience than the original English. (2) The inter-sentence link ‘the soft, spongy tissue of the alveoli’ is smoother than ‘this soft, spongy tissue’.

[3] Several of the English texts at this site have been translated from French, sometimes it seems by Francophones, as indicated by telltale errors such as the wrong tense in a later passage of the above document: “significant improvements are therefore expected when these new datasets will be implemented”. A somewhat mysterious oddity is that both the French and the English texts cited above contain the same grammatical error: the main modifications…consists of; les changements principaux…sera.