Friday, 28 November 2014

But that’s another blog

by Haris Gazdar

From left: Arafat, Sheikh Mujibur Rahman, Bhutto, Gaddafi in Lahore, Feb. 25, 1974. AFP

Who is interpreting? 
 From left: Arafat, Sheikh Mujibur Rahman, Bhutto, Gaddafi in Lahore, Feb. 25, 1974. AFP

The question first arose many years ago when my colleagues and I started to design the qualitative component of a study on access to land.  We thought we should use the opportunity provided by that study to develop our own protocols and training manuals for conducting qualitative research.  We read a lot of material that was available then on research methodology, and got busy with serious debates while trying to complete various research tasks.  What were the relative merits, functions and complementarities of qualitative and quantitative approaches (another blog)? How does a social policy focus allow qualitative research to escape the more narcissistic indulgences of contemporary anthropology (another blog)?  Why was it important to insist on the distinctiveness of qualitative social science research from similar-feeling field approaches such as participatory appraisals and action research (another blog)? How the term ‘data’ needed to be constantly rescued from the monopoly of statisticians (another blog)? Why was rigorous qualitative research anything but woolly, and usually much harder work than numbers (definitely another blog)?

I must admit asking these high-brow questions made me feel clever.  But I now think they pale into insignificance before what I almost dismissed then as a logistical problem: how do we do social science research in a place where we must work with multiple languages?  Our reports and publications must be in English. Our clients, national and international, worked in English.  The academic and policy literature that we needed to read to keep up with our clients and fellow-researchers was all in English too.  (This was just as well for me – a person who knew English but couldn’t tell apart a cotton plant from a rose bush and under whose care both would surely die, was principal investigator in a study of land and agriculture.)  Much of the team’s verbal communication was in Urdu.  And empirical material that we used, whether generated by ourselves or by others, usually involved the use of third and sometimes fourth languages.

In qualitative studies questions had to be translated first from English into Urdu, and then interpreted from Urdu into other national languages used in the country.  Interview responses, similarly, had to be translated back, sometimes stage-wise, into English.  There were logistical issues, of course, in working with multi-lingual teams who had to conduct interviews in different languages, and then convey back responses from their interviewees into Urdu or English.  Qualitative research training manuals that we had got our hands on rarely bothered about language, while for us language, interpretation and translation were often the most important issues.

Words mattered in qualitative research, but how about quantitative surveys?  Surely these were all to do with numbers.  We sometimes forget that all numbers in the social sciences are generated only after many words had been spoken, and yes, translated, retranslated, interpreted, and then translated back. In fact, words had to be chosen much more carefully when translating for quantitative surveys because diverse responses needed to be put in a ‘box’.  To a seemingly simple question like “do you own the house you live in?” there were multiple answers which defied ‘boxing’.  “The structure is mine but the land belongs to the landlord. He can evict me but I will take the malba (literally debris).”  “It is my house on village land, and the village belongs to my community as a whole”.  “It is my house on village land which belongs to the government, but I will need to move if the village community evicts me.” And let’s not even get started on what it meant for a woman to own a house.  These arrangements could be found in a single village and people had words in their own language to refer to each of them and more – words that were not incidental but fundamental to an understanding of social organisation, economic opportunity, and political affiliation.

So, if much of what we do is translation or interpretation, what are the rules? What is of value? Translation and interpretation imply conversation and dialogue.  Who are the interlocutors?  Our clients? The wider academic community which we would like to regard as our peers? Individuals from whom we solicit information? The subjects of research or policy-making? Problem?  Surely, our clients and academic peers justify their interest in the subjects because they regard them as the principals - oxymoron!  We say we are interested in landlessness because we want to address the interests of the landless.  So is this a conversation initiated by our clients or our peers who want to know more about the condition of the landless? Or do we say it is a dialogue initiated by the landless who want to convey their problems so that academics would apply their minds and policy-makers their resources to solving them? I’m still working this one out.

Meanwhile, we certainly don’t do what we do out of charity or pity. That would be patronizing in terms of personal conduct and also compromise our ability to do research, oops, interpret well.  Overcome by pity we will direct conversations to the self-serving goals of charity, but that’s another blog. And, of course, we know it’s not charity because we are paid reasonably well, another blog!