Afterword: Political Ethnography as Art and Science

Adam Ashforth has written one of the recent political ethnographies I most admire. His Witchcraft, Violence, and Democracy in South Africa draws on a total of about three years’ residence during the 1990s in Soweto (South West Township), an Apartheid-built black suburb of Johannesburg, plus subsequent visits to his adopted family and friends there. Earlier, Ashforth wrote an impressive historical analysis of the process by which Apartheid took shape (Ashforth, 1990). But preparation for his book on witchcraft, violence, and democracy plunged him shoulder-deep into ethnography. Through first hand observation, personal intervention, and incessant interrogation of his acquaintances, Ashforth built up a powerful picture of coping, strife, and hope amid vicious violence. Ashforth's ethnographic involvement forced him to abandon many a preconceived category and explanation of struggle during and after Apartheid.

Ashforth's ethnography yielded remarkable, even disturbing, results. His analysis persuades me, at least, of two surprising conclusions I long resisted when hearing them from Adam: first, that no one can make sense of local South African politics without understanding the enormous part played by fears about, accusations of, and reactions to witchcraft in Soweto's (and, by extension, South Africa’s) everyday politics; second, that no one can hope to deal with South Africa's devastating AIDS epidemic or build local-level democracy without confronting witchcraft directly.

Many a political ethnographer will resonate to Ashforth's reflection:

Fortunately, from my first day in Soweto I was blessed with remarkable friends who guided me through the pleasures and perils of life in the township. They steered me toward what little understanding of their world I can now claim, though they do not always agree with the way I have come to understand this place. I have read widely in the years since I began getting to know Soweto, but the essence of whatever I know about this place I have learned through my friends: how I know it is by being there as a friend. This is both the strength and the weakness of what follows. For what I came to understand—dimly, slowly, over many years of fumbling in the dark—is that their world is my world, and mine theirs, and yet we also live in worlds apart (Ashforth, 2005, pp. x–xi).

In order to do his ethnography, Ashforth had to become at least moderately competent in Zulu, Sotho, Xhosa, and the special brands of English Sowetans speak. It helped that he learned to play the violin Zulu style with other musicians in local drinking places, and that he was ready to defend his adopted brothers and sisters from recurrent threats of attack by weapons and witchcraft. His deep involvement in local life allowed him to reconstruct South African politics at the levels of persons, households, and small groups.

To the extent that politics actually consists not of big structures and prescribed roles but of dynamic, contingent interaction among persons, households, and small groups, political ethnography provides privileged access to its processes, causes, and effects. It makes little difference in this regard whether we take politics in the extremely broad sense of all interactions involving the exercise of power or in the narrower, more manageable sense I prefer: interactions in which at least one government participates as actor, object, and/or influential third party. In either the broad or the narrow sense, political ethnography brings field workers into direct contact with political processes instead of filtering that knowledge through other people's testimony, written records, and artifacts of political interaction.

As the superb reports in this collection indicate, to be sure, “political ethnography” commonly includes a continuum of procedures for collection of evidence, from intrusive to inobtrusive:

• 1. in-depth interviews

• 2. conversation

• 3. participant observation

• 4. passive observation of interaction

• 5. covert observation of interaction

• 6. inobtrusive observation concerning residues and consequences of interaction

With the exception of Matthew Mahler (who relies on a version of item 6: analysis of non-fiction retrospective reports of political involvement), every author in this special issue employs more than one of these approaches to political ethnography. Each approach has distinctive strengths and weaknesses. To avoid turning this brief introduction into a methodological treatise, however, let me concentrate on examining what our authors actually have to teach us about political processes, and about how to study them.

Here is my main point: if you believe (as I do) that how things happen is why they happen, then ethnography has great advantages over most other conventional social scientific methods as a way of getting at cause-effect relations. Most methods depend on correlations and comparative statics, asking whether observed variation corresponds to plausible consequences of one condition or another. Ethnography engages the analyst in looking at social processes as they unfold rather than reasoning chiefly from either the conditions under which they occur or the outcomes that correlate with them.

Effective political ethnography resembles good clinical medicine in connecting art with science. A clinical artist picks up clues about patients and their maladies that even probing mechanical and chemical tests fail to detect, then assembles them into competing causal accounts that suggest alternative therapies. But only systematic knowledge of the body's operation, of the symptoms presented by different diseases, of test results, of the available therapies, and of those therapies’ probable consequences–in short, of the relevant science–allows a diagnostician to move from preliminary observation to treatment. On the whole, political ethnographers stop short of intervening directly to cure the ills they observe. Otherwise, the analogy holds: art involving shrewd observation integrates with systematic use of accumulated knowledge. First-rate political ethnography cannily combines art with science.

Yet the analogy fails us in one crucial respect: the range of questions being asked and answered. Despite the somewhat different orientations of epidemiology and public health, by and large medical clinicians are trying to figure out what caused some individual's pathology, and what will alleviate that pathology. Political ethnographers ask a wider range of questions. In the set of papers at hand, we can group those questions in three rough categories: 1) How does a given cause produce its effects? 2) What explains the variable processes that occur in ostensibly similar situations? 3) How can ethnographers produce valid, credible knowledge of social processes? Seen in this light, the papers cluster as follows:

From cause to effect

• Wendy Wolford: How do people get involved in political mobilization, and what impact does it have on them, especially on their sense of what happened?

• Rosanne Rutten: In patron-client relations, how does shame affect a worker's interactions with employers, and how do activists overcome the inhibitions produced by shame?

• Patricia Steinhoff: How do protest participants and police negotiate limits to acceptable performances?

• Pamela Price: How does the decline of authoritative inequality affect interpersonal interaction patterns, and participants’ interpretations of those interaction patterns?

Variation in processes

• Tammy Smith: How does the presence or absence of interpersonal trust affect the capacity of groups to confront collective problems?

• Kathleen Blee and Ashley Currier: How and why do social movement groups vary in their response to national electoral campaigns, and how does the character of their involvement in such campaigns affect their retrospective assessments?

How to create knowledge

• Elisabeth Wood: What ethical choices confront ethnographers in conflict zones? How should (and do) those choices affect their work?

• Matthew Mahler: What analytic approaches yield adequate accounts of individual experience in political engagement?

I won't spoil readers’ enjoyment of these rich papers by summarizing how the authors answer their questions. But notice two features of the questions. First, all of them overflow the immediate situations in which the authors did their ethnographies; they give the lie instantly to the idea that the chief value of ethnography is to provide more interesting or adequate descriptions of social situations. Second, they all concern processes rather than correlations or comparative statics.

I rest my case.