Translating Articles in the Humanities and Social Sciences.

Franck Barbin @en

Image 1 : Cartoon by Priestman Atkinson from the Punch Almanack (1885).

Aim of the study : debunking stereotypes on translation.

My study mainly consists in analysing precise examples rather than developing long theories, as the readers of this journal are not translation experts, but would benefit from a translator’s viewpoint for their own research. The study focuses on written articles and leaves oral communications aside because they entail different translation issues (e.g. ways to address an English-speaking audience, simpler syntax, or words easy to pronounce for a non-native English speaker).

The corpus of the article is extracted from my own experience as a freelance translator and tackles different fields, i.e. management, law, psychology and even geography. All the examples are either translated from English to French or from French to English.

My study aims to debunk misconceptions about translation, more specifically in academic environments. I chose to divide my presentation in six main stereotypes :

1) Translating is easier and quicker than writing. Wrong. Translating is a long-term process which involves, among other things, language skills, content-based knowledge, appropriating the rationale and ideas of the author, and re-writing the whole article, the translator becoming a co-author of the article.

2) Machine translation yields good results. Wrong. Even if machine translation is a useful, time-saving tool for translators and researchers, it cannot replace human translators who are still needed at least for post-editing.

3) Literal translation ensures the quality of the translation. Wrong. Being faithful to the words in the source language text would often sound awkward in the target language ; the translated article would sound like a translation, without intending to do so for stylistic reasons. A case in point is, when authors can translate their own articles, they rephrase their own work and usually choose a different perspective from their original article ; they do not convey exactly the same meaning in the translated article.

4) Everyone speaks the same English. Wrong. Each country — or even states for larger countries —use English differently and have developed their own English. Authors should therefore write their articles differently when they submit their paper to a British, American, Australian, Canadian or Indian journal. This factor also depends on the scientific community you belong to, as it will superimpose specific terminology, phraseology and standards to tackle the same subject.

5) Scientific discourse is unambiguous. Wrong. Even among experts in the same field, they sometimes misunderstand what their colleagues intend to say and could judge their work as inaccurate even if they basically agree with all their ideas.

6) Concepts are universal. Wrong. They often create misunderstandings because authors do not define notions in the same way. Concepts may encompass different realities from one country to the other.

Each of the aforementioned misconceptions about translation will be exemplified in the body of the article. But, before giving such examples drawn from my experience as a freelance translator, I found it interesting to compare traditional methods to translate in the Humanities and Social sciences (HSS) with a new method I devised during a research project in management.

Question of method : translating questionnaires.

My interest for translating survey instruments started when I joined a research project in management in 2007. This joint project on mentoring practices and emotions at work was carried out between English research unit CREW (Centre for Research into Emotion Work and Employment Studies) at Brunel Business School, and French group of researchers Ekip (which means “team” in Albanian) from MINES Paris Tech, Nancy ICN Business School and Rouen University.

Conventional method.

After reviewing the literature on the translation of survey instruments, I quickly realized that I would not be able to use the conventional methods applied by disciplinary experts and that I would have to devise a new method instead. I am not the only one to have come to this conclusion. Janet Harkness, who wrote extensively on this topic (Harkness 2003, 2007, 2011), is well aware of the fact that translation procedures should be improved and questionnaires could benefit from the expertise of professional translators and linguists, as much of the literature neglects modern translation theory and research (skopos1, cognitive theory):

Even today, many projects rely on procedures variously called “back translation” to check that their survey translations are adequate. In its simplest form, this means that the translation which has been produced for a target language population is re-(or back-) translated into the source language. The two source language versions are compared to try to find out if there are problems in the target language text. (Harkness 2011, p. VIII-2)

Nevertheless, the back-translation method is still the most common method used to translate questionnaires and is arguably considered as an effective procedure to limit translation and interpretation errors (Davoine and Méda 2008). This procedure relies heavily on the “ask-the-same-question” (ASQ) principle (Smith 2004), in other words it is tied down to the source text. This widely employed approach presupposes that closely translated questions will produce an adequate instrument in the target languages. At the time of the project (2007-2008), Sylvester Ivanaj, Professor of Information Systems at ICN Business School, Nancy, also wanted us to adopt a 6-step back-translation process : (1) translating the English questionnaire (Q0) into French (Q1) ; (2) back-translating Q1 into English (Q2) ; (3) comparing the original source questionnaire (Q0) with Q2 ; (4) validating the items having the same meaning ; (5) revising the items having a different meaning ; and (6) following the same process with another translator. We managed to convince the team members to drop this method and use a different procedure instead. The main disadvantage of back-translation is to mask comprehension difficulties which could have been avoided if discussed early : translators tend to automatically correct obvious mistakes without saying it. This is an old problem : as early as 1956, some authors had already suggested that questionnaire translation and wording were the weakest link in comparative studies (Kumata and Schramm 1956).

Original approach.

The main advantage of the method I came up with was to integrate translation into the study design. Translation issues began well before the source questionnaire had been finalized, almost at the same time. The original questionnaire written in French lost its sacrosanct status in translation. My role was no longer to merely translate the French questionnaire into English ; I could alter the original French questionnaire when needed and discuss it with the other team members. These constant alterations made to the source text are related to what is described by Michaël Oustinoff (2001) and could be called “backlighting” : the two texts mutually complement and enrich each other as two versions of the same text. I attended all meetings between the English and French teams, which enabled me to have a good understanding of the objectives, theoretical tenets and constructs of the study. It proved to be an invaluable asset. My role in the project was to ensure that the items of the questionnaires were equivalent in French and English, in other words that they conveyed the same meaning and triggered the same reactions to both audiences. It is a decisive step if we want to compare the results of the questionnaires, as intended in the study protocol. I had to make every endeavour to prevent misunderstandings and possible ambiguities in both versions of the questionnaire. Among other questions, I wondered whether the item was clear enough in both languages, or if the wording was strange or shocking. The translated questionnaire had to read naturally and be tailored to the British audience. The interview and pre-test phases enabled us to further refine the wording of both questionnaires. In order to render an instrument appropriate for fielding in other cultures, the instrument has to be adapted, which usually affects content and formulation. This step is crucial to reach culturally comparable instruments.

The method I used with Kate Darlington, a visiting researcher to The Centre for Research in Emotion Work, Brunel Business School, was divided in two phases. We alternated places half through the process in each phase. In the first phase, we implemented a role-play based on the English questionnaire : one of us played the role of the interviewer and the other a respondent. The point was not to answer the questions as such, but to rephrase them and explicit the ideas behind them. We found it was the best way to set us free from the linguistic yoke and come up with a clearer, more idiomatic wording. In the second phase, we tried to answer the questions according to our own experience. The intent was twofold : once again, to rephrase the answers, but also to find new ideas and add missing answers to the questionnaire. We found this creative approach quite useful to improve the wording of the questionnaire and it helped to trigger equivalent answers in both countries. In sum, it reinforced the validity and reliability of the questionnaire. Obviously enough, the other team members could also share in new ideas, thus providing a fruitful exchange of views (Barbin 2011).

Just like Janet Harkness, I am fully aware of the fact that, in practice, “translation rarely is seen as part of questionnaire design and usually is treated as an addendum” (2003, p. 35), coming at a later stage. However conclusive the new approach I devised was, it should be noted that even members of this project tend to come back to the easiest translation procedure, despite its flaws : “All questions were asked in French, which is the local language. Therefore, we used a back-translation process to offer proper equivalence between English and French (Malhotra et al. 1996)” (Harfouche and Pezet forthcoming, p. 15). It proves that there is still a long way to go to adopt a team approach to survey translation. What is more, even though the translator’s voice is better heard when choosing the wording — because of his or her linguistic and topic expertise — the final word comes to the disciplinary experts.

Debunking stereotypes on translation : examples.

After giving the example of an approach to improve the quality of translation in the humanities and social sciences, I will illustrate the six main stereotypes I listed in the first part of my paper. As they are the most obvious, I will analyse the first three preconceived ideas more quickly and explain the last three more in depth. To apply the principles of confidentiality and privacy, I have decided not to include the names of the authors in all the examples I will develop below, unless the author agreed to appear in the paper.

1) Translating is easier and quicker than writing.

Even though translating is often viewed as a straightforward process, it not only requires a linguistic expertise in both languages, but also an excellent knowledge of the subject (terminology, phraseology, state-of-the-art ideas, etc.) and a very close analysis and comprehension of the context, leading to choose the most appropriate translation tailored for the audience.

To illustrate this first stereotype, I chose to analyse the translation of a real estate contract. The “As is” Residential Contract For Sale And Purchase I had to translate had been approved by the Florida Realtors and the Florida Bar. When translating this American form into French, I had to convey the same terms and conditions of the contract while abiding by the phraseology and terminology used in French contracts. Contracts differ between the United States and France, but the translator has to follow the norms and conventions of each country. It not only requires a good knowledge of American and French contracts, but also to conduct extensive documentary research.

For example, the American contract starts with the mere indications :

The French translation had to add the phrases “dénommé ci-après” (“hereinafter referred to as”) and “d’une part […] d’autre part” (“on the one hand […] on the other hand”) to comply with French standards :

Likewise, for similar reasons, the French translation of the contract had to expand the simple verb “agree” into “a été fait et convenu ce qui suit” and add the whole phrase “sous les conditions suspensives ci-après énoncées” (“subject to the conditions set forth below”) which was not in the original contract :

“agree that Seller shall sell and Buyer shall buy the following described Real Property and Personal Property (collectively ‘Property’) pursuant to the terms and conditions of this AS IS Residential Contract For Sale And Purchase and any riders and addenda (‘Contract’).”

Il a été fait et convenu ce qui suit :

Le vendeur, par ces présentes, vend, sous les conditions suspensives ci-après énoncées, à l’acquéreur, qui accepte d’acheter, le bien immobilier, composé du bien réel et des biens personnels, (dénommé ci-après ‘le bien’), conformément aux clauses du présent compromis de vente de biens immobiliers en l’état et de ses éventuels avenants et addenda (dénommé ci-après ‘le contrat’).”

2) Machine translation yields good results.

Even though the quality of machine translation gradually improves, it is no match for human translation. It is a useful tool if you want to get a general idea of an email for instance, but it shows its limits when high level translation is needed. It is particularly true in research. A linguist colleague of mine asked me to help her to choose which machine translation was the best for her abstract. She had used Systran and Google Translate. Even if the translation yielded by Google Translate was slightly better, it was punctuated by many mistakes and even sometimes made no sense. So I ended up translating the whole abstract for her.

One sentence of the original French abstract was : “Il [l’article] met en outre au jour des schémas de répétitions qui rappellent du ‘déjà-dit’ et engagent un à-dire.” The translation produced by Systran read : “It puts moreover at the day of the diagrams of repetitions which point out the ‘already-known as one’ and urge one with-to say,” which literally means : “Il met de plus lors de la journée des diagrammes de répétition qui soulignent le ‘déjà-connu comme répétition’ et qui incitent un ‘avec-à-dire’. The translation produced by Google Translate read : “It also puts to light regimens repeats reminiscent of the ‘already-said’ and engage in meaning”, which literally means : “Il met également à jour des répétitions de régimes qui rappellent du ‘déjà-dit’ et s’engagent dans la signification”. Needless to say that neither translation was satisfactory, even if Google Translate version was slightly better. A foreign researcher fluent enough in English could have written : “Moreover, it puts to light some schemata of repetitions which remind of an ‘already said’ and engage a ‘to-say’”. It would have been better and more easily understood, although it would have signalled it was a translation. Here is the translation I offered instead : “It also unravels repetition patterns echoing an ‘already-said’ and initiating a ‘yet to say’”. There is no need for long speeches to demonstrate the difference in quality between machine and human translations.

3) Literal translation ensures the quality of the translation.

Being faithful to the words in the source language text would often sound awkward in the target language. It can be sometimes intended when translating literary work — for stylistic reasons for instance — but it is most often unintended in academic writing. The translation must read naturally, as if the text were originally written in the target language, and it has to follow the terminology and phraseology of the field in the target language — how native experts would have approached the matter. I will take an example from an article in psychology I translated into French.

The original English sentence read : “In order to encourage such academic research to take place in a controlled design, it is important to provide and highlight the existing clinical data with this regard”. Two main pitfalls had to be avoided when translating it. First, the French translation of “academic research” should not be patterned after the English phrase (“recherche académique”). It is more accurate in this context to speak of “études universitaires” in French. Second, you should not take the words “controlled” and “design” separately, and prefer “cadre expérimental contrôlé” (3 780 certified instances) to “plan expérimental contrôlé” (122 certified instances). The translation I offered was : “Afin d’encourager la réalisation de telles études universitaires dans un cadre expérimental contrôlé, il convient de fournir et de diffuser les données cliniques existantes à ce sujet”.

After having described my new approach to translate questionnaires in the previous section, I think it is time to illustrate it here. I have chosen to take a key question which initiated a series of related questions and items. Considering its impact on the rest of the questionnaire, we gave special attention to it. The first translation of this question was too literal and not explicit enough :

Versions 1 : “A laquelle allez-vous vous référer ici d’un point de vue chronologique ? 1ère, 2e, 3e, autre.” // “Where is this mentor placed chronologically amongst your influential relationships? First/second/third.”

Therefore, we came up with a new wording, both in French and English :

Versions 2 : “Par rapport aux dates des rencontres avec ses personnes, à laquelle allez-vous vous référer ?” // “Out of these influential people, where does the most influential one stand chronologically ?”

This second version of the question failed to get its point over. We used this wording in the pre-test phase, but the respondents judged that it was not immediately understandable in the post-test interviews. The team thus offered another version of the question :

Versions 3 : “En fonction de l’ordre dans lequel vous les avez rencontrées, à laquelle allez-vous vous référer ? 1ère 2e 3e autre.” // “Please choose the most influential manager and place him/her chronologically amongst all your influential relationships.”

I was still unconvinced by the wording of the question because this third version did not intrinsically make the question clearer. I succeeded in convincing the other team members to opt for a longer question which was less ambiguous and more idiomatic :

Final versions : “Parmi ces personnes, choisissez celle à laquelle vous allez vous référer dans ce questionnaire. Situez-la par rapport aux autres en indiquant si elle est la 1ère, la 2e, la 3e… que vous avez rencontrée.” // “Please choose the most influential manager and indicate if he/she was the first, second, third… mentor you met during your career.”

Even when the translator joins the team from the very beginning, he or she can give an outside perspective on the questionnaire, pay more attention to the wording of the questionnaire and avoid literal translation so that the questions have the same impact on the respondents in both languages.

4) Everyone speaks the same English.

The most obvious difference among English speaking people is how differently they master the English language and specialised languages. When using technical terms in an article, you should always make sure your readers understand them, even if they come from the same scientific community. Researchers tend to forget this as they assume that, even if their readers are coming from different countries, they share the same field of expertise and therefore know the terminology used. It is most especially true when they use a term which is culture bound.

As I was invited to deliver a lecture by Hervé Regnauld, professor in physical geography at Rennes 2 University, I will first illustrate this point with a geographical example, more precisely with the term “kistvaen” I used in my PhD on the translation of Devon folktales (Barbin 2005). I doubt all researchers in geography know this Celtic word that designates a stone coffin especially found in Devon, England. The translator therefore needs to explain the implicit of the source text to overcome an information deficit, what Jean-René Ladmiral calls “incrémentialisation” (1994, p. 219), a type of explicitation designating a target addition on the signifier and/or signified level. More precisely, Michel Ballard indicates that delayed explicitation (“incrémentialisation différée”) forms a comment added each time the word is used. This process is not limited to translation, but can appear in any text ensuring cross-cultural contact (Ballard 2001). Let us examine to what extent this process of explicitation can be helpful when translating the word “kistvaen” :

“The kistvaens, the country-folk say, were set up for the purpose of hoarding vast treasures” (Whitcombe 1874, p. 50). No other indication is given about kistvaens in the near context. The translator is left with two choices. Either he/she decides to stick to what is strictly said in English and not to guide the reader : “Les kistvaens, comme les appellent les gens du coin, étaient érigées en vue d’abriter de vastes trésors.” This would probably obscure the text, which was not intended in the source text. Or he/she chooses to explicit the text as he/she assumes the reader will not understand this word and is not familiar with this English county : “Comme les appellent les gens du coin, les kistvaens, ou tombes que l’on trouve dans les landes désertiques du Dartmoor, étaient érigées en vue d’abriter de vastes trésors.” This second option is also favoured by Marianne Lederer :

Le simple fait de faire figurer le vocable inconnu à côté de ce qui en est l’explication simplifie la tâche du lecteur sans pour autant modifier le texte. Il est parfois nécessaire de fournir une explication plus poussée lorsque le texte renvoie à des faits culturels ignorés du lecteur et sur lesquels le contexte n’apporte aucun éclaircissement. (1998, p. 166)

We can compare it with another quotation which chose to directly explicit the word “kistvaen” in the source text :

“Turnhill, or Blackslade, Kistvaen (marked as a ‘Cairn and Cist’ on the OS map) is a Bronze Age tomb in which the cremated remains of a prehistoric warrior were once interred.” (Bamberg 1993, p. 7)

This second quotation has the advantage to add two other terms pertaining to the archaeological vocabulary, i.e. “cairn and cist”, which helps the reader to understand the meaning of the word “kistvaen”. Once again, the translator can choose between two translations. Either he or she decides that no further explanations are required because the French reader will understand the words “cairn” and “ciste” :

Le kistvaen de Turnhill, ou de Blackslade, (marqué sur la carte d’état-major en tant que ‘cairn et ciste’) est une tombe datant de l’Âge de bronze dans laquelle ont été autrefois enterrées les cendres d’un guerrier préhistorique.”

Or he/she chooses to add footnotes to explain the meaning of these two words :

Le kistvaen de Turnhill, ou de Blackslade, (marqué sur la carte d’état-major en tant que ‘cairn2 et ‘ciste3’) est une tombe datant de l’Âge de bronze dans laquelle ont été autrefois enterrées les cendres d’un guerrier préhistorique.”

The reference to “OS maps”, or Ordnance Survey maps, is much more common in the field. Great Britain’s national mapping agency should be known by geographers and the translation would anyway replace it by its French equivalent, “carte d’état-major”.

Image 2 : Kistvaen in Drizzlecombe on Dartmoor in South Devon, UK.

The problem to understand English is not limited to technical terms ; it expands to much more usual words. It would be misleading and illusory to believe that, even among native English speakers, the English language refers to a unique language. The English language has evolved differently in the varied countries using English as an official language. Therefore, the same notion has often be declined in different ways from one country to the other.

A case in example is the idea of “lawyer”, which is very intricate and ever-changing with new regulations. I do not claim to define this idea once and for all, but I simply intend to prove that the legal profession is called differently in major English-speaking countries and/or that the same word can have different meanings from place to place. In the United States, there is no real distinction between “lawyers”, “attorneys” and “counsels” or “counsellors”, which are general terms for anyone licensed to practice law (Hill 2002). These three words are often used interchangeably, even if a slight distinction is made by the American Bar Association ( 2012). Likewise, a term can be preferred according to the situation : the word “counsel(lor)” is mostly a term of address for an attorney in court, just like in the sentence “Counsel may speak to the defendant”. As a title, only the word “attorney” is used : “Attorney Robert”.

In England, there are two kinds of lawyers having different roles. Barristers are traditionally lawyers who have been admitted to plead at the bar, i.e. represent their clients in court (Bar Council 2014). They play a similar role as “trial attorneys” in the United States. Conversely, solicitors traditionally only provide legal support and advice to clients. They can handle wills and the transfer of property, just like a “notaire” in France. But, since the Courts and Legal Services Act 1990, the distinction between these two branches of the legal profession has no longer been as cut and dried as it used to be. Solicitors can now appear in the lower courts and even increasingly in the higher courts.

Canada does not draw a clear distinction between barristers and solicitors, and has even fused the two professions. Canadian lawyers are qualified as both barristers and solicitors and can therefore practice both (Studin 2010).

In South Africa, Scotland and India for instance, the word “advocate” is favoured and roughly encompasses the same role as a “barrister”, that is to say advocates have the right to appear in any court. It should be noted that this word can be confusing : in England and the United States, it has a much broader sense and refers to any person supporting a cause or policy. In sum, researchers should always keep in mind that the same English word can have different meanings from place to place and make sure they choose the right word in a given context.

5) Scientific discourse is unambiguous.

Ambiguity is allegedly the exclusive domain of literary discourse while scientific discourse is characterised by its monosemy — a one term-one meaning relationship. However, ambiguity is inherent to all discourses, whether fictional or non-fictional. Researchers tend to forget that even scientific discourse cannot communicate ideas in a totally accurate way and can be unintentionally ambiguous, i.e. can be understood in two or more possible ways. I will more particularly tackle here structural ambiguity, that is to say a sentence or a clause that seems to have several meanings, what Massimo Poesio calls “perceived ambiguity” : “perceived ambiguity is characterized in terms of inference” as a conflict between the hypotheses produced during discourse interpretation, which is “subject to two constraints : the Anti-Random Hypothesis (interpretations are not generated at random) and the Condition on Discourse Interpretation” (ambiguity has to be resolved) (Poesio 1996, p. 197-198).

The translator has no other choice but to clear up any ambiguity in the source text. As it is almost impossible to keep the same ambiguity in the target text, the translator has to choose what the most likely meaning of the sentence is, using typical disambiguation techniques (e.g. contextual evidence, syntactic and semantic constraints, collocations, field knowledge, etc.). I will illustrate this point using two examples taken from the same psychology article :

“Children in the sample included males and females that were between 8 and 14 (mean 11.26) years.”

I was first tempted to translate this sentence by : “Les enfants de l’échantillon comprenaient des garçons et des filles âgés de 8 à 14 ans (moyenne de 11,26).” The wording in English is equivocal : does the sample consist of adults and children or only of children ? The near context would urge the translator to favour the first hypothesis. However, this first impression would have been wrong because, two pages later, the author indicates that : “A heterogeneous sample of children with working memory impairments (mainly ADHD4) completed the standard protocol,” proving that the sample consisted of only children. With the benefit of hindsight, I changed my first translation into : “L’échantillon était constitué d’enfants de sexe masculin et féminin âgés de 8 à 14 ans (âge moyen de 11,26).” This final translation clearly indicated from the start that the sample only included children, preventing the reader from misunderstanding the author’s original intent.

The second example suggests another level of ambiguity, deep down the structure of the language :

“In particular, the analysis focused on parent ratings of inattentive symptoms in children, as well as adult self-report of inattention, ADHD symptoms, and cognitive failures.”

Two different translations are grammatically correct, but only one proves a good understanding of the field. The problem originates from the scope of the word “adult”. If “adult” not only governs the phrase “self-report of inattention” but also the two other phrases “ADHD symptoms” and “cognitive failures”, the French translation should read :

Cette analyse s’attache tout particulièrement aux évaluations parentales des symptômes de l’inattention chez l’enfant ainsi qu’aux auto-évaluations de l’inattention, des symptômes de TDAH5 et des échecs cognitifs chez l’adulte.

Even if this first interpretation follows grammatical rules, it is not in accordance with the psychological truth : “parent ratings of inattentive symptoms in children” are obviously made by parents and “adult self-report of inattention” by adults. Conversely, “ADHD symptoms” and “cognitive failures” affect both adults and children. So the translation should not state it otherwise. The right translation should be :

Cette annexe s’attache tout particulièrement aux évaluations parentales des symptômes de l’inattention chez l’enfant, aux auto-évaluations de l’inattention chez l’adulte, aux symptômes de TDAH et aux échecs cognitifs.

Therefore, when an author needs to translate one of his articles, he or she must choose an expert translator in their field to prevent this type of errors. Another minor change was made to make the translation clearer : the word “analysis” was not logically translated into “analyse”, but the word “annexe” (“appendix”) was preferred as this analysis was outside the article (in appendix 2) and not in the article as such.

In sum, a translated article is always less ambiguous than the original article because it has already gone through a privileged reader — the translator — who had to follow the process of understanding it and thus tries to facilitate the reading of the article.

6) Concepts are universal.

Researchers tend to believe that concepts will be similarly understood by people sharing or using the same language or at least by people having the same area of expertise. It is actually far from it. Even researchers sharing the same culture, language and domain can put a different meaning on the same notion. That is why researchers should clearly define the notions they use in their articles to ensure that their ideas are not misunderstood.

A case in point will be the translation of the title of a research project on mentoring I participated in. A major problem quickly arose when we had to translate the concept of mentoring :

Titles 1: “Recherche sur les relations de mentoring” // The relationship between mentors and their work-based mentees”.

We can right away rule out any purely linguistic issue from the researchers : half researchers were French, the other half were British, and some were even bilingual. The root of the problem was quickly diagnosed : cross-cultural comprehension. The notion of “mentor” has a totally different impact in France and the United Kingdom : the idea of mentoring is too restrictive in France while it encompasses a much wider reality in the United Kingdom — it is even used in TV programmes for children. Thus, if we had kept these two original titles, most French managers would have felt unconcerned by the survey and would not have answered the questionnaire. Conversely, many British managers would have missed the point of the project and considered too many individuals as mentors, which would have triggered many unwanted answers. The aim of the project was totally different : rouse the same interest for our questionnaire in both countries. This led us to alter the titles in both languages :

Titles 2 : “Recherche sur la coopération entre managers” // “Research project on the relationships between managers”.

This second attempt was still unsatisfactory. The French word “coopération” was once again too restrictive and was lacking an emotional dimension, which was an essential part in the project. In contrast, the English word “relationships” was too general and respondents could have thought we were focusing on personal relationships. Therefore, an adjective had to be added in both cases to refine the purpose of the study :

Final titles: “Les relations interpersonnelles entre managers” // “Influential relationships within British and French environments”.

As can be noted, there is not a strict identity between the French and English titles of our project, but rather an equivalence on the discourse level. What we had in mind was to arouse the same interest for our project in France and the United Kingdom and obtain comparable answers. For instance, the French team considered it was useless to indicate in the French title that the project aimed to compare French and British working environments as it was already clearly stated in the introduction : “Ce questionnaire est proposé à des managers français et anglais ayant une expérience professionnelle d’au moins 10 ans” and “à des fins de comparaison internationale”.

I will take a second example from my experience as a freelance translator6 proving that concepts are not universal : the French notions of “parenté” and “parentalité”, which can be translated by parentage, parenthood or even kinship, depending on the case and the country. The concept has also evolved in time. The terminology used in the field is somewhat inconsistent and the legal, biological and social conceptions of the term often interfere with each other. I do not attempt to settle the debate. My aim is to make researchers realise how difficult it is to translate concepts. Thus, they need to clearly define the concepts they use. This is all the truer when several terms are used to convey the same idea in the same article, mainly for stylistic reasons to avoid repetitions. They often fail to keep in mind that different terms cannot have exactly the same meaning.

In France, to put it simply — at the risk of sounding over-simplistic — the concept of “parentalité” involves the process of taking care of children educationally while “parenté” involves the biological relationship between parents and children. It should be noted that only the notion of “parenté” has a legal basis in France. Highlighting the flaws of the concept, Yann Favier (2014) advocates to clearly and legally redefine the genealogical link between adults and children, and redefine the legal relation between education, duties and inheritance.

In the United States, “parentage” is used to officially establish a parent-child relationship between a child and parents. This definition was extended to unmarried parents after the Uniform Parentage Act (UPA) was passed in 1973, but only a handful of states have adopted the 2002 version of it, which means that paternity laws are very different from one state to the other. For instance, in many states, only the mother or the alleged father can bring a paternity suit to identify the father of a child. However, in states that have adopted the 2002 UPA, any interested party — e.g. presumed fathers or equitable fathers — is allowed to file a paternity action in court. Likewise, since the 2002 UPA, the legal definition of “father” has been expanded to include other types of persons, thus changing radically the notion of “father” in some states (LegalMatch 2012).

In Australia, the problem to define the notion of “parenthood” is quite similar to that of the United States :

Each Australian state and territory has legislation to guide decisions about who is a parent for the purposes of other of their laws that refer to the obligations or “rights” of a parent. These differ significantly between jurisdictions, especially at this time when most jurisdictions are engaged somewhere in the process of law reform on the issue. (Library Council of NSW 2012)

In Canada, “parentage” is defined as follows :

Who a person’s parents are. Generally, if a child is conceived naturally, a child’s parents are his or her birth mother and biological father. This can be different if a child is adopted or born as a result of assisted reproduction. (Legal Services Society 2014)

In some cases, the concept of “psychological parenthood” — a legal parent-child relationship based on the assumption of a parental role — is legally recognised in Canada and has been considerably extended since the 1968 Divorce Act (Bala and Ashbourne 2012). In sum, the Canadian law only recognises three types of filiation : by blood, by adoption and by a parental project of a person or a couple. However, once again, being officially considered as a father or mother of a child varies by country and province. For example, in Quebec, a child can only have two parents at a time, but it can be two fathers or two mothers (Educaloi 2014).

In England, the term “parenthood” is preferred in a legal context and is directly dependent on three other regulations :

It is more likely for children not to have a legal relationship with their biological father than it would be with their biological mother. This is due to a set of complex rules deriving from the law of marriage (1), the law relating to birth registration (2) and the law relating to the context of artificial reproduction (3). (Steiner 2006, p. 3)

Even if, just like in Canada, the child’s best interests prevail, the determination of legal parenthood is a very contentious issue and tensions arise between different conceptions of parenthood. To put it in a nutshell, comparing the definition of the concepts of parenthood and parentage in different countries have clearly indicated the need to define concepts which are definitely not universal.

Guidelines for writing and translating articles in HSH.

When researchers do not seek the services of a translator, they should directly write their articles in English instead of first writing them in French and then translating them into English on their own. Their final articles would result in a text easier to read and understand. They should keep in mind that the translated article is not a mere copy of the original article, but an alter ego of the original article having its own way of presenting ideas. To trigger an equivalent comprehension and reaction in the target language, researchers have to adapt their writing and their way of tackling the topic. Otherwise, their ideas are likely to miss their points. Hence, literal translation is not a guarantee of quality. Researchers sometimes fail to realise that languages are culture bound and that each language has its own way of capturing reality.

Similarly, researchers tend to believe that the English language is unique, and that technical terms and everyday words are the same in English and have the same meaning. They sometimes fail to note that the English language has evolved differently in the varied countries using English as an official language. Therefore, researchers should clearly establish if they have to write their article in British, American, Australian, Canadian, South African, etc. English.

When researchers are not native English speakers, they do not totally master the English language and their way of writing English is very often culturally biased by their native languages. Therefore, attempting to write elaborate and intricate sentences in English would often lead to mistakes and misunderstandings. They should choose a simpler syntax instead and not hesitate to break down long sentences into two or more sentences. Researchers tend to believe that simple and clear sentences will be considered as over-simplistic, but clarity is always a quality to be sought after and valued in articles.

Researchers should equally be careful about automatic translation tools. Even if these tools seem convenient, they need to have a deep understanding of the English language to be able to choose the translation which is the most adapted to the situation, while discarding inappropriate translations. These tools can be used when they already have a clear idea of the translation they want, otherwise it will obscure their message or even make it nonsensical.

Researchers should pay special attention to concepts and notions used in their articles and clearly define them so as to ensure that readers — even when they are also experts in the field — can easily and properly understand what is intended in the article. More generally, researchers have to avert misunderstandings and clarify ambiguities. Likewise, they should take into account the fact that their article will be also read by people who do not share the same culture as theirs. Thus, they have to look for all cultural references in their article and precisely explain what they mean to bridge the cultural gap with their readers.

Finally, researchers have to comply with the editorial requirements of the journal. Even though it should go without saying, my own experience as a chief editor of a journal and reviewer for several journals proved that this prerequisite is not always applied. Researchers have to precisely follow the editorial requirements and adapt their style, terminology and phraseology to the expectations of the editorial board of the journal. Hence, their reviewers will be more willing to accept their papers.

When researchers seek the services of a translator, they should always choose a translator who masters the field of the article and is familiar with writing research articles. Otherwise, even though the article would be linguistically correct, it would sometimes convey a different meaning as the translator would have failed to properly understand the original idea of the author. The choice of the translator is therefore a key element to have better chances to be published, but researchers should not blindly trust the translator : they should double-check the translation and interact with the translator when questions arise.

This article is a work in progress and only a first step to guide researchers in writing and translating articles. Further research is needed to help non-native English researchers writing articles in English.


Bala, Nicholas and Christine Ashbourne. 2012. “The Widening Concept of Parent in Canada : Step-Parents, Same-Sex Partners, & Parents by ARTAmerican University Journal of Gender Social Policy and Law, vol. 3, n° 20 : p. 525-560. Ballard, Michel. 2001. Le nom propre en traduction. Paris : Ophrys. Bamberg, Robert W. 1993. Haunted Dartmoor : A Ghost-Hunter’s Guide. Newton Abbot : Peninsula Press. Barbin, Franck. 2005. “Problèmes liés à la traduction des récits populaires du Devon.” Thèse de doctorat, Université de Caen/Villeneuve d’Ascq : Presses de l’ANRT. —. 2011. “Dépasser la dualité en traduction : l’équivalence en matière de questionnaire” in Mariaule, Michaël and Corinne Wecksteen (dirs.). Le double en traduction ou l’(impossible ?) entre-deux, vol. 1, p. 133-144. Arras : Artois Presses Université. Educaloi. 2014. “Filiation : the Legal Link Between Parents and ChildrenEducaloi. Favier, Yann. 2014. “Réflexions sur le dualisme parenté/parentalité et sa traduction juridique” Texte présenté lors de l’atelier “Parentés et parentalités” à l’EHESS (Paris), le 7 avril 2014. Harfouche, Antoine and Eric Pezet. Submitted/Forthcoming. “The Dinner Game Effect : A Study about the Antecedents of Unethical Behavior in Sales Contests with Evidence from the Banking Sector ” Human Relations. Harkness, Janet. 2003. “Questionnaire translation” in Harkness, Janet A., Fons J. R. van de Vijver and Peter Philip Mohler (eds.). Cross-Cultural Survey Methods, p. 35-56. Hoboken : John Wiley & Sons. —. 2007. “Improving the comparability of translations” in Jowell, Roger, Caroline Roberts, Rory Fitzgerald and Gillian Eva (eds.). Measuring Attitudes Cross-Nationally, p. 79-94. London : Sage. —. 2011. “Translation” in Survey Research Center (ed.). Guidelines for Best Practice in Cross-Cultural Surveys, p. VIII.1-VIII.35. Ann Arbor : Survey Research Center, Institute for Social Research, University of Michigan. Hill, Gerald N. and Kathleen Hill. 2002. The People’s Law Dictionary : Taking the Mystery Out of Legal Language. New York : MJF Books. Ladmiral, Jean-René. 1994. Traduire : théorèmes pour la traduction. Paris : Gallimard, coll. “Tel”. 2012. “Attorney vs. Lawyer Lederer, Marianne. 1998. La traduction aujourd’hui. Le modèle interprétatif. Paris : Hachette, coll. “F”. Library Council of New South Wales. 2012. “Legal status of parenthoodfindLEGALanswers. LegalMatch. 2012. “Uniform Parentage Act and PaternityLegalMatch. Legal Services Society. 2014. “DefinitionsFamily Law in British Columbia. Maxwell, Beverly. 1996 “Translation and Cultural Adaptation of the Survey Instruments” in Martin, Michael O. and Dana L. Kelly (eds.). Third International Mathematics and Science Study (TIMSS) Technical Report. Volume I : Design and Development., p. 8.1-8.10. Chestnut Hill : Boston College. Oustinoff, Michaël. 2001. Bilinguisme d’écriture et auto-traduction. Julien Green, Samuel Beckett, Vladimir Nabokov. Paris : L’Harmattan. Poesio, Massimo. 1996. “Semantic Ambiguity and Perceived Ambiguity” in van Deemter, Kees and Stanley Peters (eds). Ambiguity and Underspecification, p. 159-201. Stanford : CSLI Publications. Steiner, Eva. 2006. “The Tensions Between Legal, Biological and Social Conceptions of Parenthood in English LawElectronic Journal of Comparative Law, vol. 3, n° 10 : p. 1-14. Studin, Rebecca. 2010. “Rebecca Studin’s Fox Scholarship Report 2009-2010” The Bar Council. 2014. “About barristersThe Bar Council. Vermeer, Hans J. 1989. Skopos and translation Commission. Heidelberg : Heidelberg Universitat. Whitcombe, Mary Elizabeth. 1874. Bygone Days in Devonshire and Cornwall, with Notes of Existing Superstitions and Customs. London : Richard Bentley.


1 Skopos theory is a functional approach of translation determined by its purpose or skopos (Vermeer 1989).
2 “Monticule ou tumulus de terre ou de pierre élevé par les Celtes en Europe, les Tibétains en Asie.” (Trésor de la langue française 1977, p. 15)
3 “Cercueil en pierre, composé d’un caisson de quatre dalles posées de champ, recouvert d’une cinquième et déposé en pleine terre ou sous un tumulus.” (Trésor de la langue française 1977, p. 847)
4 ADHD stands for Attention Deficit Hyperactivity Disorder.
5 TDAH is the French equivalent of ADHD and means “trouble du déficit de l'attention avec hyperactivité”.
6 This idea took solid shape when I had to translate Yann Favier’s abstract on this topic (Favier 2014) and I have compared the French situation he described to the situation in some English-speaking countries.


Debunking typical misconceptions on translation, more specifically in academic environments, helps me to establish guidelines for writing and translating articles in humanities and social sciences (HSS). Six main stereotypes are examined : (1) translating is easier and quicker than writing ; (2) machine translation yields good results ; (3) literal translation ensures the quality of the translation ; (4) everyone speaks the same English ; (5) scientific discourse is unambiguous ; and (6) concepts are universal. My study mainly consists in analysing precise examples — derived from my own experience as a freelance translator — and showing how to translate articles, rather than developing a comprehensive theory. However, the first part presents the new approach I devised to translate survey instruments.

To refer to this post

Franck Barbin @en, "Translating Articles in the Humanities and Social Sciences.",, , 18.11.2014