As someone who can spend a fair amount of time looking for moderators in Angola or wondering how best to organise depth interviews on human rights in Iran, I was intrigued to read recent AQR Networker exchanges concerning a project on issues of translation in qualitative research.

The more I read, the more they became relevant. Data collection in one language and presentation in another, translating cultural differences, who should translate what in the course of a qualitative project ­ all are pertinent to my work as a researcher, and to my background in modern languages.

Equally interesting was the response. There was much sympathy from those active in international research, plus interesting but depressing anecdotes about the consequences of not making proper allowance for language issues. I also detected the occasional inference that language and translation issues were not really a problem.

By contrast, I see a language as the key to a culture, which may function in very different ways to our own. It gives insights into how people conceptualise the world around them, and expression to their view of that world and their place within it.

I'm sure many of us have seen instances where inadequate translation can reduce serious ideas to subjects of derision. The result, whether this be over-reliance on automated translation or the assertion that anyone can translate or interpret if they know a little of the language, is the same: the point can be well and truly missed.

Those exchanges rightly made the point that interpretation, not just translation, is key. That's fine if you are in a position to interpret your raw material yourself, but much more difficult when you are reliant on third parties for your understanding of what is going on.

When I'm working with Albanians or Bengalis, whose languages I don't understand, I have to know what to make of the translated transcripts. I've seen enough to know that, had I not been there, my understanding would have been very different on the basis of the printed material alone. I can't always get to see all the groups I commission, so I need to be sure that the translation is properly done. Only then can I begin to interpret.

For many of us, it is perhaps a misfortune to have been brought up speaking the world's principal lingua franca, since it can easily desensitise us to crucial cultural differences. And with AS/A2 entries in modern languages in sharp decline, things don't look set to improve.

Translation issues will not go away, and will remain central to proper and meaningful interpretation of our international data.