Lost in translation: when research requires more than one language

IR/security research is often by its very nature international, and can involve empirical work in any number of different places and/or languages. When your research is done in one language and has to be written up in another one (in Western academia, usually English), translation is an essential part of the research and writing up process. However, when I began my PhD I found there was very little advice available on the issues surrounding translation, and most of what I learned before my fieldwork came from informal talks with colleagues and supervisors. So, I thought it might be useful to write a blog-post about some of the problems and pitfalls of translation during research, many of which I have first-hand experience with, followed by some ideas on how to solve them, or at least minimise their impact.

As anyone who has carried out research involving translation will know, it is rarely straightforward or clear-cut. There are no clear rules for how to approach it, but while there is no ‘right way’ to do it, there are many bad ways to go about it. Some of the key issues to think about are listed below, but beware: these come largely from my own experience and what worked for me may not work for you. Likewise, you may disagree with some of the opinions I’ve formed based on my experience, which is fine (of course) – the main point of this is to get you thinking about the big issues, and to find your own solution to them.

There often seems to be an endless list of problems with translating, but without some form of translation we would struggle to understand the world outside of our own little corner. When something is translated, another layer of interpretation is added to the original, and some meanings, nuances or details may be lost. Translation is never neutral: it necessarily involves inserting the translator into the data and transforming it. The translator is cannot be extracted from the translation process: s/he has a lot of power as a meaning-clarifier and maker. Personally, my research has a discursive focus, which has influenced both my thinking on and my approach to translation. Looking at ‘discourse’, particularly if you’re planning to use discourse analysis (either in a language other than your native one or in your native language which then has to be translated into another language during your writing-up) poses some further problems (more on this later). Below I cover some common issues countered during translation, divided into ‘practical’ and ‘linguistic/cultural’ problems, and then suggest some strategies to avoid these.

The title of this blog post may or may not have been a cheap ploy to justify illustrating it with a picture of Scarlett Johansson.
The title of this blog post may or may not have been a cheap ploy to justify illustrating it with a picture of Scarlett Johansson.

Practical problems:

  • Firstly and most important of all: are your language skills good enough? Here you need to bear in mind that the skill set needed depends very much on the type of research you’re doing. Do you need to do interviews? Archival/documentary research? The type of data you’re using will affect the skills that you need, and will help you target your language learning: think about whether speaking/listening may be the most important for you, or whether perhaps reading comprehension is more important. Ethnographic research will of course require much more extensive language skills than other forms, but also has more literature providing guidance on this.
  • Will you be or have you considered using a translator or interpreter? (More on that below)
  • This is slightly controversial, but I don’t think you have to be fluent, or a native speaker to research in another language. But you do have to be good enough for what you need to do, and that requires some reflexive thinking and self-awareness.
  • If you’re not a native speaker, it’s important to be open and disclose your language skills (and limitations) in the research outputs.
  • The time factor: if you’re not a native speaker, translations can take a lot of time. This is particularly true if your research involves processing a large number of documents.

Linguistic/cultural problems:

  • Things don’t always translate well – not everything has an equivalent in another language. Some texts, words or expressions will therefore not have a direct translation, and you need to have a strategy for dealing with these.
  • On top of this, some terms or words can have multiple meanings in the original language, which can be either lost in translation or mistranslated. This was a big concern for me, doing research in Chinese – and one of the many reasons I decided to conduct my interviews in English. Many words have multiple meanings (particularly in Mandarin), and your interviewee may be purposefully vague. If so, your translation may put words in her or his mouth that s/he would not have said, by giving their words a more specific meaning in the translation. It’s worth noting that this is a problem for native speakers too (in cases where they have to translate during the writing-up process).
  • Knowing a language, even to a fluent or native standard, does not automatically mean that it will be easy to carry out research interviews in that language. In practice, ‘research language’ is very different to usual language skills. Official discourse is often very different to everyday talk and in many cases uses a different vocabulary, which can make it tricky even if you’re a native speaker used to ‘everyday talk’.
  • Lastly, cultural knowledge is an important part of understanding a language. When something is translated, it is taken out of its discursive and cultural context, and in the process it can lose some of the meaning, or things can appear significant when they are not (and vice versa). A couple of examples: this article is by a Russian speaker who researches North Korea, and explains how some things translate much better from North Korean into Russian than they translate into English, because North Korea and Russia have a shared communist history and many North Korean expressions are borrowed from Soviet Russian. Translated into English, this context is lost. During my own research on energy security in China, I found a lot of discourse around China’s ‘scientific’ outlook and approach to energy security. It was only during a conversation with a fellow (non-Chinese) China-researcher that I began to view this more as a Marxist trope common in Chinese official discourse than something with more significance to China’s approach to energy security specifically. It’s easy to overlook or miss things, but it’s worth remembering that it can be just as easy for native speakers (who sometimes get home blindness). It does mean it’s important to be extra careful, however.

Until the day this magic button appears..
Until the day this magic button appears..

Avoiding pitfalls:

Strategy 1: Don’t translate

While this strategy may sound counterintuitive, in many instances translation can be avoided. For example, if you’re studying official discourse, it’s often possible to access official translations of key documents in English. Using this strategy, you need to be aware that the texts are prepared with a foreign audience in mind (which has important implications). Personally, I decided to avoid translating where possible. I conducted my interviews in English, which limited number of interviewees (not ideal, but I decided it would be better than the risk of misrepresenting my interviewees). In my case, this was less of a problem as my research relied primarily on documentary material. Many of my key documents were available in official translations, though I had to translate others. By translating documents where no official translation was available I also avoided limiting analysis to documents ‘pre-approved’ (and, therefore selected for translation) by the Chinese government. In the case of some particularly important documents I also checked official translations against their original (Chinese language) versions. This was important as official translations can at times exclude particular (often potentially controversial) sentences or sections of documents when produced for a foreign audience.

Strategy 2: Use a native speaker

If you don’t speak the language you are translating from to a native standard, get your translations checked over by a native speaker. If possible, do this anyway as a second pair of (bi or multilingual) eyes can only help.

Strategy 3: Use a translator/interpreter

Using a translator or interpreter can save a lot of time and hassle, but can also be problematic (and expensive!). There’s no such thing as a perfect translation between two different languages: your translator/interpreter is using their own knowledge and skills to filter, manage and transform the information given by your interviewee/document into a language that you understand. This can add extra layers of meaning to your data, so the usefulness depends very much on the focus of your work. It can be particularly problematic if you’re using discourse analysis, as you’re no longer analysing the discourse of your original source but translated version. If you do decide to use a translator or interpreter, make sure that you are clear with them about exactly what you need them to do – if you need as close to a perfect translation as possible, make sure that they know this. If you’re conducting an interview, go through the questions and your interview strategy with your interpreter beforehand – the wording of your questions may not actually make sense if they are translated literally.

4: Online translation engines

When processing a lot of information, online translation engines like Google Translate can be helpful to filter your material. While they are largely useless when it comes to actually translating articles and documents, they can help you identify what’s actually worth translating or spending more time on. If you have a lot of material, this can be a helpful strategy to narrow down your focus.

Lastly, it’s important to remember that there’s no such thing as a perfect translation. Translation is not a science, and your skills will get better with experience.

Ps. If anyone can recommend any useful resources about using translations and translating during research, please either add it in the comments section below or email me, and I’ll add them at the end of this post!



Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s