Thinking about Rwanda with Harvey Weinstein
I don't know how many of you recall the three posts (1, 2, and 3) I put up in mid-January in which I wrote about Eric Stover and Harvey Weinstein's new book "My Neighbor, My Enemy; Justice and community in the aftermath of mass atrocity"?
Well, ain't the world-wide web a wonderful thing? Sunday evening, I got a really interesting email from Harvey, in which he gently challenged a couple of things I'd said there.
(I had a similar experience not long ago when the Israeli researcher and writer Daniel Sobelman, whose work I mentioned when I posted here about Hizbullah, in December, likewise got in touch with me. There, without even a challenge. Yes, the WWW truly is remarkable.)
Anyway, back to Harvey M. Weinstein, who you might remember is a psychologist and a clinical professor in the School of Public Health at UC Berkeley. Okay, I can't totally remember how much detail about his professional credentials I put into those earlier posts. But take it from me he's one heck of a smart, well-informed, compassionate, and visionary guy.
His email gave me, as they say, "pause for considerable thought". So I went back and re-read the chapter in the book that I'd voiced some criticism of in light of his comments to me, and got back to him. That was Chapter 10, an important chapter in which he and two co-researchers present the results of a 2,000-person attitudes survey that they had a big team conduct, in 2002, in a four significantly distinct kinds of locations around Rwanda.
I wrote my criticisms in the 2nd of those three posts linked to above.
Here is our correspondence this week:
Subj: your review in Just World News
Date: 2/6/2005 6:23:33 PM Eastern Standard Time
From: Harvey Weinstein
To: Helena Cobban
Dear Helena
I have just read your comments on our new book "My Neighbor, My Enemy". Thank you for your taking the time to read it and for your kind words. I do remember our pleasant interchanges at Airlie House. Unfortunately, I think that your impressions of Chapter 10 are not quite accurate or helpful to anyone trying to understand what is happening in Rwanda currently. Let me try to clarify your concerns and explain why we think this paper is very important.
1. Studies such as these are not "quasi-epidemiological". This is a classic epidemiological cross-sectional survey design whose validity was reviewed with a fine toothcomb by the epidemiologists at the Journal of the American Medical Association where an aspect of this work was published.
2. A major point of the book is that data can and should be gathered in different ways to check whether the data are accurately portraying a picture of what is happening in the country. By "triangulating" epidemiological data with qualitative methods, we can assure that the picture is an accurate representation. In this case, the quantitative data accurately mirrored the qualitative data gathered through focus groups and interviews.
3. I want to emphasize that our researchers are not academics who parachuted into Rwanda to do this study. Tim Longman (whom you have met) has been living and working in Rwanda since 1992; Phuong Pham lived in Rwanda for three years and started the School of Public Health; Alice Karakezi of the Centre for Conflict management is a Rwandaise. They are not people who visited the country for a few weeks and became "experts". We are well aware of the issues raised by the concept of "ethnicity" in Rwanda and we also are aware that ethnicity is not something that is immediately recognized - it is usually not clear if someone is Hutu or Tutsi and measures can be taken in training interviewers to assure a "neutral' presentation. It is not at all the same as e.g. a Black-White or Asian-Black issue (and even then, we know that ethnicity is not always apparent). The interviewers did not reveal their identity and were asked to dress and appear non-identifiable; they were trained extensively in this and were supervised closely in the field. The resulting responses were what we expected to see from each group based on our earlier studies. Our statistical analyses confirmed what we expected to see based on the literature and prior studies. The geographic differences were as expected based on the genocide events. Our scales were tested and found to be reliable. In those few sensitive questions where one could express opinions counter to the government's view, a respondent did have the option of not responding.
4. What is important is that qualitative data can give biased information; by choosing geographic regions based on the proximity to the massacres and to the genocide, and then doing random sampling (where selection bias is decreased), we could obtain a population-based assessment that could enhance the validity of what we were finding through our other methods.
Our understanding how countries rebuild after genocide or ethnic cleansing is still in its infancy. It is too early to cut off any promising or tried and true methods. The best we can do is to use multiple strategies and assure that our methods are rigorous and meet the highest scientific standards.
I hope this is helpful,
Best wishes,
Harvey
--
Harvey M. Weinstein, MD, MPH
Associate Director, Human Rights Center
Clinical Professor, School of Public Health
University of California, Berkeley
---
Date: 2/6/2005
From: Helena
To: Harvey
Dear Harvey,
I was very happy to hear from you. I hope you're well!
You may not believe it but i have a note on my current orange Post-it 'to-do' list that says "Write to Harvey Weinstein", underlined. I was a little deterred by the need to retrieve your email address from God knows where. But now, you sent it to me!
You make some excellent points in your letter, and they have caused me to re-think my attitude to Ch.10 quite deeply.
(I do hope that you read all three of the JWN posts where I was writing about your book? ... It was sort of a rolling process of writing about the book.)
So I'd like to say the following:
(1) I'm sorry I said "quasi epidemiological". "Epidemiological" is a fine description of most of the studies in the book.
(2) I understood that you were using multiple, layered approaches for triangulation purposes. (Though I did not see focus group and in-depth interview data brought into Ch. 10, as such.)
(3) I guess I'm still concerned about the use of Tutsi interviewers to work with Hutu (or presumed Hutu) respondents, since I have a strong belief that Rwandan peole are much better at sensing, from a multiplicity of cues, the "ethnicity" (caste?) of other Rwandans than westerners ever are. I'm probably less concerned about the use of Hutu interviewers with Tutsi respondents. Given the Tutsi-Hutu power imbalance in the country, I think the self-preservation factor would be higher for Hutus perceiving themselves interviewed by Tutsi interviewers.
Of course, i recognize the problems of even trying to design an interviewing process that would respect this difference. It might have to involve a two-step interview process, with the first step conducted by a deliberately "mixed", two- or three-person interview team, who would then ask permission for "one of us" to come back and complete the interview later...
In addition, in any authoritarian society, I think there's a real problem for almost anyone "with a clipboard" going around and asking probing questions: respondents might well find it hard to believe this data doesn't go back to the authorities.
That may well be an even greater obstacle to honest data provision in Rwanda even than the ethnic-difference issue?
Having said which, looking at the data in Ch. 10, I would say it's amazing and very courageous of so many Rwandans to have expressed the opinion to your interviewers that "The Arusha Tribunal should try members of the RPF who committed war crimes" (42.0% total), or "Crimes committed by the RPA should be included in gacaca." (37.8% total) So maybe the respondents were less cowed by fear of the data getting into governmental hands than I would have expected?
... I have a couple of quick observations about Table 10.4. First, a big point: I find it notable how high the proportion of "uncertain" or "not informed" responses is, for all these questions about the ICTR. Second, a small point: the very first cohort of three numbers in the top lefthand corner ("Strongly agree" on qun 1) don't seem to make sense. If the proportion of Hutus who said that was 2.4% and the proportion of Tutsis who did was 1.7%, then surely the proportion of the whole respondent-set who said that should lie between the two figure-- unless we assume that the 14% of respondents who declined to state their "ethnicity" who answered "strongly agree" was wildly outside that range and thus managed to drag the total number upward?
More generally, though, I want to say that I think your grassroots, investigative approach to all these questions is extremely valuable indeed. Indeed, I started to go back to an even more basic set of questions when I was interviewing people for my project. (Sadly, I didn't do this in Rwanda. But I did it in Mozambique and South Africa.) My goal was to try to clarify what social goals people in societies emerging from recent atrocious violence considered most important, with the idea of then building up on that foundation to try to see what policy or other social tools they would recommend using in such situations. (I wrote more about this approach in this short paper.) I don't think we should assume, for example, that when people in other societies say they want "justice", they necessarily mean by that the same thing that a western criminal-justice specialist means by "justice"...
In addition, if I'd had more time, I would have spent more time investigating people's own views of what conflict-transcending approaches and mechanisms have worked in their own lives or their own societies in the past, and asked them whether these mechanisms might indeed be useful in a post-atrocity situation. In Mozambique, indigenous cultural resources were the single main thing that helped the society retain resilience and then heal after 17 years of truly atrocious civil war... There's no need at all for a society with largely intact cultural resources like theirs to have to import poorly tested mechanisms from the west; but too often UN bureaucrats and major western aid donors don't even think to ask what cultural capabilities people in low-income countries have... all they ask about is "needs, needs, needs"-- and "needs", moreover that a western-based institution can then offer to meet for them (on a contract basis)...
Oh well, I probably don't need to tell you all that.
Anyway, thanks so much for writing to me. I was wondering how you'd feel if I put our correspondence up on Just World News? Actually, I'm thinking of starting a separate section of the blog that would be just for a discussion of transitional justice issues... When I have time...
Warm good wishes--
Helena.
------
Date: 2/8/2005 12:47:58 PM Eastern Standard Time
From: Harvey
To: Helena
Hi Helena
Thanks for the quick response and I have read all three pieces. They are thoughtful, important, and worthy of much prolonged discussion! We just spent a week with Profesor Byanafashe, a very esteemed Rwandan historian trying to sort out historically the roots of difference that came to the fore during the genocide.
I am delighted to see how you are pursuing your ideas.
It is fine with me if you post my initial response to your piece.
Let's keep in touch
Best, Harvey
Comments