Ethnography as Commentary

)8,23+6%4,=
%7'311)28%6=
Writing from the Virtual Archive

.SLERRIW*EFMER

(YOI9RMZIVWMX]4VIWWDurham & London 2008

∫ 2008 Duke University Press
All rights reserved.
Printed in the
United States of America
on acid-free paper $
Designed by Jennifer Hill
Typeset in Adobe Jenson Pro
by Keystone Typesetting, Inc.
Library of Congress
Cataloging-in-Publication Data
appear on the last printed
page of this book.
Contents u

vii Preface

1 Introduction
Closing House—A Late Ethnography

21 Chapter One
An Event: Closing the House

39 Chapter Two
A Text: Made, Not Found

55 Chapter Three
Kahenga’s Work

73 Chapter Four
Kahenga’s World
 Contents

91 Chapter Five
Kahenga’s Thought

111 Chapter Six

Endings and Ends

125 Notes

133 Works Cited

vi 137 Index
Preface u

This book neither looks nor sounds like a textbook and

yet, if an excuse were needed for its having been written,
it would be that this book can serve as an example, an il-
lustration of, as well as a model for, writing ethnography
‘‘in the presence of texts.’’ That notion will be explored at
length; in essence it is the idea that the possibility to de-
posit our documents in a virtual archive on the Internet
changes the conditions and constraints of presenting
knowledge: The anthropologist can now write and his
or her readers can read around materials that are no longer
limited to short excerpts and quotations. The only limits
to making and keeping ‘‘ethnographic documents’’ directly
accessible are those of the time and energy it takes to set
up, maintain, and consult virtual archives.
 Preface

The approach I take is retrospective: reflecting on times passed and times

past, calling up memories, and thinking about the role of memory in the
work of ethnography. While such preoccupations may seem to be the
privilege, or affliction, of senior practitioners, this study is not offered as a
memoir. On the contrary, it is an experiment that looks to the future. It is
meant to encourage students who are introduced to anthropology, who
prepare themselves for field research, or face the daunting task of earning
their professional credentials with a dissertation.
Beyond that, I hope to make a contribution to debates about the nature
viii of ethnography in these postcolonial and ‘‘global’’ times. We seem to have
overcome anxieties that brought us close to rejecting ethnography as an
inherently imperial form of inquiry. Ironically, while we were busy worrying,
almost everybody else became enamored of the concept of ethnography and
the practices associated with it. Between critical despondency and naive
enthusiasm there is a space where ethnography can survive and thrive as the
core of anthropology—if the researcher is competent, prepared to work
hard, relentless in questioning the legitimacy of his or her claims to knowl-
edge, and, in my experience, plain lucky.
As an experiment in ethnographic writing, Ethnography as Commentary
could be called an exercise in form. However, the approach I have been
advocating in my work has always emphasized substance, the content of
experience and communicative interaction. I see our task as (re)presenta-
tion and interpretation of historically situated events and practices, based
on objectivations, the ‘‘documents’’ we produce or find. In this case I focus
on a conversation that followed the performance of a ritual and was re-
corded before many readers of this book (I hope) were born. The event took
place in Lubumbashi, a large town in the Democratic Republic of the
Congo, a world with which they are not likely to be familiar. Because the
point of my experiment is to explore an alternative to the monograph with
its encyclopedic rubrics, I keep general information about time and place to
a minimum. This does not mean that I consider such information unim-
portant. But, much like the ethnographic text that will be the object of this
study, abundant knowledge of its social and cultural context is present and
easily accessible on the Internet. For matters that require specific knowledge
 Preface

beyond what the text itself contains, I provide conventional references to

the literature.
By the time one gets around to writing what might be called a ‘‘late
ethnography’’ it becomes impossible properly to acknowledge debts to per-
sons and institutions. Therefore I want to mention only those to whom I
owe most: Kahenga Mukonkwa Michel was my interlocutor in the conver-
sation that is at the core of this book. All attempts to contact him and
involve him in my project proved futile. This makes me sad; it also strength-
ened my resolve to publish this work.
ix
Next I must acknowledge the work of my colleague at the University of
Amsterdam, Vincent de Rooij. As the webmaster, editor, and co-founder of
Language and Popular Culture in Africa he assures the presence of docu-
ments such as the conversation with Kahenga. Our website exists because
of time and energy he puts into it above and beyond his regular obligations.
Field work in Lubumbashi (1972–74) was made possible by support from
the National Endowment for the Humanities and the Rockefeller Founda-
tion program of aid to the National University of Zaire (records of which
are by now probably forgotten or buried in their archives).
Finally, I am grateful for a stay at the Stanford Humanities Center (2005–
6) and for the intellectual conviviality I enjoyed with my fellow Fellows. All
of them participated in my travails as sympathetic and critical listeners to
daily progress reports at lunch time. Their own projects widened my hori-
zon; Arnold Zwicki, to name just one of them, was an inexhaustible source
of information and guidance in linguistic matters. Conversations with Hay-
den White allowed us to renew a friendship that goes back almost as far as
my meeting with Kahenga, and I would like to take this occasion to join the
well-wishers when he celebrates his eightieth birthday.
Xanten, April 2007
Introduction:
Closing House—A Late Ethnography

A Time and a Place: Events and Documents

In the fall of 1974 I was getting ready to leave Lubum-

bashi. I had lived there for two and a half years, a stay in-
terrupted only by a few short trips to the United States
and Europe. This was my second visit to the mining
region of Shaba/Katanga in the southeast of what was
then Zaire and is now again the Democratic Republic
of the Congo (the first one, in 1966–67, had been my ini-
tiatory dissertation fieldwork). I had arrived in the re-
gion in 1972 with a project of research titled ‘‘Language
and Labor among Swahili-Speaking Mine Workers in
Katanga.’’ Fieldwork was carried out first in Kolwezi,
where most of the mines and treatment plants were
located, and later in Lubumbashi, the capital, in two
smaller establishments that made neo-African furniture
and other craft items for sale mostly to expatriates.
 Introduction

Going on the mass of documentation assembled, mainly in the form of

notes and sound recordings made at the workplaces and in the homes of
workers, the project was certainly productive. Yet, as the material piled up,
it dawned on me that my agenda had been conceived too narrowly as a
sociolinguistic and lexical-semantic study. Workers’ language, I began to see,
could not be isolated from speech and communicative practices that had
developed in this urban-industrial region and that, through Swahili as the
linguistic medium, had created a lively, multifaceted popular culture. At
about that time—the period of funded research was coming to an end—my
2 stay mutated into a stint of teaching and administration at the local campus
of the National University, allowing me to bridge a difficult period between
losing one job and finding another in the United States.
Work at the university ruled out sustained and focused research. Yet
some of the contacts and discoveries made during this period would later
become more important than the sociolinguistic project I had started out
with. During free time at night or on weekends I kept in touch with
members of the religious movement I had studied in the sixties, and I could
easily have been drawn once again into their circles of thought and instruc-
tion. But I also took the habit of going out for a drink and chat in one of the
cités, the townships set up for Africans in colonial times. I had become
acquainted with the Mufwankolo troupe of actors and enjoyed their com-
pany. I also met and had conversations with many of the popular genre
painters of Shaba/Katanga. There was so much to be curious about.
During the final months before departure, at a time when it was clear that
I would soon leave the university and the country, two chance contacts led
me to take up, or rather stumble onto, ethnographic inquiries I had not
planned and carried out without the kind of mandate or the obligations that
come with being supported by a research grant. One of them was meeting
the painter and historian Tshibumba Kanda Matulu. One late afternoon I
was driving home from the university when I saw him walking on a street,
carrying several paintings. I stopped to take a look at his work and we began
to talk, first on the roadside, later at our house. Our conversation en-
couraged him to take up a project that had been on his mind for some time:
a painted and narrated history of Zaire that has by now become a monu-
ment to his genius and to the urban popular culture that produced him. It
 Introduction

is fair to say that Tshibumba not only represented history, he made it

(Fabian 1996).
On September 30, one week before the first of four sessions with Tshi-
bumba was recorded on October 6, a similarly intensive exchange had taken
place in our house on Mpolo Avenue with Kahenga Mukonkwa Michel, by
trade a provider of health and protection. It was not the first time we had
met. An acquaintance from the Lubumbashi art scene, the painter Mwenze
Kibwanga, had recommended Kahenga to us when my then-spouse told
him about aches and pains which did not respond to conventional treat-
ment. She became Kahenga’s patient during the weeks that followed while I 3
went about my daily work and faced my daily worries, among them the
break-ins and thefts all expatriates had come to fear. Either the subject had
come up by chance or I brought it up because I wanted to try out a local
solution to my problems with safety; Kahenga offered his services and I
became his client. He performed a ritual that I was allowed to watch, and
when he returned a few days later with some herbal medicine he accepted
my invitation to talk about the ritual and about his work in general. That
conversation was recorded and then filed away among notes and documents
I found interesting but had no concrete plans for.

A Project Emerges:
Ethnography and the Virtual Archive

Almost exactly thirty years went by before I listened to that memorable

conversation again. I transcribed and translated the recording and deposited
the text in a virtual archive where it can now be looked at by everyone who
has access to the Internet (http://www2.fmg.uva.nl/lpca). This book will
be ‘‘based on’’ that text, as we say, a convenient expression that allows us to
make empirical claims—that what we report is supported by a document of
research (data, if you wish)—without having to specify how we get from an
‘‘ethnographic text’’ to a text we then call ‘‘an ethnography.’’ There was a time
when it was easier to get away with claims concerning empirical validity and
simply go about the business of ‘‘writing up’’ at home what we had noted
down and recorded in the field. That time is long gone; when we present
knowledge now we are expected to lay open, account for, the processes by
 Introduction

which observations and experiences become documents on which we may

base the interpretations and insights we call ethnography.
This critical task is not made easier when the interval between the events
that were documented and work on these documents has become long
enough to accommodate a generation. Such a distance makes questioning
the givenness of data even more urgent than it would have been if research
and writing had taken place during the more usual period of a few years.
When so much time has elapsed, problems with assuming that what we
write about is simply there are complicated by a heightened awareness of a
4 then. While this does not mean that short-term and long-term recourse to
data differ in kind—every note, document, recording, every ‘‘fact’’ is then, a
thing of the past—it may well be that an undertaking such as the one I am
embarking on in this book changes the nature of ethnography. What we do
becomes historiography as well as autobiography, by necessity rather than as
a stylistic choice. There is no alternative to telling a story, or stories, when
we go about re-presenting knowledge gained and recorded in the past; it is
impossible to make that past present without recourse to the ethnogra-
pher’s personal memory or memories.
To realize this predicament of a ‘‘late ethnography’’ and to spell out some
of the epistemological conditions of such an enterprise is one thing; to
conclude that all the ethnographer can do is tell self-centered stories (be-
cause this is all that is left for postcolonial, postmodern, or post-theoretical
anthropology) is another. In ways that are not quite clear to us, a decidedly
modern technical development, the Internet-connected personal computer,
may invigorate our commitment to ethnography. Ethnographic texts can
now be consigned to a virtual archive where they are accessible to the writer
of ethnographies as well as to his or her readers because, beyond being
accessible (which physical archives may also be), texts in a virtual archive
take on a kind of presence that even the most skilled ethnographer could
not achieve as long as possibilities to refer to, cite, or display documents
were severely limited by the conventions of publishing prior to the Internet
(or to CDs that could be distributed with the printed book).
I have begun to think about the consequences that virtual presence of
texts may have for our writing and will take this up later when I discuss the
question of genre; here I limit myself to a point that is relevant to this
 Introduction

introduction: We have become used to opposing virtual to real and tend to

forget that ‘‘virtual’’ originally connotes effectiveness, strength rather than
weakness. Consigning documents to a virtual archive makes them more
real, not in any ontological sense but in terms of their ‘‘practicality,’’ that
is, as regards their potential to mediate between the events that were noted
or recorded and our efforts to represent the knowledge gained as adequately
as possible. This—a heightened presence calling for acknowledgment and
response—has been what made me embark on this book. To make it clear,
my point of departure has not been to start with a subject matter, a topic,
much less a theoretical question and then seek evidence or support in a 5
document. I began with the transcription and translation of a recorded
conversation and, repeating somehow what happened when I recorded the
conversation with Kahenga, I had at first no other plans than to deposit the
texts in Archives of Popular Swahili. Only when I had almost completed
the task of producing the texts—for reasons that will be discussed later on,
this took several months of hard work—the idea of a book emerged.

Crossing Borders: Ethnography as Transgression

Recalling, as I do now, the story of the project as a series of almost inad-

vertent steps, I realize, suggests a procedure that may appear to violate prin-
ciples of scientific inquiry (or the image we have of it). But then, such
transgressive behavior echoes the crossing of borders that constituted the
object of this study. Not counting linguistic and social barriers that had to
be crossed earlier, at least three kinds of transgression led up to this project:
The first one occurred when the ethnographer became a patient and client,
something that in my mind (and in my memory) was not the same as mov-
ing from observation to ‘‘participant observation.’’ There was no ‘‘method’’
involved, no attempt to change roles for ethnographic purposes. Crossing
that border implied, second, a transgression of sorts when we left the
confines of Western biomedicine to seek treatment by an herbalist and,
third, it meant stepping outside the boundaries of rational conduct when I
became the client of a practitioner of magic or, perhaps, sorcery, to use labels
that are as handy as they are inadequate.
Of course, there would be no document of what happened nor a story to
 Introduction

tell if these acts of transgression had been completely unselfconscious and

free of the duplicity without which ethnographic research would be impos-
sible—a duplicity that makes us cross borders but not without establishing
a record that lets us return to our professional roles and habits. When I
speak of the ethnographer’s duplicity, this is not to be taken as confessional
breast-beating. I want duplicity understood in a sense that Nietzsche, in his
essay on truth and lying, called aussermoralisch—neither immoral nor amoral
but extramoral (1976). Put positively, duplicity is involved in performing,
playing roles, which we do when we work ‘‘as’’ ethnographers; in that sense it
6 is an epistemological concept helping us to understand how knowledge is
produced. Not being forthright at all times because, without a certain
shiftiness, a stranger in a society might not be able to move at all; putting on
an act, say, by trying to communicate and interact above one’s linguistic and
cultural means and competence; assuming multiple roles and guises as
serious researcher, disinterested observer or hanger-on, affable conversa-
tionalist, convivial companion or friend—to the moralist all this must have
something unsavory or outright repulsive. Yet, I would contend, ethnogra-
phy, much like living our ordinary lives, could not be done in any other way.
So, how straightforward was I when I asked Kahenga to ‘‘close’’ our house
in Lubumbashi? As best as I can remember I wanted to have protection.
There certainly was no intention on my part to trick Kahenga into giving
me an ‘‘ethnographic’’ performance (see also Fabian 1990b). When the day
of the house-closing came I did not really know what to expect and had
made no preparations for documenting the event. I was curious, of course,
and professional habit, or compulsion, made me ask Kahenga whether I
could watch him at work. Without hesitating for a moment he agreed and
the ritual took its course. What I saw was fascinating. When it was over and
Kahenga had left I was unable to resist the temptation to take a few notes
on what I had observed—and that, at the latest, was the moment when I
crossed the border between client and ethnographer. Or did that happen
when I asked him whether we could meet again and talk about the things I
had seen and heard? I don’t remember; all I can recall is that I wanted to
record our conversation.
In fact, as I am writing this I realize just how precarious memories can be
because the text, the protocol of our exchange, will show that borders were
 Introduction

crossed even earlier during the days when Kahenga did his work as a healer.
He had allowed his patient to accompany him on a search for herbal
medicines and to take photographs, and note down the names, of specimens
he collected. A botanist at the university examined the photographs and
came up with a list of tentative identifications. Specimens and photographs
were before us when we talked. They played an important part in structur-
ing our conversation and the text that is before us now.
Perhaps I should leave it at that and get on with formulating thoughts,
insights, and findings which, as I announced earlier, will be based on that
text, except that I anticipate a question I should answer, at least briefly, right 7
away: What does ‘‘based on’’ mean? In a general way we use this phrase to
indicate (show or claim) that the knowledge we present is grounded on
evidence. But what does it mean specifically when the evidence is a text? Or
worse, from the point of view of the ‘‘empirically’’ minded, when there is
seemingly ‘‘nothing but’’ a text—no statistical data generated by surveys or
standardized questionnaires, no compilation of facts found in archives or in
‘‘the literature’’? Does that not make my project vulnerable to accusations of
‘‘textual fundamentalism’’? This introduction, a circuitous story of how the
text was produced, should have shown the contrary. Our document is not a
foundation; it is not a ground on which conclusions may rest, if only for the
simple reason that the text does not rest or just sit there. Its current state of
fixation, I hope to have shown, was but a phase in a series of events. The
text is not a depository of facts but a mediator. Its presence makes it a pièce
de résistance on the road from past experience to future representation.
At this point, the reader hopefully has an idea of the long prehistory
of this book; he or she may still wonder what exactly it is about and how
I plan to present material and findings. As is customary I should now
provide a concise statement of the subject matter, followed by a sketch of
the way in which the presentation will be structured in the chapters that
follow. Much as I want to respect custom, I find myself compelled to
commit yet another transgression by leaving this task in suspense, for
reasons I must discuss now.
 Introduction

Ethnography and Form: The Question of Genre

Somewhere along the path I have described so far I decided to conduct this
project as an experiment in breaking up conventions of ethnography that
enjoin us to be clear about what we want to do before we start and to do one
thing at a time. The working title of the project was Closing House: A Late
Ethnography. It gave an indication of what I had in mind: putting an event,
the ritual performed by Kahenga, at the center of attention. At the same
time the pun on ‘‘closing shop’’ was intended: I undertake this project at a
8 time when I begin to look for some closure in my professional work.
Circumstances may see to it that this book will be my last and it is difficult
to accept this without at least trying to take stock and coming to some kind
of conclusion about matters that have occupied me in my work. How can I
do this without making the ethnography of a rite of protection a mere
pretext for an intellectual memoir? Either one may be of interest but can
they be fitted into a coherent form?
Anthropology and readers of anthropological writing have by now be-
come used to ethnography with a heavy dose of autobiography. That is not
really my problem; I never thought it desirable or even possible to keep the
author out of accounts based on research requiring his or her active pres-
ence. What preoccupies me is to find out what it means to write, as
indicated in the provisional subtitle, a ‘‘late ethnography,’’ one that is be-
lated, more than thirty years after the event, late in the ethnographer’s life,
and above all late in an historical constellation—something we are trying to
grasp with our theorizing about the postmodern, postcolonial condition. Is
it not, to put this bluntly, simply too late to be writing ethnography?
True, we all but abandoned the once canonical form of the monograph
when we realized how compromised it was by its roots in the imperial gaze
of a science of mankind conceived as a branch of natural history. For a while,
this made us agonize about, or, as some would prefer to put it, experiment
with, genres such as personal or historical narrative, dialogue, poetry, and
essay. Yet we still designate as ‘‘ethnographies’’ most dissertations and pub-
lications that are not textbooks or purely theoretical treatises. Whatever the
outcome of these developments will be, it is difficult to imagine that, given
 Introduction

its growing popularity outside anthropology, the term ‘‘ethnography’’ will be

abandoned in the near future (Fabian and de Rooij, in press). In the
meantime, any contribution to anthropology will also have to be a contribu-
tion to the debate about genre, that is, about the quest for legitimate forms
of presenting knowledge of cultures and societies.
Such a conclusion, I think, is inevitable. We are bound to worry about
genre, no matter how tired we may be of literary introspection and how
much we sympathize with the colleague who exclaimed ‘‘genre be damned’’
(Webster 1986; I think this was before the phrase became a fad). On the
other hand, the irreversible turn of attention to literary form is no sub- 9
stitute for continuing to worry about the production, not just the presenta-
tion, of knowledge. In fact, I would argue that the only excuse for leading
anthropology up the path of literary theory is epistemological; attention to
genre ultimately means attending to generation, that is, to the making of
knowledge.

Commentary as a Genre

What could be the genre commensurate to the task I set myself in this
book? I must admit that it has been tempting to present what I have to say
as a collage. Parts written in genres such as dialogue, personal and historical
narrative, and interpretive essay could be assembled to form an evocative
and coherent picture. In the end I could not envisage such a project. As I
understand it, collage is more than juxtaposition of elements. At any rate,
when applied to writing, collage is a metaphor and it is successful, rather
than confusing, only as a poetic creation. Anthropology has a few masters of
poetic collage; I am not one of them. And as to collage in its more literal
sense, I am not about to embark on a pictorial mode of representation
(ethnography as exhibit or tableau vivant) after all the criticism I have
heaped on ‘‘visualism’’ in anthropological discourse (Fabian 2001a).
So, for better or worse, the question of genre must be faced and this book
will be an occasion further to develop a position I began to formulate some
years ago (Fabian 2002b): ‘‘Commentary,’’ I predicted, is likely to emerge as
a genre of ethnography. The point of departure of the argument, introduced
 Introduction

above, was the new kind of presence ethnographic texts take when they are
deposited in publicly (or at least widely) accessible virtual archives. As
briefly as possible I should now spell out what this view entails.
First, to qualify as a genre, commentary must be more than just a gloss on
a brief excerpt from a source, or an annotation to, say, a diagram or an
illustration. As the form of a piece of ethnography, be it an article or a book,
commentary requires the co-presence of a substantial text and the interpre-
tive, analytic, or historical writing based on that text.
Second, commentary as a genre not only determines literary form, it
10 defines a practice of writing. Commentary is made ‘‘practically’’ possible by
the virtual presence of text(s) and it can be realized, practiced, within a
community of writers and readers who have access to the Internet. While
this may be considered limiting (despite all the talk about globality, access
to the Internet remains limited), disadvantages are outweighed by the ad-
vantages of freeing ethnographic writing from many constraints inherent in
print publication, foremost among them the injunction to keep presenta-
tions of ethnographic text (especially in the original language) to a mini-
mum of quotes.
Third, these practical advantages are by no means only practical. They
also have theoretical significance: the possibility of a form of ethnography
that is not predicated on the absence of its object or, to be more precise, on
the object being consumed by devices of presentation such as tabulating
quantifications, drawing graphics, figures, and diagrams, or simply by a
prose that, being either strictly descriptive or predominantly expository,
manages to withhold from (some may say spare) the reader the events and
documents on which it must nevertheless ground its authority. When I
propose commentary as an alternative, this should not be misunderstood as
just another version of recent calls (often long on ethics and short on
epistemology) for ‘‘giving a voice’’ to our sources. Nor would I want to
advocate burdening our readers with a kind of methodologism that pre-
empts critical theoretical reflection by endless accounts of procedure. After
all, our task is producing knowledge, and that involves more than either
showing how it is done or just writing down what we know.
 Introduction

Commentary, Comments, and Memoranda

Now that the case is made for commentary as a genre of text-centered

ethnography, that is, as a form that informs a piece of writing in its entirety,
the question remains of how the general idea is to be realized specifically.
A condition of writing in the mode of commentary is, to repeat this, the
presence of an ethnographic text. The awkward attribute ‘‘ethnographic’’ is
meant to remind ourselves as well as our readers of the peculiar status of
texts such as the one that will occupy us in this book. In a narrow under-
standing of the term, ethnographic texts are not literature. They are neither 11
found nor written as fiction; they don’t come from, and are not meant to
contribute to, a canon of readings. Nor are they, again in a narrow under-
standing, documents of the kind historians find in archives which they
themselves did not set up. Ethnographers, one might say, can deal with their
texts without being weighed down by a canon or an archive. Just as well,
because the demands of the ethnographic text are heavy enough.
To begin with, the presence of a text puts us in a paradoxical situation. As
the protocol of a performance or of a communicative exchange, a text’s
presence signals the absence of the event it documents; the text may be
present, the event is past. Commentary, therefore, is writing in the face of
that tension. One may say this is the case with all narrative and even with
descriptive writing (as writing, narrative and description are always ‘‘after
the fact’’) but tension is exacerbated, instead of relieved, by the seemingly
nonproblematic presence (givenness, availability) of a text. Hence it would
be wrong to think of commentary as a relatively relaxed genre, so to speak,
free of the constraints imposed by conventional monographic rubrics that
must be filled or analytical schemes that must be completed.
Thinking about this last statement set me off on a metaphorical reverie.
Let us say that commentary is made up of many comments and that
comments relate to commentary like bricks to a house; you build up a
commentary with comments. This sounds plausible but there is something
fundamentally wrong with the image (fundamentally: more than what is
wrong with every image). Brick construction not only requires elements
that are fabricated, made, not found, but also that these elements be identi-
cal (if not absolutely then at least ‘‘for practical purposes,’’ the ones a builder
 Introduction

has). The first requirement is met by ‘‘comments’’; they must be made,

formulated. The second one, if fulfilled, would mean that every comment
would have to come from the same mold and that would mean that writing
commentary consists of filling preexistent rubrics and the result would not
be an alternative to, but a kind of, monograph.
If the brick doesn’t work as an image, what about the stone and stone-
construction? A stone is found (or quarried), not made; it gets its shape from
being hewn rather than molded and fired. As to the identity of elements,
stone masonry has requirements similar to those of brick construction and
12 would be useless as a metaphor, unless we consider building with field-
stones, in which case the elements may be quite dissimilar (big or small,
rough or smooth, almost round or almost square, and so forth). That would
come much closer to the idea I have of building commentary with com-
ments if it were not for an inconvenient, unacceptable implication: Field-
stones are found in the field. Ethnographic comments, tempting as it may be
to latch onto the ‘‘field’’ (as in ‘‘fieldwork’’) as another metaphor, are definitely
not just found.
So, short of declaring this reflection an idle exercise, what can we take
away from it? Thinking of writing commentary as building it up from, or
out of, comments may be misleading unless we take metaphorization one
step further and postulate that when we write comments we make stones
the are shaped like fieldstones. In plain words, we should not feel compelled
to write comments that fit a mold or come out as standardized elements; we
should pick them up in whatever shape they come and put them together in
a structure that holds up.
What kind of building will be constructed in such a manner? If commen-
tary is realized by writing comments on a text and if we keep in mind what
was said about presence and absence, present and past, it follows that such
comments may be thought of as memoranda: observations, statements, ex-
plications, references written down as reminders, anything between what
the text reminds us of and what we think should be remembered when we
read a text. Remembering and memories are crucial to ethnography, as I
have argued with regard to recognition of alterity as anthropology’s central
theme (Fabian 1999), in a reflection on remembering in ethnographic prac-
tice (Fabian 2007: chap. 10), and will show in detail later on when I confront
 Introduction

the conversation with Kahenga. If memory is as important as I think it is in

this undertaking, would ‘‘memoir’’ not be a label more appropriate than
‘‘commentary,’’ especially in a project that includes elements of autobiogra-
phy? There are reasons to reject that option. First, this account will not live
up to the expectations raised by calling it a memoir because it will be
narrative only in parts and, at any rate, it does not aim to tell a single,
coherent story. Second, in another (French) connotation, memoir may
designate a thesis or dissertation and that is not what I envisage either.
For better or worse, I made a commitment to commentary as the genre of
this book—after trying it out in two earlier, shorter pieces of writing (Fabian 13
2003 and http://www2.fmg.uva.nl/lpca/jlpca/vol1/fabian.html)—only to
run into problems that began with deciding on a table of contents. What
should be the major divisions, what their order or sequence? I toyed with
three or four outlines and began to have serious doubts until I realized that
my troubles were caused by a failure to distinguish between commentary as
a practice of writing and commentary as a form of representation. I was
reminded of the debate concerning dialogue as ethnographic practice and
dialogue as a literary form. From the fact that much of our research consists
of dialogues with our ‘‘informants’’ it does not follow that dialogue is the
most, let alone the only, appropriate genre of ethnographic writing. Simi-
larly, adopting commentary as a practice of dealing with texts does not
mean that the presentation of knowledge gained must conform to a genre
that exists in our literary canon. All that is required is that the final form of
writing a commentary must be such that it can be traced back to commen-
tary as a practice. Even monographs that followed rules of data collection
and respected rubrics of presentation that had become established in our
discipline came in many different kinds; this will be the case, even more so,
with commentaries. Their tasks may ultimately be determined by anthro-
pology’s current agenda but they must meet the exigencies of specific texts.

A Text Never Comes Alone: Con-texts

If relying for evidence almost exclusively on texts makes the ethnographer

suspect of textual fundamentalism (see above), working with just one text
should be even more problematic. It is time, therefore, to spell out what text-
 Introduction

centered approaches assume about texts and, perhaps more importantly,

what they don’t assume. From the ‘‘empirical overkill’’ of which Franz Boas,
one of the founding fathers of anthropology, was accused because of his
massive text collecting, through the making of vast bodies of texts (‘‘myths’’)
the grist for structuralist analytical mills, to text(s) as a key metaphor for
culture (the latter accompanied by the disappearance of ethnographic texts
from writing in the ‘‘interpretive’’ mode), anthropologists have tried out
almost every conceivable approach without being able to establish theo-
retical consensus or at least widely shared habits and conventions in deal-
14 ing with texts. Such a state of anarchy may be deplored; for me it has
made working with ethnographic documents an undiminished pleasure. Of
course, it also forces me to rethink theory and method every time I approach
a text. Specific issues will be addressed as they come up later on; here I want
to make only a few general statements regarding my present views.
Text-centered ethnography has been for me the practical consequence of
taking a language-centered approach, which I adopted as the result of a
conception of ethnographic inquiry as a (predominantly) communicative
undertaking. In much of cultural anthropology knowledge production is
interactive. As language and speaking mediate between the ethnographer
and the people he studies, texts mediate between communicative events
(experience) and the representation of knowledge.
In other words, when ethnography is called text-centered this regards
above all its epistemological foundation; it does not mean that collecting
texts is the principal purpose of research. A text, to become ‘‘collectible,’’
must be relevant to a project that is always wider than what a specific text
may document. This also applies to the document selected for interpreta-
tion in this book. It is not ‘‘unique,’’ no matter how singular its history may
appear. Perhaps not when it took place but certainly afterward, the conver-
sation with Kahenga became part of a corpus of texts, a term that may have a
technical meaning in literary theory but never loses completely its meta-
phorical connotations. Like a body, a corpus has size, volume, weight,
articulation of parts and members; as long as it is alive it grows and changes.
The latter, growth and change, certainly fit the corpus of documents I
produced (more often than found) in the course of my work as an ethnogra-
 Introduction

pher. What in the beginning (during dissertation research in the sixties)

may have looked like data needed to carry out the proposed study of a
religious movement; what I went after when I later worked on language and
labor in the seventies; and what I came upon during shorter visits in the
eighties—all this turned out to belong to a ‘‘body,’’ something I eventually
grasped theoretically with the help of a concept of African popular culture
(Fabian 1998). That culture—minimally understood as contemporary prac-
tices of survival—was the context in which Kahenga worked, and this makes
the document of our exchange a co-text with many others.
If what I said about the new kind of presence texts acquire on the Internet 15
is true, then the virtual archive should also influence our ways of dealing
with a corpus of texts, and this should in turn affect the writing of commen-
tary. For instance, specific issues (semantic, topical) that come up in the
commentary on one text can now easily and quickly be searched for in other
documents placed in the same archive (in Archives of Popular Swahili some
are already cross-referenced; many more links could be established). Con-
versely, given their connections to one and the same easily accessible corpus,
the text-centered ethnographies on popular religion, theater, and histo-
riography which preceded the project introduced here will hopefully gain
coherence and grow into a body of writing that has a chance of staying alive
when current topical and theoretical interests fade.
What exactly will the commentary on my conversation with Kahenga
recorded thirty years ago add to the ethnography of popular culture in
Zaire/Congo? An easy answer would be to say that it opens the domain
of popular medicine. My dealings with Kahenga might then become a case
study in medical anthropology. I have no objection to such a reading as
long as it is understood that an interest in medical anthropology was not
what brought about the events and their records that will occupy us in this
book. At any rate, it would be somewhat anachronistic to project back into
the early seventies the image of a subdiscipline that thrives today but was
then, as ‘‘ethnomedicine,’’ at its beginnings. These qualifications are im-
portant because they let me comment on a conversation about medicine
without an obligation to relate my findings to state-of-the-art medical
anthropology.
 Introduction

Appropriation/Expropriation:
The Ethics of Commentary

I have always thought that to practice anthropology reflexively and critically

is all that is needed to make our work legitimate. We must be alert to the
quandaries we face as members of institutions and as citizens or residents of
countries that pursue political and economic interests, more often than not
to the detriment of the societies we study. As a scientific discipline, anthro-
pology should be governed by rules and habits of ‘‘disciplined inquiry’’ (not
16 to be equated with ‘‘methods’’) but I don’t believe that anthropology should
be subjected to a code of ethics any more than mathematics or philology.
Probably this sounds hopelessly out of tune with developments during the
last thirty years that have made ‘‘ethics’’ a central concern. It means that I
cling to a notion of my field as an (academically supported) intellectual
endeavor rather than a profession whose relations with clients and sponsors
require rules similar to those that are devised to guide the conduct of
lawyers and physicians. I dread the possibility that, being caught between
human-subject regulations and cultural-property exactions, it may become
impossible for us to conduct ethnographic inquiries in a manner and for
purposes—communicative ethnography producing shared knowledge—that
made anthropology a distinctive form of inquiry.
This being said, I want to address, however briefly, questions of legiti-
macy encountered by text-centered ethnography of the kind I will pursue
here. Concerning those that regard legalities of proprietorship and copy-
right to texts deposited in a virtual archive I refer to a statement on our
website. As long as it is not clear what international law will eventually
impose on us, we act on the assumption that documents deposited in a
(virtual) archive are publicly and globally accessible. We also assume that,
with exceptions (we have not made any so far), accessibility means freedom
to quote, copy, or work on the texts.
Another issue I want to dispose of right away is usually referred to as the
‘‘protection of informants,’’ a requirement allegedly met by withholding the
names of our interlocutors or using pseudonyms. In the past, I have dealt
with this by simply not publishing anything that, to the best of my knowl-
edge, could be damaging or dangerous to those with whom I worked. Very
 Introduction

little ever was; a more difficult problem has been to distinguish between
damaging and critical statements. I never subscribed to the precept that
ethnographers should be neutral toward (the representatives of) cultures
they study, if only because I cannot imagine ‘‘neutral’’ communication other
than interaction in a clinical setting, which is not my idea of research. If
there is a rule about withholding names it should be this: In our accounts
ethnographers and interlocutors are both agents; why should the author be
named and others remain anonymous?
Another question that has become inescapable in recent years is: Should
our interlocutors be listed as coauthors of our writings? My view of the 17
matter is that, yes, contributions should be acknowledged and, if possible,
documented but, no, those who worked with us should not be burdened
with the responsibility of authorship in a more narrow sense. Did I consult
Kahenga about these matters? I did not, though I tried to contact him
recently, simply to find out how he is and to inform him of my project. So
far, my attempts have not been successful and I don’t expect this will
change. I fear we must assume, as we do about Tshibumba, that he may no
longer be alive.
Finally, there is ‘‘appropriation,’’ a matter that is often discussed as a
problem of ethics (or politics) although it regards, in my view, above all
questions of theory and epistemology. Does ethnography as a kind of
knowledge, does commentary as a specific form of representing knowledge,
‘‘appropriate’’ its contents? And if so, why did appropriation—making some-
thing one’s own—acquire such an unsavory taste when it is discussed in
anthropology? Allegations of appropriation come easy to those who are
prepared to reduce anthropology to its colonial-imperial history or to the
services it may render to postcolonial interests. While these charges are
transparent and may fit transparent crimes of cultural robbery, they depend
for effect on the conceptual fog from which they emerge.
A measure of clarity can be achieved when we keep in mind that ‘‘prop-
erty,’’ the concept that underlies appropriation, is an equivocal, shifty
beast. In talk about ideas or about culture and its creations, property/
appropriation is capable of changing, within the space of one argument,
from its literal meaning to analogy, to metaphor, or rhetorical trope and
back. Is it at all possible to speak of ethnography/anthropology as a form of
 Introduction

appropriating other cultures in a way that illuminates our work rather than
just denouncing it globally?
Answering this troubling question may have to begin by asking a counter-
question: Can the stuff we ‘‘make our own’’ when we learn about and
understand, say, the imaginary characters of a story, the prescribed actions
of a ritual performance, the movements or rhythms of a dance, and so forth
be considered objects in an ontological sense? Are they things? Are not even
manifest objects—we call them ethnographic or art objects—collected not
just as things but as artifacts, culturally or aesthetically speaking? Between
18 cultural diffusionism throwing hoes, myths, and kinship terms into one and
the same bag of objects for the purpose of mapping the distribution of
things cultural in space, and sociological functionalism-cum-positivism en-
joining us to study social relations like things, comme des choses, as Durkheim
put it, anthropology went through a long history of producing allegedly
certain knowledge about objects whose status remained uncertain. Yet,
without agreement about what can be considered an object (ontologically,
to begin with, but also intellectually and legally), let alone about what kind
of object can become property (and whose property it would then be), talk
about ethnography as appropriation has no distinct referent.
Matters are made worse when, for ideological rather than logical reasons,
appropriation is equated with expropriation. Under such an indictment
ethnographers could not make anything their own without taking it away
from the people they study. Which brings us back to the ‘‘ethics’’ of my
project in this book. I would have had to be a villain or a fool to have
undertaken the labors it took to transcribe, translate, annotate, and com-
ment on our text if I had thought my efforts to convey an understanding of
what happened when Kahenga closed our house left him or his culture
dispossessed. I felt enriched by our conversation and I like to think the
feeling was mutual. Kahenga was under no obligation to talk to me after he
performed, for a fee, his services as a ritual specialist. Perhaps he was just
being polite when he agreed to meet afterward; quite likely he had expecta-
tions similar to mine about discussing the nature of his work. At any rate,
like all my other interlocutors, he was not offered and did not ask for
payment for the conversation we had.
 Introduction

What We Talked About:

An Overview and Guide to the Text

This attempt to trace my project back to its remote beginnings and to place
it in a context of current concerns with the nature and legitimacy of eth-
nography may have taxed the reader’s patience. That Kahenga and I con-
versed; why, when, and where our meeting took place; what remained as a
document and what I intend to do with it—all this had to be brought up, at
least summarily, so as to stake out the arena in which the text will be
confronted. It is now time to take a first glance at the content of our 19
exchange and at the succession of topics we covered.
In the outline that follows, numbers refer to paragraphs of both the
Swahili transcript and the translation as they appear on the web site.
Prelude: Getting started, 1
Patients/clients and diseases/problems: Classification in context, 2–14
Diagnosis and treatment, 15
Etiology: Classification of kinds and causes of illness/problems, 16–23
Biographic data, 24–26
Spirits and spirit associations (1), 27–28
Distinctions of practitioners, terminology in Hemba and Swahili, 29–33
Spirits and spirit associations (2), 34
Kahenga teaching his craft, 35–38
Closing the house: The ritual, 39–41
Interlude: About great spirits and their territory, 42–43
Closing the house (continued), 44–50
God, spirits, ancestors, and Christianity, 51–53
Kahenga, the Christian, 54–57
Back to diseases/problems and herbal medicine, 58–59
Comments and explanations on specimens of plants, 60–69
Looking at a diviner’s calabash from Luba country, 70–71
Herbal and prescription medicine, 72
Popular knowledge of herbs, 73
Coming to an end, 74–76
An Event: Closing the House

Remembering What Happened (1):

The Event in the Ethnographer’s Mind

1 What we said when Kahenga and I met to

talk about the protective ritual he had per-
formed was recorded and is now present as a
text. Of the event itself I have vivid but frag-
u u mentary memories. When I now set out to
tell what happened, my account will be a reconstruction
rather than a simple retelling of an experience made or
a description of a performance observed. I will assemble
my story from three sources: recollections that are pres-
ent in my mind; some notes, including a rough sketch
of the scene, taken down immediately or soon after the
ritual; and a review of the procedure that was part of our
conversation.
To begin with, I cannot give an exact date for the closing
of the house. It took place a few days before September 30,
 An Event

1974 (a Monday), the date noted on the cassettes used to tape the conversa-
tion about the ritual. We had come to Lubumbashi two years earlier and
had moved around from one temporary lodging to another for most of the
first year. Eventually, when my research grant ran out and I found work at
the National University, we settled down in a comfortable house in a
residential neighborhood not far from the campus. Life was not bad in the
late summer and fall of 1974. The oil crisis was yet to hit the country
with its full force. Mobutu’s regime was in its golden years. The mines of
Shaba/Katanga did well and there was work in the region. Though wages
22 were low by international standards (the equivalent of forty dollars a month
was considered a good income), many Lubumbashi urbanites somehow
managed to lead a life that was not limited to bare necessities. People
furnished their houses and decorated the walls of their living rooms with
paintings that reminded them of their colonial past and of urban life in the
present. Beer was affordable, at least for a few days every month, bars and
dancing places were full, popular music bloomed. Far away in America a
president had resigned in disgrace but in Zaire another event caused more
excitement. Everybody was looking forward to the ‘‘rumble in the jungle,’’
the fight between Muhammad Ali and George Foreman that was to take
place in Kinshasa on October 30.
In late September, the dry season comes to an end. The days turn hot and
overcast, fields and gardens are dried out, and food is scarcest in the weeks
before the rain arrives. A mood of nervous anticipation spreads even among
those who do well enough not to depend on the growing season. It is also a
time when residents of comfortable homes get anxious as the young and
enterprising among the poor and hungry take to theft and burglary for
survival. A season for thieves is also a season for those who can offer
protection.
I reported in the introduction how we first came in contact with Kahenga
as a healer. It was the anxiousness that was in the air, rather than a specific
incident—we had no break-ins—that made me call on his services as a
specialist in matters of security. On the appointed day he arrived at the gate
on Avenue Mpolo just after sundown (always at 6 pm, give or take half an
hour, since Lubumbashi is at some distance from the equator). Baba Mar-
cel, our cook, must have opened the gate for Kahenga and I went to meet
 An Event

him halfway. To my surprise, he was accompanied by a young man (a

relative and apprentice, as it turned out later). Either he or his assistant
carried something wrapped in cloth, one or two bundles, which they put
down as we exchanged greetings.
Then came an awkward moment. Here he was; what was going to happen
next? Kahenga took matters in hand by asking Marcel to return to his
quarters in the back of the house. He and his family were to stay inside.
When the cook had gone, my wife and I stood there facing Kahenga, still
not knowing what to do. He quickly bent down, touched the ground, and
then used his thumb to rub some dirt on our foreheads. I am almost sure 23
that he gave an explanation for that gesture right away but it could also have
come later. At any rate, it was the sign that he had begun with his work. He
could not know that what he had just done to me triggered deep memories
of Ash Wednesday, the beginning of Lent, when we went to church to
receive the sign of the cross in ash on our foreheads from a priest muttering
memento quia pulvis es et in pulverem reverteris (remember that you are dust and
that you shall return to dust). Though the two rites had little or nothing in
common as far as content, or intent, was concerned—Kahenga’s gesture, as
it turned out later, was one of protection or purification; the ashen cross is a
reminder of mortality and perhaps a call for penance—for me they merged
as bodily experiences. Kahenga had pulled me back into a realm I had left
behind long ago.
I was awed, and stayed awed throughout. Still, professional curiosity
made me ask Kahenga whether I could watch him doing whatever he was
going to do. He agreed and, without stopping to give explanations from
then on, he began undressing under a tree in the right corner of the lot. He
stripped down to a loincloth (not the kind of undergarment a young,
modern African would wear), started unpacking his bundles near the gate,
and told my wife to go into the house, turn the lights off, and stay inside.
Outside, it was almost completely dark by now. Observing what happened
then was difficult, also because the ritual was all movement, not a tableau
holding still.
Trying to sort out my memories I come up with three sets of action.
First Kahenga dug, or scraped out, moving counterclockwise, eight shallow
holes, four in the corners of our (roughly) square lot and four in the middle
 An Event

of each side between the corners. Then he went around and placed sub-
stances he had brought along in each hole. Finally he made a third round,
now crouching over each hole and covering it with dirt which he seemed to
move with his buttocks. While doing this, or right after closing a hole, he
chanted in kiHemba, his native language. I could not understand these
incantations except that I thought I recognized litanies of proper names.
When I now make Kahenga’s actions, gestures, and calls fit some sort of
classification by naming them, I realize how far removed this is from what I
experienced following and watching the almost naked man as he moved
24 through the night. ‘‘Closing the house’’ was a ritual that obviously followed a
script and rules but there was nothing rigid about it. In fact, the perfor-
mance had at one point a slight flaw that enhanced its force and made it
clear that I had crossed the line between observer and participant. I had
watched Kahenga digging eight holes and it did not take much to figure out
the logic of their spatial arrangement which made the sequence of some
actions predictable. Thus, when Kahenga, after covering the seventh hole
(not one of the cardinal points), gave signs that the procedure was complete,
I intervened. How about the eighth hole? He took this in stride and went
on to finish the course. Had he not done this, he told me, it would not have
mattered a bit as far as the efficacy of the closing was concerned. Aside from
this brief exchange, the only other time we talked briefly was when we came
to the sixth hole (in the lower-right corner of the lot). He asked me whether
this was a dangerous point. That is what a short note on the sketch I drew
right after the ritual says but I don’t remember what prompted the question
nor what my answer was.
As I write this I discover another gap in my memory—where were our
dog and cat during the ritual? They must have stayed out of sight, either
inside the house or somewhere in the yard. Would Kahenga have wanted
them out of the way? The reason it occurs to me to note this now may be
interference from another layer of recollections from the recent past, those
of transcribing and translating our conversation, during which we talked of
beliefs about cats in relation to sorcery.
Was there someone watching what went on in our lot? I never gave this a
thought at the time. Today I would say whether or not Kahenga had
 An Event

spectators (other than me) was of no importance but this is an issue that
will have to be addressed later.

Remembering What Happened (2): The Event

Remembered by the Ethnographer and His Interlocutor

Reflecting the centrality of the ritual, the portion of our exchange during
which we discussed the closing of the house occupies the middle of the text
(see the outline provided in the introduction, paragraphs 39–41 and 44–50).
We had talked for almost an hour and had covered much ground when a 25
moment came to decide on a new subject. I was about to begin questioning
Kahenga about samples of medicinal plants but changed my mind. From
some notes I had before me I pulled out the sketch I had drawn after the
ritual and suggested we first discuss the ‘‘closing of the lot.’’ I pointed out the
enclosure and the location of the house. Then, as is the habit of ethnogra-
phers, I first asked whether there was a name for what he had done (39).
The response was not what I expected as a description of the ritual: dawa ya
kufunga nyumba, literally: medicine to close a house, not a single term but a
phrase in which a noun, dawa, a substance not an action, is foregrounded.
The expression he chose put the focus on what he considered essential in all
of his work, namely, the use of substances that have the power to influence
states or the course of events. In other words, the domain he named in his
first response was not ritual but ‘‘medicine.’’
When I probed further by asking Kahenga for a term in Hemba he again
responded with a phrase, kuzika nzibo, which I translated as closing a house.
Actually, it first came out as nakuzika nzibo. If, as I suspect, there was
interference from Swahili in the verb form, then the literal translation could
be ‘‘I close the house for you.’’ In my translation I took the word I heard as
nzibo (or zibo) to mean house in Hemba. I now think that Kahenga mixed
Hemba and Swahili. In Swahili, (m)zibo means a stopper or plug as well as
the action of placing a stopper, closing a hole or passage. So the expression
Kahenga offered would be a verb phrase with a (semantically) cognate
object (like ‘‘to sing a song’’). In my Luba dictionary, the verb -zika means
‘‘barrer’’ but also ‘‘protéger contre qqn. en empêchant le passage ‘‘ (protect
 An Event

against someone by impeding passage; Van Avermaet and Mbuya 1954: 818).
In the same entry I also found ‘‘[zika] bafu: empêcher les morts de nuire aux
vivants. Quand un nganga soigne un malade il place des objets (magiques)
sur tous les sentiers qui mènent à la hutte du malade pour empêcher les
morts d’approcher’’ (keep the dead from doing harm to the living. When a
nganga treats a sick person he deposits (magic) object on all the paths that
lead to the patient’s hut to keep the dead from getting close).
With that oblique remark about a healer (nganga) the dictionary gives us a
first hint of an issue that will have to be addressed eventually: The context
26 in which Kahenga performed the protective ritual was larger than I had
suspected. In the thinking that seemed to guide him, connections might
exist between burglars and the spirits of the dead; both could be a threat,
not only to security but also to health and well-being.
Then I brought up the eight holes (40). The significance of their place-
ment must have been on my mind at the time; it certainly is now. Was
Kahenga drawing on some spatial, perhaps cosmological, symbolism in
order to mark the terrain for the ritual operation that was to follow? The
intuition that both the number of holes and their location had a special
meaning must have made me ask him whether this was ‘‘always’’ done. He
confirmed that it was but the explanation he gave was offhand, very short,
and difficult to understand on the recording. As best as I can tell, it
amounted to saying that the procedure he followed was just a practical,
geometrical ‘‘method’’ of placing the holes in the ground where he would
deposit the things (objects or substances?) he had brought along. They were
not bizimba, magic charms, he assured me, just herbs that would stop
potential troublemakers. He then told me how he had prepared the herbs
and that he had done this before he came to the house to save time—
another prosaic remark that left questions I may have had about the secret
nature of such preparations unanswered.
When I brought up his stripping down to a loincloth this was met with a
similarly laconic response. No explanation was offered for this gesture nor
for the necklace I had seen him put on before he started to say, or chant,
something that sounded to me like prayers (41). Kahenga agreed when I
asked him whether this chanting was what I thought it was. I went on
posing questions that were, again, aimed at ascertaining the ritual character
 An Event

of his incantations, another expectation I had brought along with my

anthropological baggage. With a casual finesse I had come to know from
many among my interlocutors, he put things right: When we pray to (the
spirits) of our ancestors we don’t use fixed formulae (‘‘like songs?’’ I had
asked), we just speak to the occasion, ‘‘just talking,’’ as he put it. It appears
that I was undaunted and asked him whether he could recite such a prayer
for me in his native language. Kahenga obliged and composed a kind of
generic prayer on the spot, beginning with a sentence that announced why
he addressed the spirits: ‘‘I close the house for this white man.’’ He now
used kuzibia instead of the expected kufunga. As explained above, the image 27
is one of patching up holes or leaks rather than of closing or locking
something. This weakens the literal sense of ‘‘closing’’ and leaves open the
possibility that the ritual may be aimed more at protecting integrity than
excluding, locking out, threats. Such an interpretation would also fit the
prayer asking to give ‘‘strength’’ to the inhabitants rather than to the house
or its enclosure.
To my request for an example of the prayer in Hemba (the language he
had used during the ritual) Kahenga responded in Swahili; multilinguals
often ‘‘forget’’ which language they are speaking at a given moment. When
I brought this to his attention he obliged with a prayer in his language.
Listening to it must have reminded me of an observation I had made when
I watched the ritual. I had recognized proper names in his incantations
that made them sound at times like litanies. Kahenga confirmed my obser-
vation and named four addressees of his prayers: Mukenge Mbuyi, his
father; Nyange, whom he called ‘‘mother’’; Kayembe, a chief; and Yagamino,
a ‘‘great spirit.’’
There is something odd about this list. The persons named were spirits
except, it seems, Nyange who gave him his medicine, as he said, and whom
he called his grandmother and teacher. But by all indications his grand-
mother was alive when we talked; why pray to her? The solution may have
been alluded to elsewhere in our conversation when Kahenga gave her
full name as Nyange ya Kahenga. This made me ask whether she got her
name (Nyange) from a grandparent. She did and this could mean that the
Nyange mentioned in the prayer was (the spirit of) his grandmother’s
ancestor.
 An Event

My questioning Kahenga about names led to a lengthy discussion of

spirits (42–43), a diversion from reviewing the ritual and a topic to be taken
up later. We got back to the closing of the house with a question I had
forgotten to ask earlier: What about the preparations he had taken before
starting with the ritual (44)? Kahenga first spoke of the herbal medicines he
had brought along. We had talked about that earlier and he must have
sensed that I had something else on my mind because, rather abruptly, he
changed the topic from closing the house to protecting oneself. ‘‘I visit those
for whom I pray,’’∞ he said, and this is dangerous. To avert an ‘‘accident’’ the
28 head (of those involved in the ritual) must be protected, which I take to be a
reference to his rubbing my forehead with some substance (dirt he picked
up from the ground, in my memory, but possibly a dawa he had brought
along, according to the text). This precaution must always be taken, he
insisted, and the implication is that he had earlier done this for himself
without my noticing. Rubbing the forehead with medicine was a fleeting
gesture; reflecting on it now makes me stop and face some unexpected
complexities: It is dangerous to undertake the work of protecting a house
against danger. Caution is called for, not because the substances that figure
in the procedure are inherently dangerous and must be ‘‘handled with care,’’
as we know from countless stories about magic, but because danger comes
from the very spirits whose help is sought so that the substances may do
their job. Venturing into their presence (‘‘visiting’’ them, as Kahenga put it)
calls into question conceptions of magic ritual as an impersonal quasi-
mechanical type of action (ex opere operato, work that works because it is
worked). A mere smudge on the forehead dispenses with such a simple
notion. It is not magic that protects one when doing magic (this would call
for an infinite regress). The power of that gesture lies in setting up a
dialectical tension between the material and the spiritual, the impersonal
and the personal, between established routine and precarious happening; it
renders any merely ‘‘pragmatic’’ interpretation of the ritual inadequate.
Next we came to another action that had looked peculiar when I ob-
served it: the covering of the holes (45). Kahenga confirmed that he had
moved the dirt with his buttocks but offered no comment at first. Unlike
stripping, rubbing foreheads, or digging holes, this brought up no associa-
tions except a hunch that it was done to avoid using the hands or perhaps
 An Event

the feet and that it might have something to do with preserving purity.
When I asked about it, Kahenga was at a loss with an answer in Swahili and
switched to Hemba. I was not able to transcribe what he said; the only word
I heard clearly was the verb kuzinda (actually the text has the reflexive form,
kuizinda) meaning, according to the Luba dictionary, to clean oneself by
wiping one’s behind after defecating (Van Avermaet and Mbuya 1954: 827).
So I was more or less on track with my hunch but the connection was not
what I had expected. It was not about observance of purity by the one who
performed the action but about (symbolically) wiping off the ‘‘filth’’ that
might soil the house, meaning the intruders, burglars, and all those who 29
came with evil intentions and whom the ritual was to keep out and afflict
with a curse.
An association between theft and dirt is not only implicit in the ritual
butt-wiping. It can apparently also be made explicit, as I found out when I
looked in the dictionary for other entries under kuzinda. One of them gives
an expression, ube-zinda, which the authors translate as ‘‘tu t’es rendu impos-
sible ici,’’ you made yourself impossible here. They explain this further by
giving as an example ‘‘a child who, while visiting members of the family,
committed theft and will not be allowed to come back. It is considered as
someone who has soiled himself and smells bad’’ (ibid.).
We then moved on to another gesture I had observed after the closing of
the holes (46): Among the things tied up in a piece of cloth he had brought
there was one that contained a seed or pit (not a pebble as I had thought).
Kahenga said that it was a dawa, medicine, adding an explanation I trans-
lated as a statement of purpose: A medicine to ‘‘make sure’’ that the intruder
makes his entry from the back of the house, not from the front. What he
did with the seed was to throw it back over his shoulder so that anyone
entering with force would ‘‘have an accident.’’ This he repeated as he went
from one hole to another. I wanted to know whether he always used the
same incantation. He told me he did but did not dwell on the issue. Rather,
with a chuckle, he came up with a surprising image: When he performed
the action at each hole he placed guards or ‘‘soldiers’’ (on watch).
This came over like a summing-up, a signal that enough was said about
the ritual of closing, and I was ready to go on to a new topic when Kahenga
remembered another detail (47). He told me he had forgotten to bring a
 An Event

‘‘chicken in a basket,’’ without bothering to explain. Upon reflection he must

have meant a live chicken that was to be sacrificed as part of the ritual.
Perhaps this was not an obligatory part of the procedure. At any rate, he
was so casual about it that I suggested we move on to discussing samples of
medicinal plants (48). But then it was my turn to remember something else:
Kahenga had also used water. First he had poured some into each hole and
then he had made the round of the lot, pouring or sprinkling more. This
time he offered an explanation right away, at first hesitantly but ending with
a clear statement: ‘‘Therefore I pour cold water here, which is to say that in
30 a place that stays calm you are bound to live in comfort.’’
Did he pray to the ancestors when he made the rounds with the water
(49)? He did, he said and volunteered an example of such a prayer, first in
Hemba, then in Swahili. As he was reciting this I recalled that he had raised
an arm when he prayed. He told me that this gesture signified that he was
praying to God (50), and this ended, without an explicit statement to that
effect, our conversation about the ritual of closing.

Understanding What Happened:

Ethnography in the Mode of Remembering

naming and knowing

When I first presented Kahenga in the introduction I did my utmost to
avoid attaching to him labels that seemed to impose themselves. Only once
I indicated that his trade may be said to have been that of a practitioner of
magic and sorcery. These terms are handy because they allow one to evoke
concepts that are likely to be familiar to most readers. In anthropology,
‘‘magic’’ and ‘‘sorcery’’ have been as current (and often as central) as ‘‘culture,’’
the guiding idea of our discipline. However, the unselfconscious use of these
terms (and others such as ‘‘tribe’’ or ‘‘myth’’), if there ever was one, is a thing
of the past. There is hardly anyone left who is prepared to speak of magic,
sorcery, or even culture without expressing dissatisfaction, at times to the
point of suggesting that we may have to do without these fixtures of
anthropological discourse. I have been part of that critical chorus; like most
others, however, I have been unable to do without these terms.
This is just by way of prefacing a problem I want to address now. When it
 An Event

came to labeling what happened when Kahenga closed the house I was
much less hesitant than with magic and sorcery. I called it a ritual. Why? I
do believe that ‘‘ritual’’ is somehow less troublesome than ‘‘magic’’ or ‘‘sor-
cery’’ but that does not mean that I am comfortable with the term. To begin
with, the casual use of ‘‘ritual’’ up to this point has had its advantages. It
made it possible to recall the event as something coherent and distinctive;
though the house closing had many discernible parts, calling it a ritual
allowed me to present it as a complete happening with a beginning and an
end and as recognizably different from other kinds of action and interaction
that occurred during the time we were Kahenga’s clients. I am even pre- 31
pared to acknowledge that calling what Kahenga did a ritual opens a wider,
comparative context. But I am bothered by the power such concepts have to
make us prejudge what we are trying to understand, especially when they
come to us casually. Long ago I argued that such ‘‘technical terms’’ have, in
the course of their long history in anthropological discourse, changed from
designations that help us to describe certain phenomena or experiences to
categories ‘‘in terms of ’’ which we approach what we study, categories,
moreover, that may determine a priori what counts as an experience, in this
case, of a ritual.

the burden of foreknowledge

Of course, one may point out that to formulate theories and concepts
that can be tested through empirical study is how science works, hence also
scientific inquiry into ritual. But that was not what I set out to do here. Not
even the idea of a hermeneutic circle that should be appropriate for the text-
centered approach I am taking is of much help. True, understanding based
on interpretation only works when we already know what there is to know.
But I found that in this late ethnography, burdened as it is with accumu-
lated prescience, knowing already what there is to be known can be a serious
impediment. It is commenting on the text of our conversation that allows
me to make this point not just whimsically and generally; at least now and
then I can show how what I called foreknowledge works specifically.
What are the expectations and assumptions anthropologists (and almost
any educated person) are likely to hold about rituals? Here (at the risk of
repeating a few remarks made earlier) is a list: More than other kinds of
 An Event

performance, a ritual follows a script; the practitioner knows exactly what

has to be done and when to do it; prescribed actions are carried out quasi-
mechanically and tend to be repetitive; failing to carry out everything the
script demands, or modifying actions ad hoc, compromises the integrity
and efficacy of the ritual. The list is already getting long although it only
covers ideas about ritual in general. I should add, therefore, some precon-
ceptions I had about the specific ritual of closing. If it was justified to
classify the procedure under ‘‘magic’’ it was a case of protective magic. Its
purpose (or intended effect) was to ‘‘lock’’ the house and property in the
32 sense of closing it up against outsiders with evil intentions. In a magic ritual,
I expected, Kahenga would work with special substances, magic charms,
that were likely to be different from the medicines he used as a healer.
Finally, I also had the idea, perhaps supported by the assurance Kahenga
had shown when we first discussed the closing, that the effectiveness of
magic rituals depended solely on the right charms and a correct procedure.
If these expectations were made criteria by which to judge whether the
closing of the house qualified as a magic ritual, it would have to be declared
flawed on just about all accounts mentioned. In other words, the actual
event thwarted expectations that came from what I thought I knew about
rituals. Nothing was really ‘‘falsified’’—I had not formulated hypotheses to
be tested—yet much became uncertain. But then, having preconceived ideas
and having to let go of them is what makes interpretation move through the
hermeneutic circle.
To show this more concretely we can return to the text and briefly
recapitulate some examples of expectations that were wrong, or at least not
quite right, as they came up when we reviewed the ritual (see the preceding
section):
Regarding the concept that underlay this specific ritual (39), I was in part
misled when I translated its designation in Swahili, kufunga nyumba, as
closing the house. The image I had was one of locking the house as we lock
doors to protect ourselves against intruders. But then Kahenga used an-
other phrase which in turn made me consult a corresponding expression; in
Luba it turned out that ‘‘closing’’ meant something much more complex
than the idea of domestic security I had in my mind.
Next, when the eight holes came up (40) I tried to impose my idea that
 An Event

number and placement may have had a symbolic significance. Kahenga did
not deny this so much as brush it off. And what did it say about the ‘‘magic’’
character of the ritual when he explicitly denied using bizimba, magic
charms, which jibed with his insisting elsewhere that the substances he
brought were just herbal medicines, mizizi (literally: roots)?
Another inconsistency most likely brought about by a preconceived im-
age I had is recorded toward the end of paragraph 40. When I recalled the
ritual from memory I said he stripped down to a loincloth before he started;
in our conversation he says that he put on a loincloth (as part of the attire
required for the procedure) and also a necklace, which I did not remember. 33
The difference looks insignificant until one reflects on it: What I saw as
undressing was for Kahenga dressing up. Not quasi-nakedness (and its
potential symbolic significance) was at issue but the proper attire for the
performance.
When I named the oral components of the ritual I called them prayers
(in the text) and incantations (in the commentary) because they were
delivered in a tone of voice and a manner that met expectations we have of
repetitious ‘‘magical formulae.’’ I asked Kahenga about it, suggesting that he
may have recited these texts ‘‘like songs.’’ He rejected this and insisted that
praying is ‘‘just talking’’ (to a person). Prayers are addressed to spirits; they
are not ‘‘said over’’ something, a sacrifice or, in this case, a protective medi-
cine. An attentive reader of the text may point out that he later appears to
take this back (end of 46). I was clinging to my idea of incantations and now
took a different tack, asking whether what he said was different each time.
No, it was the same, he now told me. Perhaps he simply contradicted
himself; more likely, given the elliptic nature of our exchange (about which
much more will be said in the next chapter), he reasserted that praying was
like ordinary speech: it is neither formulaic in a strict sense nor is it made up
of uniquely creative utterances.
Kahenga had another surprise for me when we got to the rubbing of some
substance on the forehead (44). Only then it became clear to me that he
had done this not just to us but also to himself (and presumably his
assistant). The substance had not been dirt, which had made me remember
taking the ashes, but a medicine prepared for protection. Protection from
whom or what? Not from evil persons or influences the ritual was to fend
 An Event

off but from the spirits he ‘‘visits’’ when he prays to them. Once again, this
oblique remark confounded my diffuse ideas about prayer as part of a ritual:
You beseech spirits not because you fear them but because you trust them
to help you. And about who ‘‘visits’’ whom, I had assumed that prayers were
said to summon a spirit to be present where help was needed, sometimes to
the point—to note another expectation—where the spirit inhabits or ‘‘pos-
sesses’’ the supplicant.
I had observed Kahenga covering the holes by moving dirt over them
with his (almost bare) behind and thought this was a gesture meant to keep
34 his hands clean, literally or in some symbolic sense (45). Initially he con-
firmed this but then our conversation was briefly interrupted and when we
continued the recording he corrected himself. There was symbolism in the
gesture but not the one I had guessed at. He still left room for my first
interpretation but what counted was not moving dirt and thereby closing
the holes. It was wiping oneself as one does after defecating. Kahenga
offered no exegesis and it was only later, when I consulted the Luba diction-
ary, that the symbolic significance became clear.
As we continued, Kahenga mentioned something that had escaped my
attention. He had taken a seed from his medicine bundle and thrown it over
his shoulder (46). Had I been left guessing the meaning of this gesture I
would probably have taken it as ‘‘covering his back,’’ that is, as a protective
act. The explanation he gave was not that simple and made me revise
another assumption—that a magic ritual was either defensive or aggressive.
He made it clear that it can be both and the closing would not just keep
intruders out of the plot but actually put a spell on them and cause them
bodily harm (make them have an ‘‘accident,’’ as he put it). In my way of
thinking this would have meant that I had engaged his services as a sorcerer.
Had this been clear to me, my knowledge, acquired from many instances I
had heard of during the years, of the lethal power of certain spells would
have given me second thoughts. With the hindsight that I have now I must
admit that I failed to realize the ethical consequences of my pragmatic
decision to have our house protected. One might say that time has taken
care of that problem but this does not close the issue. My bringing up ethics
now has also epistemological implications. Does it mean that I share (or
shared at time) Kahenga’s beliefs in the efficacy of his work? For the mo-
 An Event

ment, I let the question stand; about knowledge and belief much more will
have to be said later.

Performance, Conversation, and Commentary

What got me started on this line of thought, remember, were flaws I

observed in the actual performance of the ritual. I then looked at different
parts and recalled moments when my expectations were deceived. I only
discussed examples because any attempt to do this systematically runs into
problems. How is one to divide the performance into its parts and how, 35
short of inventing an artificial nomenclature, does one label the divisions?
After struggling with this for a while I decided to start with an inventory of
discernible elements. While this means setting aside questions of order and
structure, the list will bring out the bewildering complexity of the event I
remembered and discussed with Kahenga.
Let me begin with the four broadest categories allowing us to conceptual-
ize what happened: The ritual required space, time, body, and matter. The
space that provided the physical as well as the social setting for the event was
an urban lot with borders marked by a fence. Time was involved not only as
duration or time needed for the procedure but also as choice of the day and
hour and as the timing of actions during the ritual. A body was needed not
just in a general, obvious sense—someone had to perform—but also as
something that made this performance distinctive (more about that when
we get to actions). The same was true of matter. Leaving aside whether an
immaterial ritual is conceivable, matter in the form of substances had a
prominent part in the closing. Kahenga had brought dawa, prepared medi-
cines. He poured water and moved dirt.
Within the frame set by these categories the performance consisted of
distinct parts. It is easy to name them but difficult to come up with a
classification that would keep them neatly apart. Here is, first, a list of
designations that come to mind: Perhaps most conspicuous were movements.
Most of the time Kahenga was busy making the rounds of the lot, first to
dig the eight holes, then to fill and cover them and to recite prayers.
Another term, already introduced, is actions, such as preparing medicines,
digging holes, placing medicines, pouring water, covering holes, throwing
 An Event

seeds, and praying. Finally, I observed certain gestures, acts I thought had a
special significance. Among them were changing from everyday clothes,
rubbing foreheads, moving dirt with one’s buttocks, pouring water, raising
one’s arm.
This is not a very convincing taxonomy by logical standards since there is
too much overlap in the properties of the classes. Movements of a person
could also be thought of as actions, actions cannot occur without move-
ment, and gestures should probably be a subclass of actions rather than a
separate category. Nevertheless these exercises in labeling and classifying
36 provide a few points of orientation for our commentary on the text that
should be resumed now.
Perhaps most remarkable about our exchange as we have it before us now
is its disjointed and incomplete character. Far from going through the event
in an orderly manner—something that should have been easy since we both
had participated in it recently—we treated topics unevenly; some we took
up only to drop them soon after, others made us branch out into discussions
of subjects that were not directly related to the ritual. If the somewhat
erratic character of our conversation could be attributed solely to my erratic
questioning, this would be a trivial observation. More than once Kahenga,
too, suddenly introduced new issues, came up with explanations I had not
asked for, or took unforeseen directions. These observations bring out
complexities in the presentation of ethnographic knowledge which we or-
dinarily could leave undiscussed but which the resolve to write ethnography
as a commentary on texts forces us to address.
It was no accident that two major reorientations in anthropology—the
interpretive turn and the acknowledgment that much of ethnography is of a
dialogical nature—came, as it were, in tandem. No one has stated this more
clearly than Dennis Tedlock (1979, 1983). Nonetheless, debates about dia-
logue and hermeneutics seem to have moved on separate tracks. My im-
pression has been that advocates of dialogue (except for Tedlock and a few
others) did not worry too much about the textual nature of the records
ethnographers brought back from the field. Some champions of herme-
neutics tended to avoid presenting ethnographic texts altogether. They
argued for interpretation as an alternative to explanatory approaches on the
 An Event

grounds that culture could be conceived according to the ‘‘model of the text’’
(Ricoeur 1971) and text became a metaphor.
Though this was not intended when I proposed commentary as an
ethnographic genre, one of the consequences will be to tackle in our theoret-
ical discussion and our writing of ethnography tasks that dialogical and
hermeneutic approaches failed to realize. A first step would be to see what
happens when in ethnographic accounts dialogue and text are not deployed
as metaphors, that is, when they are not figures but actual objects, material
records of speech, not something ‘‘in terms of which’’ we talk (or write) but
that which we write about. 37
Take dialogue. It is one thing to say that I entered into a dialogical
relation with Kahenga when we conversed, that we engaged in communica-
tion and presumably in a reciprocal relationship different from the one
implied when we talk about investigators questioning informants. It is
another matter to proceed directly from an epistemological position (field
research is dialogical) to a genre of presentation (ethnography could or
should be written as dialogue) without confronting records of conversation
such as our text. As we have seen, presenting our exchange as it was
recorded raised as many questions as it gave answers about what had
happened when the ritual was performed. The task we had set ourselves—
to go once again through the closing ritual—was, as it were, compromised
throughout the conversation by conflict and tension: (a) Between dialogue
and narrative; recalling the ritual would have required a story but the
narrative was constantly interrupted and interfered with by questions. (b)
Between describing and remembering; the aim may have been simply to
describe the procedure and to state what happened but we had no object
before us that could be described, only our memories of an event and our
preknowledge of rituals, general and specific. (c) Between information and
performance; my questions about the ritual were not only about recon-
structing but also about understanding what happened. After all, inasmuch
as I acted as an ethnographer I was after cultural knowledge. Some of the
things Kahenga could tell me about his culture came as information, that is,
as discursive answers to my questions, others just recalled Hemba cultural
knowledge as he performed it rather than articulated verbally.
 An Event

All this goes to say that dialogue ‘‘as such’’ is not the answer to our
epistemological problems (how do ethnographers get to know what they
know?), it is by confronting documented dialogue, texts, that we begin to
address them (how can ethnographers come up with trustworthy and
believable accounts of what they know?).
Moving from dialogical to hermeneutic anthropology, we should now ask
how literal texts rather than text as a metaphor influence our understanding
of interpretation. In general terms, the answer to this question is implied in
the project to write ethnography as commentary, and we have had a glimpse
38 of what this may look like. Giving the portion of the conversation that was
about the closing a prominent place in this first chapter was justified. After
all, it was the ritual that led to the documented exchange on which this
study is based. But if we want to avoid falling into the trap of textual
fundamentalism, we must as yet consider how the text we have begun to
comment on was actually made. As we will see, both the sometimes insur-
mountable difficulties and the unexpected insights I came upon during the
work of transcribing and translating the recording will dispose of the notion
that ethnographic texts are given, ‘‘data’’ to be mined, not to be accounted
for beyond relating the kind of story I told in the introduction. The genesis
of our text will be the subject of the next chapter.
A Text: Made, Not Found

2 The text is there, easily accessible on the Inter-

net. Why not get on with the commentary
instead of burdening, as I am about to do,
the reader with the story of how the text was
u u made? After all, a scientist who works with
a microscope is not expected to include in the report of
his or her findings a technical description and history of
the microscope. The answer is that ethnographic texts
are not instruments of investigation except in a vague,
figurative sense. In the introduction I also rejected a
notion of texts as depositories of facts and called them
mediators. This may be questioned; in my eagerness to
avoid an instrumentalist stance, I may seem to assign to
texts a quasi-personal role. I chose ‘‘mediator’’ because I
don’t think that the more current ‘‘medium’’ reflects what
 A Text

I have in mind. To be sure, written texts, like sound and visual recordings,
are part of ‘‘media’’ whose role in producing ethnographic knowledge has
been recognized and critically discussed for a long time—mostly in general
terms. Here the task is to explore how the making of a specific text is
involved in the process of knowledge production.
At this point, a remark is in order about ‘‘making’’ a text, an expression
that may sound awkward. One transcribes a recording or writes a text, why
insist on calling it making a text? There are two reasons I can think of; both
are epistemological. As I have done already several times I find it useful to
40 stipulate that ethnographic texts are made not found if only as a reminder of
the difference between the ethnographer’s and the literary critic’s texts. This
is not to suggest that literary texts are simply given, as if they did not have to
be appropriated in various ways before they become objects of interpreta-
tion and critique. Still, the literary critic is usually not the author of his or
her text, whereas the ethnographer usually is, at least as far as the kind of
text we have before us is concerned.
There is another reason to emphasize ‘‘making.’’ It expresses a position
according to which knowledge production involves objectivation. Only sub-
jects can know, yet in order to be presented and shared, subjective experi-
ences and insights must become objectified. They must exist materially, be
embodied, in the knowing subject, in the kinds of sensual mediations that
enable subjects to communicate through language, and in all the other
things and practices we usually call culture. Embodied communication
always is precarious, all the more so when it occurs between persons who, in
order to communicate, must cross cultural boundaries. Ethnography is a
transgressive practice (as I argued in the introduction).
Texts are among the objectivations needed for producing ethnographic
knowledge. They are crucial in language-centered approaches and they
support, not exclusively but crucially, our claims to ethnographic objectivity.
This, sparing the reader the details, summarizes the epistemological posi-
tion I adopted and developed over the years and hope to put to another test
in the current project.∞
These general and sketchy thoughts on texts and ethnography provide
the necessary background for a report on the problems and challenges I
faced in transposing the recording of the conversation with Kahenga into a
 A Text

written document. None of the many similar projects I had undertaken

previously had been quite as difficult. None had tried my skills as much as
this exchange and none had pushed me closer to my limits in struggling
with the unforgiving materiality of speech imperfectly articulated and im-
perfectly recorded. But I am getting ahead of the story.

Recording a Communicative Event

In our attempts to understand the ‘‘ethnography of communication’’ we aim

at discovering the cultural and social specificity of communicative events. 41
Like writing, verbal communication comes in distinctive genres. To gain
access to the communicative practices of a society (or ‘‘speech community’’)
the ethnographer must learn to recognize different genres and to act accord-
ing to the rules that govern local speech. This was an exciting insight thirty
years ago; as I see it, it is as important now as it was then.
The text before us is the protocol of a communicative event, a conversa-
tion, as I have been calling it. With the ‘‘components of speech events’’
distinguished by Dell Hymes (1974: chap. 1) in mind we may summarize
the relevant features. In doing this I follow the spirit rather than the letter of
Hymes’s lists, the ‘‘spirit’’ being the idea that the features of a text that is a
protocol should be interpreted according to our knowledge of the charac-
teristics of the speech event that was recorded.
The setting or scene of our conversation was the living and dining room in
our house (the same that was ‘‘closed’’ in the ritual Kahenga performed).
The room, though sparsely furnished, had a fireplace and there were paint-
ings by several Lubumbashi artists of the ‘‘academic’’ school on the walls. As
I recall it, they were by Pili Pili and Mwenze Kibwanga (of the Desfossé
School) and by Mode Muntu (trained at the local Académie de Beaux
Arts). The point of mentioning this is that these pictures contributed to
defining the setting as ‘‘expatriate’’—not as the kind of African salon, embel-
lished by popular genre paintings that Kahenga would have been familiar
with. We were sitting facing each other at the dining table on which we had
laid out some notes, photographs, and samples of medicinal plants. I have
no recollection of the exact time of day. Since the meeting took place on a
Monday, it was most likely the late afternoon or evening. Neither Kahenga
 A Text

nor I had other appointments waiting and we could take all the time we
needed. Incidentally, as we learn from the text (56), the communicative
situation was not new to Kahenga. In his village he had had similar conver-
sations before with a missionary who interrogated him and took notes.
The participants or speakers were Kahenga and myself. The tape also
records sound and noises made by members of the household who went
about in the room while we were talking. Once or twice I had to interrupt
our conversation to attend to their requests.
The linguistic medium (or the ‘‘channel’’) was Shaba/Katanga Swahili as
42 spoken in cities where I had learned it. Although I spoke it fluently my com-
petence was limited, especially as far as the lexicon (words and idiomatic
expressions) was concerned. Kahenga helped with terms I was searching for.
When I did not pronounce them correctly, he sometimes repeated them in
the incorrect form—out of politeness or simply because he wanted to main-
tain the flow of the conversation—but then corrected me later (see the
beginning of 67). Kahenga was at ease in this language but there are indi-
cations that he had had to switch to a local variety from the one he had
grown up with in his home country in northern Shaba (a dialect considered
closer to East Coast Swahili). There was also occasional interference from
Hemba, his native language, and closely related Luba. I suspect that a closer
analysis of this exchange would show that we met, as it were, halfway in a
kind of ad hoc sociolect. I said that Kahenga spoke a, not the, local variety.
This observation was prompted by an unusual trait of his speech, a conspic-
uous absence of French elements. At the time we talked, using French loan-
words and phrases, as well as switching from one language to the other for
entire sentences or passages, was a stylistic feature of language use among
urbanites of Kahenga’s age and level of education.≤ One would need more
information than our text provides to explain why he did not follow that
practice. It could have been a conscious choice, perhaps a villager’s way of
taking his distance from the ways of city people, but this remains a guess.
How to call the genre of this speech event as defined by the purpose and
topic of the communicative situation? Conversation is an adequate term,
generally speaking, but the text shows that this general frame was filled with
a mixture of forms such as interrogation (question and answer), narrative,
explanation, and responses elicited with the help of notes and photographs.
 A Text

This mixed nature is reflected in a lack of discernible order in the sequence

of our exchanges. It even shows up in a certain incoherence of speech, the
symptoms of which are multiple starts, ellipsis, interruption, incomplete
statements, repairs, and corrections. The reason for this may have been
that, even though the exchange had focal topics (the ritual and Kahenga’s
work), it lacked a clear purpose. It was not an interview, even mitigated by
the adjective ‘‘non-directed,’’ if interview means an instrument or method
for obtaining ethnographic data for a pre-established end.≥
What about the key or register in which we conducted our talk? That the
manner of our speech was informal; that we did not raise our voices since 43
we did not address an audience; that we did not argue, joke, or engage in
banter—all this is implicit in the genre of our exchange. Less palpable and
predictable (at least for an outsider) is an effect of the linguistic medium we
shared. Kahenga was also reasonably fluent in French and we could have
chosen to converse in that language. Speaking Swahili meant more than
selecting an available linguistic code. As I have explained elsewhere (Fabian
2003), given the history of this language as the medium of a distinctive
urban culture, talking Swahili creates a kind of intimacy and complicity.
Instead of ‘‘we speak in Swahili’’ people often say tunasema mu kikwetu,
literally: we speak in our-talk.
Finally, a remark on the technicalities of the recording. I used a uher
tape recorder and a microphone. At the time this was just about the best
equipment available. From what I remember and what the text shows,
Kahenga accepted the recorder without questions. He gave no signs of
discomfort nor did the (visible) microphone make him assume a formal
stance (as it had happened sometimes with other interlocutors). If any-
thing, the equipment distracted me more than him. I had to keep an eye on
the reel and turn the tape in time (43).

Listening and Transcribing

When ethnographers discuss the work of transcribing, which they don’t do

very often, they report on it as a chore (one that, if possible, is given to an
assistant who is a native speaker of the language). If they examine problems
of transposing the recording of an oral event into a literary document at all
 A Text

they may talk about difficulties encountered and solutions adopted, espe-
cially as regards ‘‘orthography’’ and other matters of graphically representing
nonlinguistic features of communication.∂ That transcribing requires listen-
ing usually ‘‘goes without saying.’’ This, I think, is quite odd, given the shift
of emphasis in our discipline from observing to listening as sources of
ethnographic knowledge. This turn occurred in the larger context of a
debate about alternatives to ‘‘ocularcentrism’’ in theories of knowledge in
general (Rorty 1980) as well as in critical assessments of anthropology’s
‘‘visualist’’ discourse (Fabian 2002a) and what can be described as a re-
44 cuperation of senses other than vision. However, whereas the closely related
literary turn in anthropology caused at least a minor rally toward an eth-
nography of reading (Boyarin 1993), listening is rarely made the topic of
theoretical reflections or stories of auditory experiences—with one notable
exception, a seminal essay by Regina Bendix (2000).∑
But my task here is not to formulate a general program for an ethnogra-
phy of listening. Instead I want to offer some thoughts, most of them noted
when I listened to this particular recording. To set the mood and to give an
idea of an ‘‘ethnographic listening’’ to a sound recording I should like to
quote from my diary/scrapbook (‘‘field notes,’’ if you wish, taken when I
worked on transcribing and translating our text). The first excerpt is dated
October 30, 2004:
Yesterday I began listening to the recording of my conversation with Kahenga.
The power that these recordings have to recreate presence never ceases to
amaze me. This effect is different from what happens when I remember the
event and check my notes to refresh my memory. Notes, dates, do correct
ideas that were formed in the course of years. An example is the date of the
recording, September 1974. I was sure that our dealings with Kahenga must
have occurred earlier; ‘‘closing the house’’ at Avenue Mpolo somehow makes
less sense just a few months before our departure. Recordings—they may have
this in common with texts—are not like parcels of information, stored away
some time earlier for retrieval and use later. The reason is that presence
must/can be experienced; it cannot be picked up, handled, or whatever is
required to store something. As every reading of a text has been said to be a
new creation so is every listening to recorded voices. Of course, to be precise
 A Text

about the kind of presence that is being created one must note the obvious: it
takes its present qualities largely from past experiences and from my having
participated in the recorded event. Which brings up the issue of remembering
as recognition, in this case perhaps best translated as re-cognition, setting into
motion once again acts of cognition. Hence the shading of thinking into
remembering and vice versa.

Other entries followed until I came to this sigh of relief on June 1, 2005:
Just finished transcription and translation of my conversation with Kahenga.
Must have been the most difficult and challenging job of this kind I’ve 45
done so far. And what a ‘‘text’’! Confusing or transgressing whatever dis-
tinctions were still left in my mind between sound recording and graphic
representation, between transcription and translation, between translation
and interpretation . . .

These impressions and reflections should prepare the reader for the
difficulties, in fact the impossibility, of limiting a report on transcription to
the technicalities of graphic representation. Much like the components of
speech events that can be distinguished analytically but must be imagined
as shaping an event synthetically, that is, by ‘‘working together,’’ recording,
transcribing, and translating cannot be conceived as mere auxiliary activities
in the service of ethnography.∏ I find it impossible, therefore, to keep com-
ments on the transcript and questions on translation separate.
The Swahili transcript as it appears on the web site was not made from
the original tape-reels but from cassette copies. This resulted in a certain
loss of sound quality, but using cassettes rather than tape-reels had the
advantage that they could be listened to on an office-transcriber, a cassette
player that makes constant stopping and backtracking as well as slowing
down the speed relatively easy. The transcript was directly entered and
subsequently edited on a pc. This method suited habits I developed during
many years but it is by now outdated. The sound recording could have been
digitized and available software would have probably made transcribing a
little easier (not much, I think, because the task of ‘‘voice recognition’’ would
still have been mine). It is also possible now to key the transcript directly to
the sound track but this is beyond the current means of our web site.
 A Text

The graphic conventions adopted for transcribing recordings in Shaba

Swahili can be dealt with here quite briefly and generally; specific problems
will be addressed later. The method I follow is best described as a common-
sense orthography not unlike the one used by speakers of Shaba/Katanga
Swahili who are literate but have not been taught a standardized orthogra-
phy in that language. I use three signs: colon (:), slash (/), and question
mark (?). Roughly these correspond to sustained clause, full clause, and
question (or to sustained, falling, rising sentence melody). Incomprehen-
sible phrases are marked . . . ? . . . and brief comments as well as notes on
46 nonverbal signals are enclosed in square brackets.
Concerning the layout of the text on the web site, transcript and transla-
tion are presented in columns and numbered paragraphs. The latter make it
possible to align Swahili and English versions, at least approximately. The
division into paragraphs was of course an imposition after the fact and did
not follow rigid criteria. It reflects changes of topic, subdivisions within
topics, whether marked by pauses or not, and may at times appear some-
what arbitrary. Its main use is that numbered paragraphs can be referred to
more easily in the commentary. Numbering lines would have been another
possibility but also an encumbrance I decided to avoid.
On the web site the texts also include notes. These were added to justify
or briefly explain certain transcriptions and translations but also to note
matters of interpretation. In fact, the accumulation of notes prepared for
texts to be deposited in the virtual archive first made me think about
commentary as an ethnographic genre. Most of the annotation that accom-
panies the conversation with Kahenga on the web site has been worked into
the current project.

Transcribing and Translating the

Conversation with Kahenga

‘‘lost in transcription’’
It was Ray Birdwhistell, I believe, who once called ethnographic record-
ings ‘‘cadavers’’ of speech. Raising these cadavers from the dead, poetically
speaking, is a challenge; the reward is a document that is going to have a life
of its own. Another image, that of hidden wealth recovered by transcrip-
 A Text

tion, however, only partially describes what happens between listening to a

recording and writing down what one hears. ‘‘Reducing a language to writ-
ing’’ was a phrase common in the nineteenth century. It meant that a
language was ‘‘described’’ by means of vocabularies and grammars. In more
than one respect, making a text requires reducing a wealth of information
and describing rather then just transcribing sound.
Traces of reduction and description can be found in the Swahili text as it
appears in our virtual archive. On a first level, reduction is involved in the
decision to present a phonological rather than a phonetic transcript. Both
modes describe rather than simply reproduce sounds. This inevitably has 47
the effect of making disappear, as it were, variation in speech sounds that
does not affect the meaning of words. But in our case it also meant that the
great variability of pronunciation that is a characteristic of local Swahiliπ
and differences in ‘‘accent’’ between Kahenga and myself were rendered
invisible (both have cultural and communicative significance).
Reduction is, above all, necessary to separate speech from other sonic
information on a recording. Kahenga and I produced many vocal sounds
that were not verbal. By far the most frequent example is ‘‘mm,’’ which can
express affirmation but mostly just acknowledges a statement as a way of
keeping the exchange going. Some vocal expressions are impossible to tran-
scribe and can only be described, among them ‘‘chuckle,’’ or ‘‘laughing.’’
Volume, speed, pitch, timbre, but also pauses, patterns of breathing, clear-
ing one’s throat, and many other audible features were not noted at all,
except when they conveyed information I thought was significant.
The recording also preserved much of the sonic environment in which
our conversation took place. It caught ‘‘noises’’ outside and inside the house
most of which were filtered out of the transcript even though they may be
important in triggering memories of the event often needed to recognize
what was being said (and what it meant). The same goes for sonic informa-
tion regarding body posture and proximity, that is, features of communica-
tive interaction.
As often as it seemed useful I included in the transcript information that
is audible on the recording but cannot be directly represented graphically.
This was done by adding more than two hundred short glosses between
square brackets as well as numerous footnotes explaining or justifying a
 A Text

given transcription. Both the glosses and the footnotes are in English, and
that is one of the many ways in which the target of the exercise, a translation
into English, affirms its presence in the Swahili text even before a full
translation is accomplished.
Almost all recorded conversations pose problems when it comes to tran-
scribing overlapping speech. Linguists have devised notations for this (and
for many of the other nonverbal features mentioned above) but I think that
attempts to reach graphic accuracy are always compromises and most of
them are practical only for short passages that one wants to analyze closely.
48 I opted for a successive presentation of overlapping utterances whenever it
was possible to separate them. Often only one of them appears in the
transcript, the other is marked incomprehensible ( . . . ? . . . ). Overlap is
often caused by one speaker interrupting the other (marked . . . ), not
necessarily because he disregards rules of (polite) turn-taking but by some-
thing one could call ‘‘anticipated’’ responses. Such interruptions need not
have a negative effect on the exchange. Close examination, I suspect, would
reveal that (much like the ‘‘mm’’) they are timed and follow a certain
rhythm, adding intensity and fluency to the dialogue.
More than sixty times I had to mark passages (words, phrases, seldom
entire sentences) as incomprehensible. Many of these gaps are due to the
kind of problems I just described, others are caused by the bad quality of
(parts of) the recording. Yet others simply reflect a characteristic of our
conversation as a whole. Being relaxed and informal about this exchange
also made us careless with articulation; often words and phrases came out
slurred, too fast, or not loud enough (in some instances I had to leave gaps
in transcribing my own questions and remarks!). Kahenga tended to con-
tract and often simply drop syllables/morphemes. Many vowels in un-
stressed position are scarcely audible so that it becomes difficult, for in-
stance, to decide whether a verb starts with ana- or ina- and that may pose
problems of determining the agent (personal or impersonal).
These observations on careless enunciation do not only apply to the
articulation of sounds but also to the formation of words and sentences.
There is a lot of—how to call it?—bricolage in the way we speak: starting
with a prefix, then stopping briefly to add another morpheme to complete
the lexeme or choose a different one. Other ‘‘stopping devices,’’ such as
 A Text

inserting nani (roughly: what was it again?) also belong here. On my side
this ‘‘piecemeal’’ way of speaking may have been a sign of hesitation due to
limited competence but Kahenga’s recorded speech is not all that different
in this respect. Of course, this could be because his command of Shaba
Swahili is also somewhat limited (although that does not affect the ease and
speed with which he speaks). All in all, I think this characteristic of our
exchange should be understood as a matter of register or style, expressive of
a certain lack of clear purpose. What made us carefree in this conversation
also made us careless about articulation. After the fact I am convinced that
these imperfections added to the productivity of our exchange. 49

‘‘found in translation’’
Taking, as I just did, these observations on the challenges of transcription
from the articulation of sounds to the formation of words and sentences
brings us back to an insight I stated at the beginning: In the making of a text
such as ours it is impossible to keep transcribing and translating completely
separate. When I transcribe I must know that what I transcribe is meaning-
ful. When I translate I spell out meanings recognized earlier. And when
that proves difficult or impossible I may change the transcript after listening
once again to the recording. Sometimes pondering the translation suddenly
makes an alternative transcription of a word or phrase plausible.
A transcript is more like a sculpture than a picture in that it is made in
three dimensions. Work on (1) a sound recording of speech with the aim of
(2) a graphic representation always also requires (3) understanding, however
incomplete or provisional, of what is being said. That is as far as the image
of a sculpture goes. The next step would be to ask whether making the text
is like carving wood or more like modeling clay. After what was said about
reduction and poesis the answer could only be ‘‘both’’ and that would not be
very illuminating.∫
I said that I modeled my transcriptions of Swahili texts on grassroots
literacy. This needs qualification. In the course of preparing the edition and
translation of a typewritten text, a colonial history of Lubumbashi called
the Vocabulary of the Town of Elisabethville, I encountered major obstacles
caused by features of uncontrolled literacy. Above all, the writer showed
little respect for exact and consistent segmentation, that is, for word and
 A Text

sentence boundaries. Spaces, capital letters, and punctuation were used

liberally but erratically, as ornaments of literacy rather than as aids for
reading. The solution I eventually found for these problems was quite
simply to re-oralize the text by treating it as it was intended in the first
place: as a script for an oral performance. A native speaker and writer of
Shaba Swahili read the Vocabulary for me, the reading was recorded, and the
recording was transcribed. It was this second text that made the translation
possible, in fact, quite easy. Most of the work, to repeat my point, had been
done by establishing a transcript in which word and sentence boundaries
50 could be recognized.Ω
In fact, ‘‘recognition’’ best describes the sensual-intellectual operation that
produces a transcript of the kind we call an ethnographic text. Recognition
is also a cue to the role of memory in ethnography in general and in the
making of ethnographic texts in particular.∞≠ Transcribing and translating
need remembering in that the one who transcribes and translates must have
phonetic, lexical, and grammatical competences—a repertoire learned in the
past and capable of being activated (remembered) in the present. Such
remembering is of course required whenever we speak or write, no matter in
what language. Usually it remains, as it were, in the background; remember-
ing is then more a condition than an activity. But in the course of transcrib-
ing a recording there are countless moments that make us stop with in-
comprehension or let us hesitate because we see alternatives that must be
weighed. It has been my experience that we resolve most of these problems
because the recording makes us remember what we hear. Such memory may
be created text-internally, that is, in the course of working on a given text
(how did I transcribe/translate a given utterance earlier?). It may also come
as remembering a speaker’s attitudes (including bodily postures, changes in
directions of gaze, signs of excitement or lagging attention, and so forth)
or as realizing the significance of indexical and nonlinguistic information
which all sound recordings are full of but which in many cases can only be
re-cognized (understood once again) by the ethnographer who was present
when the recording was made.
Because of the precarious and often fragmentary makeup of the tran-
script, the English version does not so much reproduce as reconstruct the
meaning of the Swahili text. Constructive creativity is above all required
 A Text

when it comes to solving problems that are posed, for instance, by specific
characteristics of Swahili or by the nondiscursive, performative nature of
free conversation. Among the former are problems of reference. In Swahili,
nouns, most proper names, pronouns, and verbal forms are not marked for
gender.∞∞ Outside the context of live speech it may be impossible exactly to
determine who or what is referred to by a pronoun or demonstrative. To
represent the ambiguity of gender every time it could not be resolved with a
‘‘he/she,’’ for instance, would have made the translation even more awk-
ward. So I often opted, as I do in my writing generally, for the gender of the
writer (hence, ‘‘the spirit, he . . .’’). Equally pervasive are questions of tense. 51
Like many speakers of local Swahili, Kahenga only uses three kinds: past,
present, and future (marked by the infixes -li-, -na-, and -ta- respectively).
Furthermore, the -na- tense, strictly speaking, places action not so much in
the present as in an unspecified time.∞≤ Since English has no equivalent,
many verbs in the -na- tense can, depending on the context, be translated as
actions that took place in the past or are going to take place in the future.
Another difficulty of translation has been ellipsis, not only in the rhetori-
cal sense of the term in which we may call statements in any language
elliptic: We ‘‘get the gist’’ but we sense that much remained unsaid on
purpose and we would like more elaboration. Such rhetorical omissions
also occurred in our conversation but here I want to point to ellipsis as a
characteristic of the Swahili text and as a challenge to the translator. The
reasons for this trait (of telling it all without saying everything) are complex
and diverse and this commentary is not the place to document and analyze
them in detail. Still, I would like to give a summary which will allow me to
add some observations on the ‘‘ethnography of communication’’ of our talk.
First there is the social history of Shaba/Katanga Swahili. Ever since the
1920s, when it became the African language shared by local people, most of
them recruits to the mines and other immigrants attracted by the growing
towns of the region, it was usually described by comparison to Standard
Swahili as a vehicular language characterized by a limited lexicon, a reduced
grammar, and some peculiar features (mainly morphological and lexical)
ascribed to the influence of local Bantu languages. When I say ‘‘described’’
this should be taken in an informal sense: this was how expatriates and
educated Africans talked about local Swahili. There were some early at-
 A Text

tempts but linguistic descriptions proper first appeared only in the seventies
and a thorough and definitive study is still not available.∞≥ More than a
million people, many of them residents in the area in the third or fourth
generation, speak local Swahili as their first, their principal, and more and
more as their only African language. In such a situation the qualification
vehicular (often with the sous-entendu ‘‘merely’’) has lost its sense. This
variety has emerged and developed as a popular language, that is, as a
medium of a popular and ‘‘modern’’ urban culture. Though many, perhaps
most, of its speakers are literate, Shaba Swahili is taught and learned
52 informally, outside the educational system. People use it above all in oral
communication and when they write it they do this unhampered by formal
rules. By now we know it does not make much sense to declare, say, Haitian
Creole a reduced or deficient form of French or, for that matter, to describe
any language by listing what it lacks in comparison to another, and this goes
for Shaba Swahili, too.∞∂ This being said, it is nevertheless true that docu-
mented exchanges in this language, some more than others, have elliptic
features in the sense that Shaba Swahili achieves semantic and grammatical
precision typically by means that are available only in oral communication—
especially all the ‘‘nonlinguistic’’ information from intonation to body lan-
guage I mentioned earlier. Translating a Swahili transcript like ours is an
exercise in the ethnographic description of verbal performance even before
one begins with ethnographic comments.
Second, there is a certain propensity for elliptic speech in the way Ka-
henga and I talk to each other. Especially in the part of our conversation
where he comments on plant samples and photographs (starting with
paragraph 57) speech is often indexical; brief statements are accompanied
by pointing and other gestures, for instance, of assent or negation, often
impossible to reconstruct from the audio recording. Indexical communica-
tion about objects that are physically present achieves its aims without
description or explanation and this further contributes to making the ex-
change more performative than discursive. At one point (63), one of my
glosses to the transcript notes a ‘‘sniffing sound.’’ Here and later, smell is one
of the properties Kahenga uses to identify plants. This further complicates
the situation: He not only selects samples and puts them before us but also
lifts them up to let me smell them and then smells them himself. When he
 A Text

does this, often the only verbal clues he gives are the Swahili demonstratives
ile, ule, hii, huyu, whose exact meaning or reference, as I said before, is
difficult to determine semantically. In sum, much of what was communi-
cated in these passages was lost in the recording and lacking in the tran-
script. It had to be restored from nonverbal, contextual information aided
by memory if the translation was to make sense.
Third, our exchange can at times become elliptic in ways I encountered in
most of the ethnographic texts I recorded during the years. Again, gaps are a
characteristic of all verbal communication: When something is told not
everything needs to be said (or can be said). When Kahenga seems to 53
assume that something goes without saying, there are several possibilities to
account for this. He may not be aware of making such an assumption or, if
he is, he may not care to question it. But he may also have reasons to assume
that nothing, or not much, is lost on me when he is being elliptic. The
assumption, then, is that of a common understanding that does not always
need to be articulated.∞∑
Readers of the translation may be surprised at these extensive observa-
tions on ellipsis in view of the fact that so much in our exchange appears
redundant and repetitive. Repetition of words and phrases, like multiple
starts with a word or sentence, are among the trickiest problems of transla-
tion. Why does one balk at the simple solution, which would be to trans-
pose the transcript one-to-one? Pondering this made me realize yet another
reason why the relationship between transcript and translation should be
seen as a process in the course of which understanding emerges (like the
text itself, understanding is made, not found). Redundancy, though a fea-
ture of all language communication, is more pronounced in oral exchanges
and it may be rampant in the kind of informal, roaming conversation I had
with Kahenga where we sometimes resorted to repetition just to maintain
the flow. More than once, trying to be faithful to the transcript made me
stop short of leaving out entire passages because translating them came
close to making a caricature of this conversation.
Finally, on the web site, transcript and translation are presented in facing
columns. The idea was to make it easier for readers to move between the
Swahili and English versions. But a glance at the site shows that the two
are less than perfectly aligned. As noted earlier, in the making of an ethno-
 A Text

graphic text, transcription and, even more so, its translation involve eth-
nographic description, hence the apparent gaps in the Swahili column.
Though the present alignment could be improved by shortening the pas-
sages that are juxtaposed, this would hardly improve the overall appearance.
In the end, the reasons for having transcript and translation face each other
are more symbolic and rhetorical than practical. They typographically ex-
press the theoretical argument to which this chapter has been devoted:
Much like ethnographic communication itself, the presentation of its docu-
ments consists of confrontation. A translation confronts the original, it does
54 not consume it. Making ethnographic knowledge is an always incomplete,
unresolved process, even at the level of making the texts we then analyze
and interpret.
The principle that follows from such a processual view—no translation is
final—will also be applied when I comment on the text in the chapters that
follow. When I select expressions or statements for commentary they will
rarely be simply copied from the translation on the web site. As a rule I will
be working from the Swahili text and translate relevant material anew, often
profiting from gains in understanding that result from rereading a passage
in its interpretative context. Some discrepancies with the posted English
version will be in wording only but others represent alternative translations.

Endnote: Learned Ignorance?

Paradoxically, the narrow focus on one specific text makes one aware of an
immense context of technical and scholarly writing about transcription and
translation. Because I cannot pretend to be ignorant of its existence, I feel I
should justify why I chose to ignore so much relevant literature in this
chapter. An excuse could be to point out that every ethnography enters
territories staked out and ruled by specialists and that no ethnography
could be written if it went off in all the directions opened by the questions
our documents make us ask. Of course, not citing or discussing specialized
writings may involve a higher cost than I realize. I am willing to pay the
price as part of my experiment with writing in the genre of commentary.∞∏
Kahenga’s Work

3 The introduction described how this proj-

ect originated. The chapters that followed
recalled the event that started it all and re-
traced the ‘‘chain of evidence,’’ connecting the
text deposited in a virtual archive to a time, a
place, and the speakers/actors whose conversation was
recorded. We now have an idea what it takes to make
an ethnographic document. We know how it was done.
The next step will be to remember that ‘‘document’’ is
derived from Latin docere, to teach, and ask in this and the
following chapters what it is our text has to teach us.
I resolved to write this ethnography in the genre of
commentary, a kind of writing in which priority is given to
a text that is virtually present, for the writer as well as the
reader. What, if this is to be more than a declaration of
 Kahenga’s Work

intent, does ‘‘giving priority’’ entail? Answers to this question already given
or alluded to came down to the ‘‘rule’’ that the agenda of producing knowl-
edge and the order of presenting information and findings should be de-
rived from the text rather than imposed on it by the kind of scheme that
used to guide the writing of monographs in anthropology—at least ideally.
This experiment, however, begins to show that it will not be possible
strictly to follow such a rule in practice, precisely if presentation is to be
guided by respect for the text. Thus, a breach was committed in the first
chapter in that our commentary did not begin with the beginning of the
56 text but with a passage from its middle. This was called for because that
passage covered the ritual of closing the house, the event and topic that was
central to our conversation. Before beginning to ‘‘cover’’ the remainder of the
text, the commentary had to be put onto the ground on which this exchange
stood (or moved, when topics other than the ritual were discussed). Unlike
the classical commentator on authoritative writings, for whom the text itself
may be the ground to stand on, the ethnographic commentator is beholden
to a world outside the texts he or she produced as mediations of knowledge.
The ethnographic commentary must take notice of that world, and that will
be the task of this chapter.
In deference to the text, Kahenga, the person, will be introduced later
when we get to relevant passages (24–26, also elsewhere, especially 54). We
will begin with Kahenga, the practitioner, who worked in a trade that was as
common, in demand, and accessible as many others serving the local popu-
lation in what economists with a top-down perspective usually call the
‘‘private’’ sector or the ‘‘informal’’ economy. If looked at from below, neither
adjective makes much sense without numerous qualifications. A healer’s
work, for instance, though carried out in the private sector, is in the public
eye and there is nothing informal about the competences, procedures, or
even the economic transactions that allow him or her to make a living in
the informal economy. Furthermore, that informal activities escape state
control does not mean that they are ignored. Conflicts, or meeting-points,
between professional medicine and the art of the healer have been recog-
nized since colonial times. In the Congo, attempts to bring these relation-
ships under control date back the decades before Independence in 1960 (see
Janzen 1978). The meeting with Kahenga may have been a matter of chance
 Kahenga’s Work

but it is important to keep in mind that it occurred in an existing social and

political context.

Kahenga Plying His Trade

It is now time to return to the beginning of our exchange (1), an odd, abrupt
beginning. We had already started to talk about plant specimens, medicines,
laid out on the table at which we sat down when I pressed the record button
and we were off to a running start. ‘‘This,’’ Kahenga told me, ‘‘is for rubbing
in,’’ and went on to talk about a medicine to counteract a ‘‘bad medicine.’’ I 57
let Kahenga finish explaining this before I cut this exchange about medi-
cines short by taking a different, indirect tack. Setting aside the samples
before us, I asked him to tell me about today’s work (2). After that we
backed up for a review of last week’s cases (3) he dealt with when he stayed
in Kenia township at the house of an older relative of the young man who
had assisted him during the closing ritual. We embarked on a wide-ranging
review of his typical clients and their typical problems (4). Put simply—too
simply, as we will see—we sought answers to three questions: Who came to
him, what were the complaints, what were the remedies?
Within each of the three categories we looked at (patients, ills, remedies)
our conversation brought up numerous distinctions of domains and of
classes and subclasses. Therefore, what we exchanged discursively could be
reduced to, and presented as, taxonomies. If I were to do this now I would
follow the classical path taken at the time when medical anthropology was
still ethnomedicine, a branch of ethnoscience. Ethnoscience worked with
the hypothesis that cultural knowledge is essentially classification and that
anthropologists should work out in their study of cultures the distinctive
ways in which different societies structure domains of experience in their
‘‘dictionaries.’’ These structures, it was maintained, could be read from the
logic of taxonomic arrangements (relationships of contrast, opposition,
inclusion and exclusion, and so forth).
Ethnoscientific studies were impressive on logic but less so on ideology,
rhetoric, and hardly at all on history. Above all, they had no concern for the
integrity of ethnographic texts (or the communicative events texts docu-
ment), and when texts (rather than brief statements, preferably single-word
 Kahenga’s Work

responses) were used for analysis, and then were only to be mined for nontex-
tual, nondiscursive information. Because respect for the text is the pro-
claimed ideal of writing in the genre of commentary, I will, by and large, resist
the temptation, and deny myself the pleasure, of playing taxonomic games
and stick to descriptive summaries and paraphrases until enough is said to
support reflections on a theoretical basis broader than that of ethnoscience.

clients, problems, remedies

The following list of ‘‘cases’’ (in my questions I used the French loan cas,
58 plur. macas) was compiled from our text. It respects the sequence in which
consultations were mentioned and should answer the guiding questions in a
straightforward manner—at least until we must face questions that our list
raises rather than answers.
Kahenga first mentioned two female clients. ‘‘They were women who
wanted dawa,∞ a medicine that would help them with their work.’’ The
problem was not specified in the first case. The second woman feared that
she might be fired, as had happened to many others recently. Remedy: A
(powdered) herbal medicine (2).
A married couple consulted him about problems of fertility. Remedy:
The husband’s blood gets ‘‘fixed’’ by drinking a medicine (4).
Frequently women sought help with ailments of the uterus. At this point
Kahenga described only symptoms in some detail but said nothing about
the treatment. For some reason, perhaps because treating gynecological
complaints was a specialty of his and a major source of income, he told me
that a woman would pay him ‘‘(only) if she is cured’’ (5).
Then problems at work—labor relations—came up again. When ‘‘a per-
son wants to be liked by his superiors’’ Kahenga had medicine for him (in a
bottle) or fitted the client with a kind of belt to wear. Kahenga volunteered a
number: During the preceding week twelve persons ‘‘wanted to be liked at
their workplace.’’ Each client got medicine suited for his or her case (6).
Marital problems came next. As an example he cited the case of his
assistant. The latter reported that treatment by an unspecified but probably
herbal medicine was successful (10).
People also come to Kahenga with various kinds of abdominal pain,
 Kahenga’s Work

caused by ‘‘bugs’’ (more about this in the next section when we comment on
causation); herbal medicine was the remedy (10).
In one of the most interesting but unfortunately most cryptic passages
Kahenga then spoke about being consulted by a thief (or a gang of thieves?)
planning to steal an expatriate’s car. His client wanted him to ‘‘prepare
something’’ to assure the success of the operation. At least, this is a possible
reading of the first part of the story which then abruptly switches to the
treatment Kahenga had for the thief after he was injured in his chest when
the car hit a wall during his getaway. No remedy was mentioned, just the
fact that Kahenga was paid (11). 59
Closing a house and bringing peace to it was mentioned next. Kahenga
remarked that ‘‘even back home in the village it is just the same,’’ probably
because he wanted to make it clear that the particular service that had
brought us together was in frequent demand (12).
The association with our case brought us to his treating my spouse for
her pains and the children of Mwenze (the painter who had recommended
Kahenga to us) for suffering from weight loss and sungu, swollen legs (13).
Again, the remedies were herbal medicines he also used when he was
consulted by women (including prostitutes) who wanted to be loved, by
chiefs who wanted to have authority, and by women seeking abortions (14).
The list of ailments continued with pain in the joints and limbs, itching
skin, a skin disease called bukoma, and headaches (16) and even after we had
moved on to other topics Kahenga kept adding to the catalogue: impotence
and another kind of abdominal disease (20), swelling of the back, bones
bent or broken (22), gonorrhoea (23), and, as a whimsical afterthought
toward the end of our conversation, loss of hair (58).
For the sake of completeness it should be mentioned that Kahenga,
acting on behalf of his grandmother and teacher, was also involved in
procuring medicine for politicians, such as Godefroid Munongo, an impor-
tant figure during the secession of Katanga (33), and certain unnamed chiefs
(38). I would not rule out that putting himself into the role of mere
messenger for his grandmother in these cases was his way of avoiding a
touchy subject. Having been involved with leaders of secessionist Katanga
could have been dangerous under Mobutu’s regime. Kahenga had men-
 Kahenga’s Work

tioned chiefs as his clients earlier (14) and service to politicians may well
have been part of his trade at the time when we talked.
At the beginning of our conversation I had asked Kahenga what he had
been doing just before we met; at the end I wanted to know what he had
lined up for the days to come. Several clients were waiting for him in
Musoshi (a mining town about thirty kilometers south of Lubumbashi).
After that he was going to ‘‘look after this person who has twins,’’ help
the relatives celebrate, and ‘‘do what needs to be arranged for a mother of
twins’’ (75).
60 In our conversation we then turned from an inventory of specific and
typical cases to discussing, in more general terms, Kahenga’s approaches to
illness as regards diagnosis and treatment, and this will be commented on in
the next section. Though it means abandoning the sequence of our ex-
change, the catalogue of complaints and diseases should be confirmed and
completed with information Kahenga provided when we looked at the
herbal specimens he had brought back from the collecting expedition he
undertook, accompanied by my spouse. He named each of them and told
me what it was used for: stomach trouble (60), eye trouble (61), treating
children who have digestive troubles or bad dreams (62), stomach trouble
(63), problems at work (64), another preparation for treating children (65),
one against a bug (66), others for solving problems at work and generally for
removing pressure (67), a medicine to enhance virility (68), and one for a
swollen neck (69).

diagnoses, treatments, etiologies

Kahenga had explained how an abortion is brought about (14) and proba-
bly would have added other treatments to the list when I decided to change
the approach. I paused and then directed our discussion from compiling
cases to taking a closer look at a consultation (15). I asked him to tell me
what happens when a sick person comes to see him. First, I used maladi, a
French gloss current in Shaba Swahili that can mean either illness or a
person who is ill.≤ Then I made this less ambiguous with the Swahili
synonym mugonjwa, a sick person. After the fact, I realize that this was one
of the instances of preconceived ideas or expectations guiding my questions.
 Kahenga’s Work

Unfazed by information suggesting a much wider scope of his activities that

had come up in our conversation, I had Kahenga down as a medical
practitioner. My mistake was excusable—after all, he had called himself a
munganga, a term used for ‘‘doctor,’’ be it herbalist or physician—but it gave
our exchange an orientation that could have seriously limited its ethno-
graphic value had it become the main line we followed from then on. The
corrective came from the kind, or genre, of communication we were engaged
in: not an information-centered, topical interview but an event-centered
reflexive dialogue.
Kahenga responded to my leading question—what happens when a client 61
comes?—with a sense for the practical. It depends on the complaint, he told
me (without saying it) and took as an example a woman who has problems
with her uterus. The first step is a physical examination, which in such a
case is carried out by his wife.≥ Then he listens at length to ‘‘them’’ (presum-
ably his wife, the patient, and persons close to her). Insisting that he is the
one who ‘‘knows’’ what the problem is, he told me that he procures the
medicine and gives it directly to the patient. We made another start to
clarify this and Kahenga confirmed the basic pattern: To come to a diag-
nosis in a case where probing and palpating a woman’s body is ruled out,
listening to what the patient and the munganga’s proxy have to say is essen-
tial. Notice also that Kahenga always got to the medicine, dawa, when he
explained how he worked. He did not mention in this context or elsewhere,
as far as I can see, that he might have to revise his diagnosis if a remedy
did not help but he repeatedly pointed out, in a way that made this a matter
of principle, that the efficacy of a medicine can be determined only by trial
and error.
My next move was to steer our discussion to the question of causes.
Before I comment on this, however, I should note that the topic had come
up earlier in a passage in which I tried to keep the focus on kinds of illness
(10). There he had told me that a kilulu could cause a sickness of the belly.
The term, meaning literally a small insect, a parasite, or a bacterial infec-
tion, is difficult to translate. I tried ‘‘vermin’’ when it first occurred; later (63)
I used the colloquial ‘‘bug,’’ which may be a better rendition of the non-
specific meaning the term has in Katanga Swahili.∂ To a question suggesting
 Kahenga’s Work

that all pain (meaning here: illness) may be caused by kilulu, Kahenga
responded by first redirecting me to (illness) of the belly and then specifying
that a kilulu is responsible for reproductive troubles in men and women.
Let me return now to the point in our exchange where I tried to steer the
discussion from kinds of illness to causes (16). At first, it appears that the
attempt was bungled by both of us. Kahenga took my question for the
origins of illness as a prompt to continue the enumeration of kinds of illness
we had concluded (or just stopped) earlier. I cut this short but took his
reaction as a cue for raising the issue of classification. I suggested that, as
62 there are kabila, literally: tribes, of people there must be kabila, kinds of,
illness. He acknowledged the clumsiness of my comparison with a chuckle
but he agreed: There are kabila. When I asked him about the kinds he knew
he began to show signs of impatience: Isn’t that what I have been telling you
about? Still, maybe just to humor me, he volunteered a rudimentary classi-
fication that should drive an ethnoscientist to despair: There are illnesses ‘‘of
the uterus’’ and those where the limbs get thin, the defining traits being an
organ in the first and a symptom in the second kind. Then it was his turn to
switch abruptly from kind to cause. He brought up the example of a skin
disease, mpesé, brought on a person by a dawa set like a trap (that is the
literal meaning of the verb -tega he used here).
In our conversation this was an offhand remark that did not break the
stride of our struggling, or juggling, with kinds of illness. For the commen-
tary it presents a momentous finding and challenge. There is nothing
intrinsically odd about stating that a ‘‘medicine’’ can also cause damage; in
television commercials we hear warnings about side effects of medication
daily. But this was not implied in what Kahenga told me when he said that a
dawa can cause an illness, and this sets the terms we have been using for his
work—healer, illness, patient, medicine, remedy—afloat. We will not be able
to do without them and don’t have to as long as it is understood that, even
more than this is the case with all translation glosses, the English terms
cover the semantic space of their Swahili counterparts only partially and
often, depending on the context, hardly at all.
Kahenga named another skin condition he ‘‘can do’’∑ and then picked up
the thread of causation again, speaking of an affliction whose symptom is a
one-sided headache. One can catch it by contagion, by touching a tree that
 Kahenga’s Work

was infected, as it were, with the evil intentions of another person. This is
one of many moments, some encountered earlier, others still to come, when
translation succeeds at a price. It conveys a thought we can understand, yet
choosing a word like ‘‘infect’’ transports understanding into the realm of
metaphors. By making what Kahenga tells me somehow familiar it lessens
the challenge his statement poses for interpretation. The same would have
happened had I given in to an even more tempting alternative and para-
phrased the etiology of the one-sided headache as catching a ‘‘curse’’ from a
tree where someone had put it. This would have placed Kahenga’s ideas
altogether in the realm of the irrational, cutting short any attempt at rational 63
comprehension (other, that is, than anthropology’s old tricks of understand-
ing the nonunderstood by placing it somewhere on an evolutionary scale of
prerational thought or having it ‘‘symbolically’’ serve social functions).
Infection by evil intentions was in fact the condition he had diagnosed in
my spouse and against which he had prepared a medicine. The text leaves
matters unclear. As best I can tell, Kahenga had taken the remedy from an
‘‘infected’’ tree and had thereby removed the ‘‘infection’’ from that tree. The
problem of translation here is not only that Kahenga’s remarks were once
again elliptic; an even greater difficulty—one we can only signal here but will
have to face later—is posed by a term that ranks with dawa as a pivot of
Kahenga’s discourse: miti, the plural form of muti, which refers to a tree but
also to a plant or vegetal matter in general. What goes for dawa also goes for
miti: It can heal an illness but it can also cause it.
What kind of causation is this? Both the medicine prepared to ‘‘entrap’’ a
person in a skin disease and the infected tree that gives a person who
touches it a headache are carriers of a person’s evil intentions. Therefore—at
least in the cases described—dawa and miti are presented as links in a chain
of causation that leads to a person. What exactly does that mean? Taking a
lead from the expression Kahenga used (kutega dawa, see above), I under-
stood the act as being intentional, but this raises questions if it is also to
apply to the case of the ‘‘infected’’ tree. The way it is depicted here, catching
the illness from a tree is a matter of chance. How are we then to understand
the causal link between evil intent and bad luck?
Since the passage on which we are commenting offers no help with our
queries we may as well move on to the next one where I somewhat abruptly
 Kahenga’s Work

asked Kahenga whether Mungu, God, can ‘‘send illness’’ (17). I don’t think
that it was mere logic that made me take this leap (moving from proximate
causes to ultimate cause) but rather, once again, knowledge (or memories) I
had brought to our conversation. I knew that very old people who died, as
we would put it, from natural causes, were said to die lufu ya Mungu, God’s
death. Could this induce Kahenga to clarify what he told me about illness
caused by evil intentions? Unless Mungu could be evil, how can God be
causing illness? He was unfazed and had an answer ready: God sends illness
as punishment, as a ‘‘payment’’ (malipizi), to be precise. This response
64 looked innocuous and could have been expected but what happened was
that Kahenga switched to a moral or perhaps legal discourse. God is not
evil, God adjudicates and imposes illness as a penalty for a sin or transgres-
sion (kosa). I kept probing for causation and asked whether it is the sin that
makes you ill, not paying attention to what he had just explained: God’s role
is to ‘‘return’’ to you the sin you committed. A person who does another
person wrong may find himself or herself suffering from a stomach-ache or
a bad back. In the hypothetical case he then cited he had someone I passed
on the road insult me as a ‘‘stupid white guy.’’ This, he suggested, makes me
angry and ready to exchange blows with the offender. God steps in, as it
were, and makes him pay in the form of a back-ache he is going to feel as
soon as he gets home. I wanted to be sure that this counts as an ‘‘illness from
God,’’ that is, caused by God, but Kahenga refused to enter a philosophical
discussion; he just gave me another example of a woman catching a back-
ache from bad company she kept.
Can spirits cause illness (18)? Yes, they can, Kahenga said and gave an
example. The spirit of your deceased father may be angry because you failed
to make a required offering to him. He can punish you with sterility, for
instance. If, after consulting a diviner, the proper sacrifice is made, such a
person will be healed.
We were briefly sidetracked by a remark I made at the end of this
paragraph and got to discussing questions of race and ethnicity before we
returned to our topic, kinds of illness and their causes (20). Kahenga added
impotence to the infirmities he knew how to treat but I was now more
interested in summarizing what he had told me so far about causes. I
proposed a list and he confirmed: Illness can be caused by a bug ‘‘in the
 Kahenga’s Work

belly,’’ by a person’s evil intentions, by the spirit of an ancestor who ‘‘wants to

eat,’’ by God who punishes you for having wronged a fellow human being.
He may have sensed that I still had a problem with this last kind of cause
and gave another example but was so casual about it that I only now realize
its special significance: The wrong he cites is that of a thief who gloats over
his haul while the victim ‘‘decries’’ (kulia) the loss of his goods. Such ‘‘inside’’
lamenting (literally: saying something to yourself) is crying to God who
then punishes the thief by immobilizing him. People will say—and now
Kahenga identified the victim—that ‘‘this white man’’ prepared a dawa that
works as God’s punishment. With that Kahenga inserts the closing ritual, 65
the event that brought us together, into his etiology of diseases. This means
that the distinction between medicine and magic (perhaps sorcery) which I
made as soon as I began thinking about Kahenga, the healer, and Kahenga,
the performer of what I took to be a ritual of magic protection, does not, or
not accurately, reflect how he conceived his work.
Because I failed to appreciate the significance of his response at the time, I
missed the chance to have the matter clarified or deepened. Instead, I took
his mentioning a white person in his example as an occasion to continue
with the topic of causation. I asked him whether people think that Euro-
peans ‘‘bring disease’’ (21). Initially, his response was to deny this, possibly
because he did not want to implicate the European present (a conjecture
based on context and the ‘‘tone’’ of his response). But historical knowledge
made me insist and rephrase my question: Didn’t people think that way in
the old times? Now he agreed and told a complicated story of the arrival of
sleeping sickness in Hemba country during colonial times. This was one of
the many interludes in our conversation that will be taken up later (in
chapter 5). In the two shorter paragraphs that follow (22 and 23) we
returned once more to the kinds of diseases he treated before we took up a
new topic, Kahenga’s biography.

Kahenga’s Life and Training

We began with establishing Kahenga’s name and that of the village he came
from (24) but got sidetracked. He gave his full name later on as Kahenga
Mukunkole, adding that people at home sometimes called him by his
 Kahenga’s Work

Christian name, Michel (37).∏ The brief passage that follows my initial ques-
tion is a bit confusing, even for someone who is familiar with Luba naming
practices (as I was at the time we talked). Though family names have been in
use since colonial times (‘‘Tshombe’’ is a prominent example in Katanga; a
more recent one would be ‘‘Kabila,’’ president of the country and his adopted
son and successor), they did not displace precolonial customs. In the seven-
ties most individuals had Christian as well as African names, usually two,
that were publicly known. Beyond this basic fact matters get too compli-
cated to be discussed here.π Relevant to this passage is that a person usually
66 got one of his African names from a grandparent (male or female since few
names are gender-specific). Kahenga said that he got his from his grand-
mother and this was conforming to the rule. The term nkambo, which I
translated as grandmother, can also mean great-grandparent and any de-
ceased relative in ascending generations. The context suggests that the
nkambo mentioned by Kahenga was his mother’s mother. Her name was
Nyange ya Kahenga and she was called Kahenga after her grandparent.
When Kahenga spoke of his grandmother as mama, mother, as he almost
always did, he used a customary term of address for older females that
did not describe his actual kinship relation to her. In sum, establishing
Kahenga’s identity through his name placed him in a line of descent that was
also the line along which he had acquired his knowledge and skills as a
munganga ya miti (herbalist).
As we will see, his remarks about the mama who taught him still leave us
with vexing problems, but before we address them, here is a short curricu-
lum vitae based on Kahenga’s answers to my questions (25). He was born in
1940 in the village of Kihangu, near the Roman Catholic mission post and
town of Sola. His parents were peasants and he was their only child. They
sent him to the school run by the mission, which he attended until fifth
grade, graduating with a certificat. Much later (58) he added more informa-
tion: He was baptized, married, and had children who also went to Catholic
school at Sola mission. His house and lot, large enough to accommodate
many patients, were close to the residence of the missionaries with whom he
had good relations even though at the time we met he was excluded from
Holy Communion because he lived in a polygamous marriage.
But back to our conversation. After he had graduated in 1956, the mis-
 Kahenga’s Work

sionaries∫ selected him for continued secondary studies toward the priest-
hood at the seminary of Lusaka. But then his father died and his mama was
not to be persuaded by the missionaries to let Kahenga go (again, it is not
clear whether this referred to his mother or to his grandmother; I lean
toward the latter given his grandmother’s role in his life from then on). She
insisted that he had to stay away from the mission and so he went back to
his paternal village where he worked in the fields. It was at that time (he
was about sixteen) that he began his apprenticeship as a munganga with
his mama.
She took the center of our conversation at this point (27). Going on what 67
I had understood (or thought I had understood) earlier, I started out by
saying: ‘‘It was this mama who is dead who taught you?’’ Kahenga confirmed
this and that poses the problem I alluded to before. Elsewhere in the text or
in a conversation I remembered but did not record Kahenga had told me
that, if a dawa he had prepared or a treatment he had tried did not work, he
would go back home to consult with his mama. How is this to be inter-
preted? Did he mean that he could ‘‘ask’’ the spirit of his teacher or was it
simply that he did not correct the misunderstanding in my question and
that this person was still alive? As I mentioned when I discussed forms of
ellipsis in our exchange (in chapter 2), tense in Swahili verbs often is
ambiguous. All we can conclude from his way of talking about his mentor is
that she was ever-present in his work. The exact nature of that presence
remains a riddle, at least for the moment.
Further questions about his teacher steered our conversation toward
tutelary spirits and spirit associations, a topic to be commented on in a later
chapter. However, interspersed in these long passages, Kahenga offered
further information on his life and profession that should be noted now. In
the middle of paragraph 28 I tried to clarify terminology, asking him how
people back home referred to him (and his trade). He took a moment to
think about this and came up with an unexpected appellation (in Hemba):
munganga bwainaye. He translated this for me as ‘‘a munganga who has his
mother’s dawa, medicines’’ and went on to make this more precise by saying
that he would describe himself as a mufumu, an important person, who got
his bunganga, his profession of munganga, from his mama. Why the repeated
reference to his ‘‘mother’’? It could mean that the expression just describes
 Kahenga’s Work

how he got his training; more likely it should be read as foregrounding his
teacher’s dawa, thus locating his competence or efficacy in the medicines he
inherited, in substances rather than skills. The latter took the foreground
when I asked how people in Lubumbashi referred to, or called, him. He told
me that they consult him as specialist, as fundi ya dawa (literally: a craftsman
or specialist of medicines). In the city, the proper Swahili term of address
was munganga; mufumu was used only now and then.
Because I thought that mufumu, as a loanword from Luba in Katanga
Swahili could be a synonym for mulozi, sorcerer,Ω we began a lengthy dis-
68 cussion of sorcery and returned only later to biographical matters when
Kahenga described his teacher’s methods (33) and revealed that he also had
an apprentice, a relative of his (probably a nephew on his father’s side)
whom he was teaching his craft for a fee of ‘‘two, three goats’’ (35). Such
payment ‘‘in kind’’ rather than money is significant, given the fact that
Kahenga usually worked for cash. As in the case of bridewealth, at least part
of which must consist of gifts prescribed by tradition, the goats paid as
‘‘tuition’’ to a teacher of bunganga constituted a bond of obligations rather
than a commercial transaction.
Could he teach his craft to a white man (36)? Instead of answering
directly he deflected the question, first by telling me that it would be no
problem to show me how he collects his medicinal plants, then by begin-
ning to talk about one of them from among the samples before us. After
another aside (37), where we talked about the exact location of Sola, I
asked, as an afterthought, whether at the very beginning, when he was still
working with his teacher, he also visited clients outside his home country.
He did, as an emissary of his grandmother, and gave as an example the case
of a notable who wanted kulya busultani, literally: eat the office of chief (38;
see Fabian 1990b on this idiom and metaphor).

An Idiosyncratic munganga?

Keeping our attention on the text compels us to raise this question.∞≠

Kahenga’s idiosyncrasies were not a topic we discussed (it could not very
well have been the subject of direct questions) yet it came to the surface in
the form of two passages, marked ‘‘Spirits and spirit associations’’ in the
 Kahenga’s Work

outline given at the end of the introduction. They seem to intrude into our
exchange, interrupting the section on Kahenga’s life and training.
We had come to the point where he had to leave school and return to
farm work (26). He mentioned that this was also the time when he began to
watch his mama and teacher prepare medicines and to run errands for her. I
wanted to know more about this person and interrupted his autobiographic
narrative with a question that was on that semi-conscious agenda set by
prior ethnographic knowledge (27): Did his mother usually work alone or
did she have associates (literally: people like her)? I had in mind an associa-
tion (here and later we used the term nkundi, group) of banganga, which I 69
knew existed in Luba country. Kahenga denied this (at least for the time
being), perhaps because he thought that I was asking about helpers rather
than fellow members of an organization, only to affirm later that his mama
performed with a spirit (society) called bugembe. He himself put on a
performance when he described to me the nature of that relationship with
expressions that beautifully illustrate the meaning of ‘‘performing’’ culture.
He imitated the spirit’s voice as it comes out of the medium, her chanting,
perhaps a kind of glossolalia, and evoked her ecstatic body movements—all
this preparatory, as he then added in an everyday voice, to the moment
when ‘‘she would put on her work clothes and do her work for the people.’’
My next question was of course whether he had such a spirit too. No, he
did not and he refused to recognize one when his mama insisted that he also
would have to ‘‘put’’ (utaweka) his work together with ‘‘this spirit.’’ I was a
child then, he told me, was I going to get involved with all this dancing
around? No way, he added almost contemptuously. Nor did he inherit his
teacher’s spirit, as I tried to suggest. I am against (this), he said and, to make
his position perfectly clear, he added (though, linguistically, the statement is
cryptic and difficult to translate), ‘‘working with dawa, as I do now [is all I
want to do] . . . no spirit [for me].’’
We then spent the following paragraph (28) talking about different spirit-
associations in his home country. At the end, Kahenga reiterated his refusal
to join such groups; in fact, he admitted (with a chuckle) that he had an
aversion to the performances spirit-mediums put on. But the topic kept
troubling us and we took it up one more time, again interrupting the flow of
our conversation, in an attempt to get back on track after a diversion
 Kahenga’s Work

regarding his dealings with Katanga politicians (end of 33). I wanted to

know whether the societies we had named earlier had many members
nowadays (as they used to in the old times even though the whites were
against them). He confirmed this. Was there no pressure put on him to
join? There wasn’t. Was there no way he could work together with these
(‘‘organized’’) healers? Again, his answer was categorical: No, I just don’t
like it.
When it comes to placing Kahenga’s work in a context that includes
African as well as Western healing practices we should seek understanding
70 from information that our conversation disclosed. In the passages we just
looked at (and elsewhere), Kahenga spoke about his training as a person-
alized relationship with his mama, a close relative. He described the educa-
tion he received from her as a pragmatic process of gradual learning. He
never mentioned any kind of esoteric knowledge that would have required
initiation. At the same time, he acknowledged that through his teacher and
her membership in a spirit-association he was connected to the core of
‘‘traditional’’ bunganga. Even though we talked about these matters only
briefly, he showed that he was familiar with these societies, their rituals, and
techniques and that he often met banganga who were presumably much like
his teacher. His refusal to join their organizations must, therefore, be
interpreted as a deliberate ‘‘career choice.’’ It did not make him idiosyncratic;
on the contrary, by detaching himself from a tradition that would have been
difficult to keep up outside his rural home region, he became a respected
modern practitioner of bunganga.
Nor does the label ‘‘idiosyncratic’’ fit what he told me when we continued
on the topic of training. Kahenga himself now had an apprentice, a nephew
living in Lubumbashi, where he was a university student (35). Teaching him
the knowledge of plants required returning to the village, apparently not as
an ideological return to tradition but as a matter of practicality (or ecology)
because that was were he found most of the plants he needed and he knew
where and when to collect them to prepare dawa. When, toward the end of
that paragraph, I asked him whether he received payment from his appren-
tice his response was not what I expected. Yes, there was remuneration (see
above), however, it did not pass from his nephew to Kahenga but to
Kahenga’s mama, his own teacher.
 Kahenga’s Work

Conclusion

A picture of Kahenga’s work is beginning to take shape. Remarkable is, first,

his mobility. From his base at the village/mission of Sola some six hundred
kilometers north of Lubumbashi he had, within a few weeks prior to our
meeting, traveled to Lubumbashi, visited Kolwezi more than three hundred
kilometers to the west (3), and planned another sixty-to-seventy-kilometer
round-trip from Lubumbashi to Musoshi before presumably returning to
his home. We did not get around to discussing his travels in any detail; most
likely he used available means of transportation. Between Lubumbashi and 71
Kolwezi there was a railway connection; the other places could be reached
by riding minibuses or trucks. The region he covered included rural as well
as urban areas and a surface comparable to that of the Netherlands. To call
him an itinerant munganga, however, might be misleading. He did not travel
in order to find clients; his services were called upon by people who already
knew of him or of his reputation.
Second, the list of the kinds of Kahenga’s ‘‘clients’’ (a term we kept using
after I had introduced it at the very beginning) and their ‘‘problems’’ gives us
an idea of the scope of his activities. At the same time it can be read as a
Sittengemälde, a kind of exuberant genre painting peopled by men, women,
and children; villagers and townspeople; wives, husbands, and whores;
workers and big shots, politicians and chiefs; artists and criminals; Africans
and expatriates. True, soldiers, policemen, government clerks, priests, law-
yers, physicians, engineers, and businessmen were not mentioned as clients
but neither were they explicitly or implicitly excluded by any of Kahenga’s
statements.
To say that a review of Kahenga’s work results in a vivid picture of the
society and the historical period in which he practiced should not be
misunderstood. There was nothing ‘‘picturesque’’ about life in Katanga
during the mid-seventies though it may appear that way compared to the
abject misery and political chaos in the present. Kahenga’s services were in
demand because his clients experienced countless adverse situations and
afflictions that needed therapy or ‘‘fixing.’’
The question arises whether the work of the munganga was indeed as
pivotal as well as diagnostic (that is, socially revealing) as it appears from
 Kahenga’s Work

our document. Was it perhaps a symptom of the medicalization of physical

and social-economic problems of survival, not on the same scale but of the
same kind as those rampant processes of medicalizing human life from the
cradle to the grave that feed professional practitioners and providers of
‘‘care,’’ pharmaceutical industries, and insurance companies in the West?
Intuitively one shies away from such a conclusion and it may be possible to
reject it with arguments derived from political economy showing that the
practices involved are not comparable in kind or scale. But that would not
be what ethnographers can contribute to the debate. What we may be able
72 to do is something that anthropology at its best has been capable of: to
represent and confront alternative views, concepts, and practices, among
them ‘‘medicine,’’ for which our own societies claim universal significance
and validity. The task of another chapter, therefore, will be to comment on
what the text tells us about Kahenga’s ideas, his intellectual equipment that
allowed him to render the services for which he was consulted by his clients.
Before we get to that, however, we need to prepare the ground for con-
frontation by assembling, beyond what was already presented, information
on the historical context in which our meeting took place.
Kahenga’s World

4 When Kahenga and I met we discussed the

ritual he had performed, his profession, and
as much of his specialized knowledge as
could be covered in the course of our con-
versation. We shared a language and, at least
temporarily, had common interests that made commu-
nication possible. But it is also clear that, in the absence
of a prepared questionnaire, our conversation would not
have progressed beyond some sort of minimal gathering
of information, if it had not moved under its own power.
That power had among its sources memory and reflection,
stories and explanations, in short, comments on the world
in which the meeting took place, and that is why the text
has much to tell about its social-cultural, political, and
historical context.
 Kahenga’s World

Giving a coherent account of Kahenga’s ‘‘world’’ (an undertaking which

does not imply that Kahenga’s world is coherent, or that it must be coher-
ent to be intelligible) necessarily entails a certain shift from comments that
follow the course of our conversation to a series of synthetic presentations
organized around salient topics. Inevitably, this will make the section head-
ings of this chapter of our experiment look embarrassingly conventional.
On the other hand, just because there are good reasons for abandoning
the monograph as a dominant genre there should be no prohibition against
using some of its descriptive rubrics when this comes in handy in writing
74 ethnography as commentary—as long as the selection is inspired by, and
stays close to, the text. Realizing that a confrontation with Kahenga’s
thought, his ‘‘worldview,’’ had to be grounded in the ‘‘world he viewed’’
helped to overcome lingering literary scruples.

Village and City

Kahenga, the up-country practitioner of bunganga from Hemba land, was

consulted in Lubumbashi and asked to provide housing security, a thor-
oughly urban service. What he did, what we discussed, information we
exchanged, reflections we shared, explanations we considered—all of this
happened in a world that, if a propensity to align village and town with
tradition and modernity were accepted as a fact, was neither here nor there;
or neither now nor then. Still, our conversation gave us many openings for
discussing the ‘‘versus’’ between village and city, and we should now take
stock of relevant statements.
First of all, though this is easy to forget when attention is concentrated on
one ethnographic document of a casual conversation, we met, and Kahenga
operated, in a political context, the nation-state of Zaire, as it was then
called. I worked at the national university and was, at least indirectly, a
government employee; Kahenga, as his identity card showed, was in a trade
that had some sort of government approval. I don’t remember exactly when,
but he showed me the card. Apart from a photograph and the usual
information on date and place of birth it gave his profession as docteur du
bois, a phrase that makes little sense unless one recognizes it as a literal
 Kahenga’s World

translation of munganga ya miti. This must have fascinated me so much that

I forgot to note the office or organization that had issued the document.
On the whole the presence of a state and its institutions and agencies is
assumed rather than discussed. Occasionally, however, it comes to the
surface, to give one example, in paragraphs 37–38 (a section titled in the
outline ‘‘Kahenga teaching his craft’’). At the end of that passage I noted
that this part of our conversation, if measured by the ideal of a meaningful
exchange, could easily be mistaken as ‘‘textual trash,’’ as something one
might be tempted to clean out when presenting the text. But it paid to take
a second look. We were discussing the possibility that Kahenga might teach 75
a colleague of mine who was interested in medical anthropology. I assumed
that, if this were to happen, it would be in Hemba country rather than in
Lubumbashi. When I then asked for directions to his village, Kahenga at
first evaded an answer with a hint of impatience: Why this interrogation in
the midst of his telling me about his methods of teaching? I insisted and
asked for the name of the place where he lived: Sola, not really a ‘‘village’’ but
an agglomeration that had grown around an important Catholic mission
(population ca. 3,000 according to a recent estimate). I was starting to say,
so, that is where your mama, your teacher, lives. Without letting me finish
the sentence Kahenga informed me that she had her home in Moba. That
threw me off because I knew the name as that of a larger town on the shore
of Lake Tanganyika (it was called Baudoinville in colonial times). For some
reason—maybe he was just associating colonial and postcolonial names—
Kahenga brought up Albertville (now Kalemie). At any rate, the Moba he
talked about was just a village ‘‘back home.’’ So I made one more start, this
time with a question for the ‘‘region’’ or ‘‘zone’’ (two administrative entities
at the time) to which Sola belonged. He obliged but his response was vague.
Finally, I just asked for his postal address, which (resorting to an admin-
istrative gesture) I wrote down as we talked. While we were at it I wanted to
make sure that I had his own name down correctly. After another half-
hearted attempt at getting the exact location of Sola—it was twenty-two
kilometers from Kongolo (the capital of the region), he told me, or was it
thirty-two?—we were ready to abandon this line.
Our attempts to locate Kahenga’s home geographically and adminis-
 Kahenga’s World

tratively lacked exactitude; what they revealed about his world makes label-
ing him a ‘‘rural’’ practitioner inadequate, certainly if rural also means local,
traditional, backward, and so forth. Ever since his school days (when he first
faced the prospect of traveling) he had lived in a place that, through state
and local bureaucracy and the Catholic missions, was part of a regional and
national network of villages and towns, and when he moved between largely
rural northern and urban southern Katanga he was not a ‘‘migrant.’’
One is tempted to refer to such a situation as an urban-rural continuum,
and that does catch one of its characteristics, but it would be wrong to
76 equate continuity with homogeneity. Locations where people live may be-
long to a recognized geographical or political-administrative region (leaving
aside in this argument ethnic distinctions and divisions) but this does not
mean that such entities always create feelings of belonging together. The
situation in Katanga is too complex, its history has been too troubled, for
‘‘primordial ties’’ to be formed. For instance, when it comes to terminology,
there is ample linguistic evidence in our text of the reverberations of coloni-
zation and industrialization. In a long note to the English version (para-
graph 48, note 77) I gave a summary of usages prompted by an explanation
Kahenga had given of the purpose of one of the actions during the closing
ritual, saying that it was to give ‘‘coolness’’ to the people living in (this)
mukini. Mukini? I asked because one of the meanings of the term, ‘‘village,’’
was foremost on my mind and it seemed odd that he should refer to the
place the ritual was to protect—a lot in a large city—as a ‘‘village.’’ What
caused this double-take was not a linguistic problem. I knew that mukini
was also the general term for a ‘‘settled place’’ of any size.∞ As such it was
opposed to pori, ‘‘unsettled country,’’ ‘‘wilderness.’’ The problem was one of
those intrusions, noted earlier, of sociological preconceptions (expressed in
discursive habits) under which African ethnography has always labored, a
sort of axiomatic opposition between rural and urban going together with
and equating rural with traditional and urban with modern. After the fact, I
realize that this was on my mind, not Kahenga’s.
These observations are not to deny the fact, documented in our text, that
mukini can also have the restricted meaning in expressions such as kule ku
mukini, (back) there in the village (e.g., 35), ku mukini: kwetu, in the village
back home (12). Added to this should be kule kwetu, there back home, and
 Kahenga’s World

other circumlocutions that suggest a sense of ‘‘home’’ not as in opposition

to, but as part of, Kahenga’s larger world.
There was one striking passage early on in our conversation (7) that is
worthy of closer examination as an example of confrontation between the
ethnographer’s expectations and the interlocutor’s views. We had talked
about one of the problems Kahenga routinely ‘‘arranged’’ with the help of
dawa: labor conflicts in an urban-industrial setting. I told him that I was a
little surprised by this. Is what you are doing not the work of a munganga?
Yes, it is. But, I continued (without actually saying ‘‘but’’), this is work you
learned from your ancestors. Yes, it is. But, and now I said alafu, the 77
ancestors lived in a village. Kahenga grants that, too. They didn’t know, I
continued, about labor problems like getting fired or not promoted, right?
At that point he must have run out of patience with my line of questioning
or, rather, arguing, and stopped me dead with a laconic ‘‘no.’’ Appealing to
my intelligence or imagination he reminds me that the basic issue he deals
with as a munganga—in a factory as well as in a village—is trouble among
people. In both cases, dawa can be prepared to resolve such problems. I
wanted to be sure and asked again: As you see it, then, there is no difference
between a factory and a village? Of course not, he replied, laughing, people
in a factory are people. Still clinging to my preconception, I took what I now
recognize as anthropology’s last stand in a situation like this: I gave a
temporal twist to the opposition between village and factory by placing the
village in the past (zamani). A few years later, in Time and the Other, I
criticized this as temporal distancing, a device we use in our discourse to
place our interlocutors in a time other than ours. This example from our
text shows how discursive habits, that is, the ways we write ethnography
‘‘later,’’ intrude in our conversations ‘‘in the field.’’ This is how conventions of
writing may influence how we formulate our questions and what we think
we want to find out.
When Kahenga reacted with a noncommittal ‘‘mm’’ to my invoking the
past, I tried another approach (8). Zamani, in the old times, elders and chief
had sought dawa (understood but not said: to sustain their power). As soon
as Kahenga had confirmed this I switched to the present (‘‘now, here and
now’’) and asked whether he, too, had clients among the bakubwa, the big
shots? Yes he did have such clients, not here (in Lubumbashi) but ‘‘in the
 Kahenga’s World

village,’’ patiently explaining to me that his expertise would be called upon

by persons who wanted to maintain their position or were about to kulya
busultani, ‘‘eat the office of chief.’’ Clinging to my allochronic perspective, I
failed to appreciate that he had brought up the village in response to my
querying him about ‘‘here and now,’’ that is, where and when the village was
for him. I took him through a series of further questions intended to clarify
whether bakubwa ya sasa, modern big shots, sought his services. No, he had
‘‘not yet’’ been consulted by such persons, he told me, but, of course, he
knew that they often went ‘‘to the bush’’ in search of dawa. At any rate, he
78 insisted that the problems chiefs in the village had to resolve were of the
same kind as those posed by a domestic conflict or those the heads of a
big business or factory and the supervisors or foremen in the mining com-
pany may have in maintaining their authority. In all these cases dawa may
be needed.

Kinship and Family

Generally, an allochronic perspective has been operative in setting village

against town, tradition (then) against modernity (now). Among the more
specific ways to maintain such a perspective has been anthropology’s pre-
occupation with ‘‘kinship.’’ Therefore, another expectation the ethnogra-
pher brings to the presentation of a munganga’s world is that he lives, like
everybody else in ‘‘traditional society,’’ in a network of family and kinship
relations that structure interaction and determine status. At a first glance,
our text confirms that expectation. But in several respects, Kahenga’s kin-
ship relations or, at any rate, what he chose to tell me about them (explicitly
but also by implication) do not fit the typical picture one may have of a
traditional situation.
First some of the basics: In the passages that contain information on his
biography (24, 25), Kahenga mentioned his immediate family, his father
and mother but no siblings. This was highly unusual but I let it go until we
came to a point where the episode about the missionaries trying to recruit
him as a seminarian made him mention that he ‘‘was born alone.’’ To make
sure that I had this right I asked again whether his mother had not given
birth to other children. No, she had no other children. Not only that, she
 Kahenga’s World

herself was an only child. About his own family we did not talk at any
length. At the time, Kahenga was thirty-four, lived in a polygamous mar-
riage, and had children who attended the same mission school in Sola he
had gone to when he was young (54). All this remained general and even the
few facts just mentioned leave questions open: Who were his relatives by
marriage and where did they live? With links established through two wives
there may have been many. How many children did he have and what were
their ages?
Apart from gaps of information that would be interesting to fill now but
were not talked about then, there are other problems with reconstructing 79
Kahenga’s kinship relations. Kin were mentioned frequently, but in the text
before us the exact relations are often not clear because the terms he used
are polysemic and remained ambiguous even when Kahenga offered de-
scriptions of a relationship. Take my attempts to understand how exactly he
was related to his apprentice and the young man who assisted in the ritual
(35). In fact, looking at this passage now I am not even absolutely sure that
we were talking about two different persons. When I asked Kahenga
whether he had an apprentice his answer was yes, my muyomba, a reciprocal
term that designates the relation between a person and his or her mother’s
or father’s brother.≤ I translated it as ‘‘my nephew.’’ But how could he have
had a nephew—in the common understanding of the English term—if he
was an only child? Assuming that the relationship was reckoned through an
ascending generation, it would have to be on his father’s side, since his
mother was also an only child. In that case, the young man was his father’s
brother’s (i.e., Kahenga’s muyomba’s) child and that would make him his
cousin rather than nephew. Confusion deepened when we moved on to his
young assistant. He was the muloko, usually ‘‘younger sibling,’’ of ‘‘the mama
of that mama who gave me the medicine,’’ that is, his teacher. If at all, this is
translatable only if the first mama does not mean ‘‘mother’’ but simply
female relative; hence all we get to know about the assistant is that he was a
relation on ‘‘mother’s side.’’
Who exactly was ‘‘mother,’’ the appellation Kahenga always used for his
teacher? In the biographic passage (24) we clarified this: she was the person
who gave birth to Kahenga’s mother, therefore his grandmother, nkambo
(see also 33, 35). This was my term, not his, and it could have remained an
 Kahenga’s World

outsider’s gloss if it were not for the fact that it served to trace Kahenga’s
name-line to the nkambo of his nkambo, a grandparent of his grandmother
(24), a common practice (see above). In our conversation nkambo occurred
most of the time in the plural, bankambo (and once in the collective, man-
kambo) and referred to ‘‘ancestors’’ generally, not to specific relatives. When
we later talked about prayers that were recited during the closing ritual and
I asked Kahenga to whom he prayed, his answer was baba yangu, my father
(41). Did he mean his own deceased father? Impossible to tell because a
little later he mentions baba together with nkambo and muzimu (spirit,
80 especially of a deceased person) as addressees of prayers (see also 51 where
baba and muzimu are used interchangeably).
What is the point of dwelling on ill-understood detail regarding kinship
relations when all this seems to do is to produce more questions and little
that would help us to give this part of Kahenga’s world a distinctive shape?
The answer could be twofold: Our exercise demonstrates limits of com-
mentary as ethnography. The vexing ambiguities that kinship terms leave in
our text are, as it were, artifacts of documentation. The text is all we have
before us now. Thirty years ago it would have been possible to remove
ambiguity by asking Kahenga to be more precise. But that would have
worked only to a certain degree, which brings us to the second part of the
answer: The point of commenting on terminological ambiguity was that it
helps to appreciate the fluidity of social relations, including kinship, the one
category that has often been regarded the most manifest and fixed of social
relations, so much so that studies of kinship terms could be mistaken,
perhaps with the connivance of their authors, for studies of kinship relations.
Kin were important to Kahenga; even in the limited space of our conversa-
tion he revealed how firmly linked his personal and professional identity, his
name and his dawa, were to his ‘‘mother.’’ Family obligations at one point
may have changed the course of his life when he had to forgo secondary
education. He gave no indication that he felt constrained or limited in his
mobility by kinship obligations. What I knew about Luba society could
have led me to inquire about other levels of kin-relations, such as clan and
lineage. But this never came up and it is idle to speculate how to interpret
absence of reference to clan or lineage in this text.
 Kahenga’s World

Race, Ethnicity, Politics

Moving on to another aspect of Kahenga’s world, what does our text tell us
regarding the importance of race, ethnicity, and politics in his work as
munganga? The question does not aim at, and what will be discussed does
not provide, anything near to a comprehensive account of these dimensions
in the Zaire of the mid-seventies. This would be the task of a monograph.
Commentary works the other way around and we should direct attention to
passages in our conversation in which we talked about these matters, which
we did rarely and mostly indirectly. At least that is how I put it to myself 81
when I tried to get a grip on the task. Among the difficulties I faced right
away was one that is posed also by other topics treated in this chapter:
What do statements regarding race, ethnicity, and politics tell us about
Kahenga’s world and what about his thought? What may count in our
exchange as more or less reliable factual information about a political situa-
tion and what was expressed as ideas and reflections, his and mine, on the
topics of this section?
Again, we may begin with some basics, starting with race, a term I used as
a heading of notes made during work on the text. As I confront it now I am
not even sure why I included it in the list that makes up the heading of this
section. Nowhere in our conversation did we use expressions, either in
French or in Swahili, unequivocally signifying race.≥ I can find only two
words that could be classed as expressions with obvious racial connotations.
One of them, mweusi, a color term meaning black, occurs a single time (!) in
the phrase muntu mweusi, a black person (53). I was the one to bring the
adjective into our conversation when I wanted to be sure whether Kahenga’s
reference to pères, fathers or priests, included Africans. In contrast, muzungu,
plural w/bazungu, usually translated as ‘‘white person’’ or ‘‘European,’’ ap-
pears more than thirty times. No question, ‘‘Whites’’ were often referred to
but does this mean that race was a topic?
I must admit that this seemingly straightforward question—a question
one ought to, after examining the text, be able to answer with a yes or no—
stopped me dead in the tracks of this commentary. Attempts to come up
with a statement left me vacillating. On the one hand, I should like to, and
 Kahenga’s World

could, argue that neither Kahenga nor I thought of muzungu as a racial

distinction based on skin color or other physical features.∂ On the other
hand, there is the fact—and here I can speak only for myself—that I did
experience my dealings with Kahenga as ‘‘transgression’’ (in a sense I ex-
plained in the introduction) and I have reasons to believe that the closing
ritual and our conversation are likely to be perceived as extraordinary
because, more than in routine research, boundaries were crossed when the
European ethnographer became an African munganga’s client. Was there not
a racial or racist frisson involved? And if so, was it any different from the
82 one that has always affected anthropology as a discipline? Not much more
comes to mind that would help us to return from introspection to Ka-
henga’s world except a gratuitous conjecture: Inasmuch as race in the sense
of racist thought and behavior (and there is no other sense, given the
impossibility to come up with a defensible definition of race) was present in
the social context in which our meeting took place it must have affected
interaction somehow.
What we can do is take a closer look at occurrences of muzungu/wazungu
in our text. These terms came up in our exchange for two reasons. One was
simply to designate, or ‘‘name,’’ expatriates as a category of actors. For
instance, Kahenga mentioned among clients who might consult him some-
one who wants to work for muzungu, an expatriate employer (6). In the
prayers that were part of the ritual of closing the house, huyu muzungu, this
white man, referred to me, the client who was to be protected (41, 49, 50).
Once he called a European missionary muzungu (54); the Belgian colonizers
were wazungu (21) and the expression dawa ya wazungu (72) meant West-
ern drugs.
Not quite as often, coming to the second reason, the term was introduced
into the conversation by my questions whose purpose was (with one excep-
tion, 54) not to identify a person but to steer the discussion toward a theme
that was clearly more on my mind than on his: distinctions and differences
between Africans and expatriates as patients or clients (18). I wanted to
know to what extent Kahenga’s world and the one I represented were
separate. A result of this line of questioning was that my expectations (not
consciously formulated but somehow implicit in our relationship) were not
confirmed. Not only could he teach a muzungu his knowledge of herbal
 Kahenga’s World

medicine (36) but wazungu could, and did, cause harm with the help of dawa
(20, 21) and, most surprising of all (but a matter to be taken up in the next
chapter), among the mizimu, spirits of the deceased who could be helpful as
well as harmful, there were also wazungu (19, 59). Only in one instance did
Kahenga use a form of the word that made it deprecatory. This happened
when he told me that an illness could be God’s punishment for a wrong
done to another person and gave an example: Someone meets a European
and insults him (here Kahenga slipped into the role of that person): Hooo:
angaria kile kimuzungu kile, something like ‘‘Ahh, look at this piece of mu-
zungu’’ (17).∑ Kahenga left no doubt that such expressions of contempt were 83
contemptible.
Earlier I said that neither I nor Kahenga employed a term signifying race.
A careful reader could point to a passage that seems to contradict this. The
context was our discussion of causes of illness, in this case mizimu, spirits,
who can make a person sick and have to be appeased with the sacrifice of a
chicken or goat. It must have been my compulsion to look for difference
that made me ask whether wazungu could be afflicted with this kind of
illness. Actually I stated my question in the form of a conclusion, as if it
followed logically that African ancestor-spirits could only be harmful to
Africans. Kahenga rejected this immediately. No, you people can catch this
type of illness; this ‘‘has nothing to do with kabila’’ (18). This could be
translated as ‘‘nothing to do with race’’ but many other occurrences of kabila
show that nowhere else it was used in this specific sense. I did not give up
my pursuit of difference and pointed out that wazungu did not make sacri-
fices to spirits (implying: how can they have the illness but not the cure?).
Kahenga did not accept this, telling me in effect that I should not generalize
because among all people, including wazungu, there were different kabila,
kinds. I had this nonspecific sense in mind when I began to interrogate him
about kinds of illness and used kabila, a local all-purpose term for differ-
ence.∏ I must have sensed or anticipated that Kahenga would find this
quaint because I immediately added ‘‘as there is a kabila of people,’’ which he
accepted with a chuckle. Being queried about ‘‘tribes’’ of illness (tribe is the
dominant connotation of kabila) amused him (16).
Not much later kabila in the specific meaning of ethnic identity—to use a
neutral sociological gloss—became a topic when my ideas and expectations
 Kahenga’s World

once again made me ask an ‘‘ethnographic’’ question (19). I reminded Ka-

henga that he had learned his trade in the village, hence in his kabila, which
was Hemba. But didn’t he take on clients from ‘‘all tribes’’? I expressed
surprise, pointing out that throughout the country people usually ‘‘follow
their kabila’’ in these matters. He agreed. So, why do they seek dawa,
medicine, without regard for ethnic differences?π Kahenga saw no contra-
diction and explained to me that the world of the mizimu, the spirits, as
agents of illness as well as healing, was not divided by ethnic boundaries.
We did touch on ethnic distinctions in this paragraph (our example was the
84 Kongo from lower Zaire) and later on, first when we compared Hemba and
Luba words for God (51) and then when he identified a divining gourd I
showed to him as Luba (70). In neither case did Kahenga give the slightest
indication of using ethnic labels in an evaluative, much less discriminatory,
manner. This was remarkable, given the long history of colonial invention
and promotion of ethnicity (through the ‘‘scientific’’ and administrative uses
of ‘‘tribe’’), a history that was no less real for being invented as shown in
violent outbreaks and endemic presence of ethnic strife, especially in the
postcolonial recent past. Bukabila or tribalisme were constantly talked about
in local political discourse and ethnicity was a central issue in analyses of the
very situation in which our encounter took place (for instance, Young and
Turner 1985). The politics of the day—Mobutu’s regime, the bureaucrats,
police, and soldiers Kahenga undoubtedly had to deal with in his frequent
travels—are not mentioned in the text but they were of course present in his
world and the one we shared.

Religion: The Catholic Mission

When one searches the document for statements about institutions and
organizations that affected Kahenga’s life and work most directly, then one
finds that it was the Catholic mission. He was baptized, educated in a
mission school, and his (first) marriage was Catholic. His children, presum-
ably also baptized, attended the mission school, which was convenient
because he had his home and his ‘‘practice’’ in a settlement attached to the
mission post of Sola. Among the most interesting and lively parts of our
conversation were Kahenga’s accounts of his relations with the missionaries;
 Kahenga’s World

interesting, because they confounded notions I may have had regarding the
incompatibility of bunganga and Christianity; lively, because Kahenga en-
joyed baffling me with the rapport he enjoyed with some of the priests.
Several passages are relevant to this aspect of Kahenga’s world and they
become rather intriguing, especially when we follow the rule that, as much
as possible, the course of our conversation should determine the sequence of
comments. Remembering that rule let me notice something that would
have otherwise been easy to overlook: Mission and missionaries (bapères;
later also missionnaires or (ba)padri) first came up when we talked about
Kahenga’s life history (25). In our later, more detailed discussion of the topic 85
it was not his daily, mundane, and ‘‘modern’’ dealings with the fathers at
Sola that brought up the mission but a ‘‘theological’’ passage in our con-
versation (51). It began with my asking him about the Hemba term for
God, vilinyambi (Mungu in Swahili). Once again, ethnological foreknowledge
made me steer the conversation toward a topos in the anthropology of
African religion, the relative remoteness of God and the proximity of spirits
as addressees of prayer and sacrifice. When I suggested that people may
know God but that he is ‘‘far away,’’ Kahenga contradicted me, amused by
my misconception. Yes, it may be true that prayers are said and offerings are
made to ancestor spirits, but who created them and is God not the sole
source of baraka, blessings? To make sure that I understood the meaning of
baraka, Kahenga brought up a concrete example, the importance of mayi
ya baraka, holy water, in his work as a munganga. It was a powerful sub-
stance ‘‘because it had been blessed’’ and he would get it from the mission-
aries (54).
Without making him state the connection, this brought him to the
person who did the blessing of holy water, the (Catholic) priest (52).
Kahenga spontaneously placed him in the frame we had just constructed in
our theological discussion of God, ancestor spirits, prayer, and sacrifice and
defined him as ‘‘a person who is always really close to the dead (mufu).’’ I
reacted to the singular form he used (mufu, not bafu) by asking ‘‘which
dead?’’ He did not let himself be distracted by a minor point and continued
to make the one that was important to him: It was the priest’s closeness to
the deceased, in fact his ability to see a dead person ‘‘openly (waziwazi)’’
that gave him indirect access to God and made his prayers efficacious. Then
 Kahenga’s World

I interrupted his explanation with an image I had stored away in my

memory. I knew of the habit of priests to say their breviary while walking
up and down in the cemetery that was often close to the church and rectory.
I brought this up and Kahenga immediately knew what I was talking about.
People had seen missionaries in the cemetery, reading from a book (perhaps
moving their lips), and took this ritual recitation of prayers as speaking
to the dead, something that he himself could not do, as he told me, be-
cause it took learning from ‘‘many books.’’ Such learning from books also
gave the priests strength to control (‘‘tie up with a rope’’) dangerous spirits
86 of the dead. Talking to the dead ‘‘all the time’’ gets them close to God, who
answers their prayers most of the time, and that is ‘‘why we need the priests
back home.’’
Kahenga’s remarks in this episode tell us much about the presence of the
missionaries in his world. They were more than just representatives of an
institution whose foreign, colonial nature could not have been lost on the
people. With some imagination, the villagers had found a place for the
missionaries’ rituals and book-learning in a frame of thought that made
sense to Kahenga and his likes.
But there was also the mundane presence of the missionaries in everyday
life at Sola. Kahenga had his house and clinic next door to the mission
church (54). Were they not against his work? On the contrary, they would
come to him with people they had been unable to cure; conversely, he would
ask them for help when he had problems making his dawa work. Even the
fact that he was excluded from communion because of his polygamous
marriage did not trouble the good ‘‘professional’’ relations he had with the
priests. They might even bring a sick fellow expatriate (mugonjwa ya kwabo)
and ask Kahenga to prepare a dawa, which he would do and the person
would be cured. His laconic response to my c’est vrai—is that true?—was:
‘‘(When they ask for it) are you going to refuse?’’
I wanted to know more about this surprising collegial relationship be-
tween two worlds, Christianity and Western medicine on one side and
traditional bunganga on the other, which, at that time, I still could not think
of otherwise than as being in contradiction and conflict. How was it pos-
sible to keep communications open so that missionaries would like to talk
to him the way I did in our conversation? Kahenga’s response made me
 Kahenga’s World

realize that I was not the first to show ethnographic interest in his work
(adding another aspect to ‘‘late ethnography’’—missionaries had often pre-
ceded anthropologists). He told me that he met regularly with a Father
Joseph, the Superior of the mission. They would talk at length and the
priest would interrogate him about his work, take notes, even give him
some money,∫ and encourage him to continue as long as Kahenga called on
God to make his bizimba work (like ours, their conversations would be in
Swahili). One day, he then told me (55), a traveling party of missionaries
even filmed him and his patients.
Given his good relations with the missionaries, Kahenga had many Chris- 87
tians among his clients (56). He mentioned some women who were station-
ary patients at his house and he probably would have given other examples
when, taking a guess based on what I knew about Catholicism in Katanga, I
asked him whether the movement known as Jamaa existed in his place.Ω It
did and had many followers at Sola. Not only that, Kahenga and his wife
had been members or candidates (only married couples could join the
Jamaa) until he disqualified himself by taking a second wife. He then went
on to tell me about conflict between two groups, a typical situation that had
the ring of authenticity. Knowing that the movement’s doctrine prohibited
recourse to ‘‘pagan’’ practices, I asked him whether they did not reject the use
of dawa. They did, with the exception of strictly herbal medicine (dawa ya
mizizi) for which he was frequently consulted by them although they would
insist that he did not put ‘‘anything else’’ into his preparations.

Landscape and Memory

There is one passage, relevant to the topic of this chapter (21), at which I
should have stopped for comment earlier, only to skip it repeatedly because
I found it difficult to decide whether it should be commented on in this
chapter on Kahenga’s ‘‘world’’ or in the next one about his ‘‘thoughts.’’ That
distinction between world and thought, which at first seemed a convenient
and elegant way of ordering what I wanted to present, has by now become a
burden. The problems with maintaining it reflect of course a tension be-
tween the generic demands and constraints of a monograph and those
of a commentary. Facing the recalcitrant paragraph now will not resolve
 Kahenga’s World

my problems of presentation but it will make it possible to add another

facet to Kahenga’s complex ‘‘world’’ and prepare a transition to the chapter
that follows.
Problems of presentation are not the only ones posed by this passage.
Between the extraordinary story Kahenga had to tell and my eagerness to
understand what I heard, the conversation turned into a groping exchange,
full of multiple starts, repeated questions, and laborious attempts to get
matters clarified—with limited success, going on the challenge this part
turned out to be when it came to transcription and translation. As we will
88 see, in the end, finding a place for this episode and determining what
Kahenga had actually said came down to taking two decisions, one topical,
the other linguistic.
We had discussed matters of race and ethnicity (17–20) when it occurred
to me to press Kahenga on a question we had barely touched on: Did people
here think that wazungu bring disease? He set out to deny this when, with a
single-mindedness I now find embarrassing, I reformulated my question
without letting him finish the sentence: But was this not what people
thought in the olden times? With that I changed the focus from a theoreti-
cal reflection on causation to memory and history. Kahenga immediately
rose to this and began to tell a story that must have been very much alive in
popular memory. It began simply enough when he answered my question
about Europeans bringing disease. Yes, that is what people thought, and the
gist of what followed was that the wazungu ‘‘of old’’ (‘‘1918’’ was his guess) had
brought sleeping sickness to the region of Kasongo.∞≠ Starting with the first
sentence, the text contains surprising elements and some vexing ambigu-
ities. It began when Kahenga, in a move that made colonial memories
uncomfortably present, asserted something about Europeans and referred
to them, not as wazungu but as ‘‘you:’’ It was you who ‘‘locked up’’ disease, he
began to tell me, then he hesitated and paused as if searching for a better
beginning until he got his narrative on the way.
The going, however, was not to be smooth. What was meant by kufunga
maladi, locking up disease? In Kahenga’s story, local presence of the disease
was somehow tied to a certain large hill or mountain near Kasongo that had
been an abode of spirits. It was there, I understood, that the illness was
deposited as a dawa. Considering that the issue was whether whites had
 Kahenga’s World

caused illness, did his answer mean that they had brought sleeping sickness
as a dawa by enclosing it rather than locking it up (and thereby controlling it)
in that hill and that it spread from there? Much like the question that
started the discussion, this interpretation was suggested by colonial history.
Connections between colonization and the spread of diseases (involving not
just importation but transition from endemic to epidemic occurrence) have
been a subject of research and debate in which sleeping sickness had a
prominent place (see, for instance, Lyons 1992).
My understanding of history, such as it was, had followed the wrong trail.
This became clear only a sentence later when Kahenga removed ambiguity: 89
It was not the disease that was ‘‘locked up,’’ as I had been led to believe by
the elliptic phrase kufunga maladi, but a dawa was deposited and somehow
sealed up: ‘‘They placed a dawa and locked it up’’ and ‘‘there was no more
sleeping sickness.’’ Then (he said sasa, now) the Whites left and—was it that
the seal was broken (a chuckle indicates that this may be speculation)?—
there was a lot of sleeping sickness there in (the region of) Kasongo. I was
still confused, so I tried again: There was this huge (kabambi) mountain and
when the Whites came, what were they looking for? Whatever they were
looking for, he answered, they found that many people were dying of
sleeping sickness. So then they buried (banazika) their dawa (against/for?
the disease) after having talked at length with the guardians of the spirit
who lived on that mountain. I wanted to know the mountain’s name but
Kahenga did not remember, except that it was near Kasongo. I must have
imagined that the Europeans were associated with the mountain because
they had built a house (an observation post?) on its summit. There was no
house, Kahenga told me, in fact, he began to suspect that I did not know the
meaning of kilima, mountain. After this was cleared up I wanted to know
whether one could see the place where the dawa had been deposited. They
(the Europeans) used to go there often, he told me, but we avoided the spot
out of fear and if you go near it there is a wind blowing that will make you
catch sleeping sickness.
Then we talked about the return of the disease after the end of Belgian
colonization. The Whites had been ‘‘chased away’’ and their tricks (mayele)
no longer worked. By now we had been through the story several times but I
was still not ready to let go. Didn’t the Whites use vaccination (meaning:
 Kahenga’s World

rather than dawa) to fight disease? Kahenga refused to follow me along that
line. Although he undoubtedly had witnessed many vaccination campaigns
he stuck to the story of the locked-up medicine. Above all he stuck to a story,
to a shared memory of the colonial past that had a place in a prominent
topological feature. Memory, as often is the case, was embodied in a land-
scape, one that was inhabited by powerful mizimu, spirits, to whom the
Whites owed their success in fighting sleeping sickness.
It hard to resist an allegorical interpretation of the tale: There are the
Whites going for the mountain peak to work their power by burying their
90 dawa in African soil. Their cleverness had sleeping sickness ‘‘locked up.’’
And there came the evil wind of postcolonization blowing from that moun-
tain, making people sick. But indulging in allegories was not what Kahenga
had in mind. At any rate, an attempt at allegorical reading runs into trouble
with an account that remains ambiguous to the end. On the basis of the text
alone, it is impossible, for instance, to say whether the placing of dawa was
said to have happened on, in, or at a certain huge mountain that may have
been just a big hill. And did the Whites control a disease that plagued the
country or had they brought it in the first place, which had been my
question in the beginning.
If Kahenga was not engaged in allegorizing what was his intention and
what the significance of the story? To me, it was the tie between past history
and present landscape, or between memory and ecology, that revealed a
dimension of what we have been trying to catch as ‘‘his world.’’ All it took
was a change from third to second person (not ‘‘the Europeans,’’ but ‘‘you’’)
to make the disquieting point that the European ethnographer was already
part of a world he may have imagined as being ‘‘out there,’’ belonging to his
interlocutor.
 Kahenga’s Thought

‘‘Belief in Belief ’’

5 As soon as I had decided on the title of this

chapter I had second thoughts. It may seem
reasonable, though not mandated by the suc-
cession of topics in our text, to proceed from
Kahenga’s practice to the ‘‘theory’’ that guides
his work. However, the underlying assumption—that the
two can be distinguished or even kept separate—is ques-
tionable. Above all, presenting work and thought in suc-
cession may give the wrong impression that Kahenga’s
practice merely implemented his thought and knowledge
—as if he had worked with a set of ideas that, once ac-
quired or accepted, did not need constant work to be
workable. Of course, to lump practical skills such as the
recognition of medicinal plants, the ability to match illness
and cure, and to decide on a treatment, as well as views
 Kahenga’s Thought

regarding causes and agents, judgments concerning good and evil, and
ontological concepts under one common heading would mean resorting to a
time-honored strategy. It usually consisted of calling, and thereby bracket-
ing all of the above as, belief (with variants such as belief system, worldview,
doctrine, ideology).
The effect of such bracketing, at one time considered a theoretical achieve-
ment but questioned more recently,∞ has been to constitute certain kinds of
thought and knowledge (among them religion and magic) as objects of
inquiry by opposing them to scientific thought and ‘‘Western reason.’’ This
92 may have kept them in an arena of rational debate; it also made it unneces-
sary (in fact, impossible) to confront religion, magic, and other ‘‘beliefs’’ at
eye level. As someone put it in a formulation whose source I have been
unable to trace, anthropology was in the business of showing ‘‘why we know
and they believe.’’
That was not all. The function of this strategy (to use a neutral term and
avoid getting entangled, at this point, in having to justify terms such as
‘‘purpose’’ or ‘‘consequences’’) was not only to set other kinds of thought and
knowledge apart from the ones we call scientific but also to establish intel-
lectual control of what had been set apart and thereby constituted as an
object. Following the maxim divide et impera, establishing dominion required
‘‘dividing,’’ that is, making and defending distinctions within the domain of
belief(s), above all the one between religion (itself divided as high or world
religion vs. primitive or tribal religion) and magic (usually also divided as
white vs. black, benevolent vs. noxious magic, with categories such as sor-
cery and witchcraft covering much of the latter). At one time, sense was
made of these distinctions by placing them at different stages of evolution.
Then, under the paradigm of functionalism, more or less the same categori-
zations were approached as socially ‘‘functional,’’ institutional differentia-
tions, and not much changed when functionalism was hyphenated with
structuralism. Science, religion, magic, sorcery, and witchcraft were thought
of as different configurations of symbols that, in opposition or contrast to
other such configurations, constituted a cultural system of beliefs and con-
comitant practices. More recently the Foucauldian notion of discursive
practices, again applied to all of the above, seems to have become the
bracketing device du jour (though it may be doubted that Foucault himself
 Kahenga’s Thought

intended his radical historizations of sex, punishment, and other regimes to

serve as epistemological sedatives).
Such has been anthropology’s ‘‘belief in belief,’’≤ deployed as a strategy for
establishing intellectual control (also called explanation) of scientific objects
classed as religion, magic, or sorcery. This almost-caricature of the anthro-
pology of religion and magic is not to be taken as a gratuitous dismissal of
the search for reason in the apparently irrational (often it had considerable
success≥). If I still think that the great theoretical treatises and detailed
ethnographies our predecessors produced provide little guidance for a proj-
ect to write ethnography based on a present document of past communica- 93
tive interaction, this is due to the failure or refusal to recognize the contem-
poraneity of their objects of study. Such failure may not have been a
necessary result of using terms like religion, magic, witchcraft, or sorcery
(anthropologists did not invent them and we are as yet unable to do entirely
without them) but it was certainly fostered by making these labels technical
terms and then turning them into rhetorical devices of an allochronic
discourse that kept our interlocutors’ practices and thoughts at a safe
distance by placing them in a time other than ours with the help of theories
of evolution, change, or modernization.
This brief reminder of the history of anthropology’s road toward ‘‘belief in
belief ’’ shows why it is difficult to avoid approaching Kahenga’s thought and
knowledge without designating what he thinks and knows as ‘‘beliefs’’ and
pressing what he says and does into preconceived categories. Are there
alternatives that allow us to confront his views and assertions? Of course, no
ethnographer should delude himself into being able to start completely
afresh. Still, I have come to think that writing ethnography as commentary
makes it possible, if not to get rid of our conceptual arsenal then at least to
keep it in abeyance often and long enough to make what I call confrontation
productive. An immediate consequence of such a position (or project) is
that we lose some of our most cherished certainties, among them the
ontological distinction between a real world and an imagined, thought-up,
or postulated world. But is it not precisely such a distinction that makes me
present Kahenga’s ‘‘world’’ and his ‘‘thought’’ in separate chapters?
Be that as it may, critique of imposed categories should not make us blind
to categorizations if we encounter them in our documents. Even the most
 Kahenga’s Thought

resolute refusal to press what we get to know about bunganga into familiar
categories does not exempt the ethnographic commentator from respecting
and presenting categorizations our interlocutors make when they tell us
what they think and know.

Thought and Knowledge

Kahenga and I met at a time when I was engaged in projects of eth-

nographic research that had not started out but eventually came together as
94 inquiry into ‘‘popular culture.’’ In most of the expressions I had studied—
popular religion, historiography, painting, and theater—a common denomi-
nator was ‘‘thought’’ (kuwaza, to think, and mawazo, thoughts). In the Jamaa
movement, for instance, thought was, as a kind of gnosis, the pivot of their
teachings. The painter-historian Tshibumba insisted that a historian thought
the past.∂ When I searched our text for -wazo/waza and related terms the
results were striking.∑
The noun mawazo does not occur at all and, with one exception, only I
used the verb kuwaza four times, twice in questions about what ‘‘people
thought’’ (21) and twice when I asked Kahenga to think in the sense of
remember (4, 6). Such shifting of meaning from thinking to remembering
and back is exemplified in the only instance where Kahenga employs the
verb kuwaza.∏ We were coming to the end of our conversation and as one of
several afterthoughts I asked him how common knowledge of plant names
was among people in the village. They knew all the names, he told me, and if
a person had forgotten one he or she would just ‘‘put it (the plant) there and
think’’ (73). When I transcribed this passage I made a note to myself to look
more closely at this semantic constellation of thinking and remembering in
Kahenga’s answer to a question about knowledge.
We will get to knowledge presently but first an observation on another
conspicuous absence in this text: In Swahili, as in European languages, the
phrase (mi)nawaza, ‘‘I think,’’ seldom occurs in the marked sense of ‘‘I
cogitate.’’ Most often it introduces statements in the sense that make it
synonymous with ‘‘it is my opinion,’’ ‘‘I am not sure but . . . ,’’ in short, ‘‘I
believe.’’ But I don’t recall ever having come across a case in conversations I
had through the years, many of which were intensely searching and reflex-
 Kahenga’s Thought

ive, where a speaker’s ‘‘I believe’’ was a statement of ‘‘belief ’’ in the sense
discussed the introductory section above, that is, in contrast or opposition
to ‘‘knowledge.’’ The Swahili verb -sadiki, to believe (never used in our
conversation), always means ‘‘to believe in’’ and is used only in religious
discourse; it never means ‘‘I believe that.’’ Kahenga did not qualify any of his
statements as convictions. While this does not mean that he had no convic-
tions, it is expressive of his altogether factual attitude to the matters we
discussed.
If terms for thinking and believing are conspicuously absent from the
transcript, the contrary can be said about the verb that signifies knowing. It 95
has two alternate forms in Katanga Swahili, jua and -yua with the former
counting as more refined (kiswahili bora). I don’t recall ever hearing the
nouns mjuzi or maarifa, knowledge, listed in dictionaries (though a search of
other texts on our web site might prove me wrong). Acquired knowledge,
learning or education, is called elimu, but that term is also absent from
our text.
First, an interesting linguistic detail that showed up when I looked at the
frequency and distribution of occurrences. I used the verb twenty-six times,
with one exception always in the -jua variant; Kahenga employs it thirty-five
times, always in the -yua form. Remarkable about this is that the choice of
single phonological variant (an allophone in technical terms), of which I
was probably not conscious at the time and now only discovered almost by
accident, appears with such regularity as a distinctive feature marking ques-
tions in contrast to answers.π
Since our conversation was in a mode of inquiry it is not surprising that
the topic of knowledge was brought up most often in the questions I asked,
always with verb phrases, and that Kahenga’s answers frequently repeated or
echoed my use of -yua—lexically. Semantically, a reading of this text with a
focus on expressions relating to knowledge and on their contexts produces a
picture of great complexity. Here is, first, an inventory of the verb yua and its
many shades of meaning. Kahenga asserts that he ‘‘knows that’’ or is con-
scious of (17, 30); that he ‘‘knows whether’’ something is the case (38); he is
‘‘known as’’ (37); he knows the function of an object (70); he knows God (51)
and the practices of divination (71). To know can mean to recognize (18, 19)
or to experience/feel (20). Another cluster of significations is formed when
 Kahenga’s Thought

he uses -yua in a diagnostic, technical, professional sense of to recognize,

identify, and to ‘‘know how to’’ (6, 9, 15, 16, 17, 23, 29). To know can refer to
the ability to identify a person as a sorcerer (29, 30, 31). His apprentice
knows what he has been taught (36) and Kahenga knows names of plants
‘‘well’’ by writing them down (33, see also 35), something that is different
from the general knowledge of plant names among people of his country
(73).∫ Finally, -yua may refer to linguistic competence; one knows (the
meaning of) a word (21, 76) and one knows a language such as Swahili (29)
or Hemba (54).
96

Knowledge and Power

Among phrases that belong to this cluster I found one, mwenye kujua dawa, a
person who knows dawa (29), that invites further exploration. In the same
context Kahenga calls such a person fundi ya dawa, a craftsman or specialist
of dawa. Most often -enye is used to form an attribute of a person or to make
of a verb phrase the equivalent of a noun (mwenye kujua could be translated
as ‘‘a knower’’). But it may also connote regular possession or, more appro-
priately in the case under consideration, proven competence with an under-
tone of power. It was this hint of a pragmatic conception of knowledge that
made me attentive to the verb –weza, whose connotations can range from
liberty or ability (an equivalent of the English ‘‘I can’’) to capability, mastery,
and power. I found two examples of compound phrases combining -weza
and yua. In the first one Kahenga said of his teacher that she anaweza kuyua a
sorcerer (29), which clearly says more than ‘‘she can know’’; he told me that
she had the competence, or power, to identify such a person. Not much
later (31), when I asked him whether he, too, ‘‘can know’’ the work (kazi) of
sorcery—meaning: could he practice sorcery?—he responded (twice) with ‘‘I
can know’’ only to negate this apparent admission by making it clear that he
did not accept such an insinuation and assuring me that he did not like to
exercise his power back home in situations when ‘‘the whole village’’ was
out to find a sorcerer, though he would use it in a case such as mine when
my house may be threatened by ‘‘such people’’ (bale benyewe) who seek to
harm us.
The next step in this exploration of ‘‘knowledge’’ was to search the text for
 Kahenga’s Thought

occurrences of -weza, which again revealed a wealth of connotations and

produced some insights. One of them emerges when we line up uses of the
phrase naweza kufanya, literally: I can do or make. Kahenga used this expres-
sion, sometimes without an object, when we talked about different kinds of
illness/problems he was able to treat (4, 5, 7, 11, 16). What is remarkable
about this is that he presents his knowledge not as a taxonomy of concepts/
terms but as a sort of catalogue of his competences. As far as I can see he
never responded to my inquiries about kinds of diseases with only a term or
phrase referring to a condition; he always got to talk about treating it: He can
(naweza) prepare medicine or dispense it (20, 54, 58), heal a person (19, 20), 97
mend conditions, such as sterility (4, 22), and, as he reminded me, close a
house (13). The negative form siwezi also occurs (15, 33, 34, 56, 72) but only
in the sense of I am not allowed to, I can do nothing about it, it is not
possible that; none of them negates what the meaning of -weza we are
exploring here asserts.
Concentrating on -weza one begins to realize that in Kahenga’s thought
power as the ability to treat afflictions is closely related, to put this cau-
tiously, to the power to cause them. We touched on this in the preceding
chapter (in the sections on race and religion) when we considered the
possibility that God, who according to Kahenga is the ultimate source of
the munganga’s power to heal, may also cause illness. He has the power
(anaweza) to ‘‘return, or turn on you, the wrong you committed’’ and ‘‘make
you pay for it’’ by afflicting you with ‘‘this illness’’ (17).
Not only God has this power; mizimu, spirits, whom one may implore
(19) and who can see to it that a person is richly rewarded (52), may
also bring illness (18). A spirit is also able (anaweza) to travel to another
part of the country (19). Catholic priests can see the spirits of the de-
ceased (anaweza kumuona) and converse with them. They get this ‘‘strength’’
(nguvu), which Kahenga says he does not possess, from having ‘‘learned
many books’’—a rather surprising attribution of such powers to literate
knowledge. Above all, priests have the power to ‘‘chain’’ evil spirits and that,
he adds, is why we need them (52).
Almost as an afterthought it occurred to me probe the text for mayele, a
word that is common in Katanga SwahiliΩ and had become part of local
French, especially as spoken by long-time European residents who may have
 Kahenga’s Thought

picked it up from the pidginized Swahili many spoke as a work- and

command-language. Mayele, a noun, has connotations that can approach
that of akili, intelligence (no occurrence in our text), for instance in the
phrase muntu wa mayele, a clever person. (I remember my old friend and
mentor Kalundi exlaiming mayele ya bazungu!—White man’s cleverness—as
we were negotiating a maze of freeway crossings on the outskirts of Brus-
sels.) Mayele can denote resourcefulness in general but most often it means
something like a trick, the solution to a problem, or a specific device used in
such a solution.
98 Kahenga used mayele on five occasions. When I was interrogating him
about his work in ‘‘labor relations’’ my question implied that bunganga
belonged to knowledge bankambo, his ancestors, and had developed in a
village context. How could it work in a modern factory? He rejected the
implication categorically. They had knowledge of such matters banayua
sababu tu ni mayele, best paraphrased as ‘‘because it is simply a matter of
finding solutions,’’ for instance, to conflicts between people, and such prob-
lems are the same in a village and factory (7). A similar argument emerged
when we discussed the role of the munganga in assisting rulers and politi-
cians. He must have sensed that I was going to register this as being limited
to politics and the public sphere when he offered an example that ‘‘hit
home’’: You see, here in your house you don’t get along with the wife (he
knew more than I did at the time). How are you going to go about this? You
must find a mayele so that you get along with the wife (8). Later we talked
about mizimu and their ability to work their mayele beyond regional and
ethnic boundaries. A white person’s ancestral spirit could follow you to
Africa and make you well. The same goes for people from the Kongo ethnic
group in the lower Congo who now live in Katanga (19). And when the
Whites managed to ‘‘lock up’’ sleeping sickness (see the preceding chapter)
they could do this because they had brought ‘‘that mayele of theirs’’ along.
These exercises in epistemic archaeology in the sense Foucault (1973) gave
to ‘‘archeology’’—collecting the results of searches and using (commenting
on) the findings to reconstruct Kahenga’s conception of knowledge—do not
make easy reading. Nor do they succeed in presenting Kahenga’s thought as
a coherent system or ‘‘theory,’’ but they show a metalevel in his thinking.
Evidence for this is not as explicit and pronounced as in other kinds of local
 Kahenga’s Thought

discourse where terms like kuwaza and mawazo, thinking and thoughts/
ideas, are deployed in a reflexive manner.∞≠ This does not mean that Ka-
henga was not capable of such thought but that we concentrated in our
conversation on his work as a munganga.

Thought and Knowledge in Action

We can now go on and direct our attention to statements in which Kahenga

articulated contents, objects, or uses (rather than forms) of knowledge and
their ties to cultural practices and institutions. Our aim will remain not to 99
bracket his ideas as ‘‘beliefs’’ even when, or especially when, some of the
things he told me may seem beyond belief. Come to think of it, has being
‘‘beyond belief ’’ not always been what qualified certain kinds of ideas or
thoughts as (mere) beliefs? Writing ethnography in the genre of commen-
tary takes away the pressure of constantly having to gauge the truth value of
Kahenga’s assertions or of evading this by attributing social functions or
symbolic significance to actions that do not easily fit our own habits of
thought. Interpretive commentary goes a long way (but of course not all the
way) in enabling the ethnographer to respond with a ‘‘no comment’’ to
demands for ‘‘explanation.’’

fansia: doing medicine

What exactly was Kahenga’s ‘‘work?’’ Healing, curing, treating, therapy?
More than a dozen times he referred to the effect of what he did as -pona, get
well; only once does he use this verb in the causative form -ponyesha, make
well, cure, and that in an indirect statement reporting what people say a
charm (kizimba) can do (54). Much of what the text can tell us about how
he called and presumably conceptualized his work is expressed by the verb
-fanya whose definition in the dictionary is ‘‘cause to do, cause to be use-
ful or of avail, hence make. One of the commonest verbs in Swahili, al-
ways implying some result, purpose, or object beyond mere act.’’ I counted
eighty-three occurrences of -fanya in Kahenga’s speech, many of them rele-
vant to the present topic (for example -fanya dawa, make, prepare a medi-
cine), a daunting wealth of information to comment on. Fortunately a
derivative form he also used, -fansia,∞∞ literally: ‘‘to cause to make for,’’ allows
 Kahenga’s Thought

us to get a better grip on the semantics of ‘‘making’’ people well. Admittedly,

‘‘to cause to make for’’ (causative and applicative) does not look promising at
first—until one realizes that it packs a kind of double-barreled sense of
causation. Examples will show that our own commonsense understanding
of medical treatment or therapy does not adequately describe the work of a
munganga.
When a client consults Kahenga he or she simply says unifansie, literally:
you should cause to make for me, without specifying an object or a ‘‘what,’’
and, as he says, niko nafansia, I do it (2, similar forms: 6, 8, 11, 16). In other
100 occurrences an object is named: a dawa (16, 67) or nguvu, an effort, a strength
or power to get well (19, 41, 56) or simply bantu, people, elliptic for ‘‘making
something for people’’ (27, 33, 71). In sum, when Kahenga expresses what he
does as –fansia he inserts his activity into a chain of causation that he not so
much sets in motion but directs toward a purpose or a person, a kind of
causation that is multiply mediated, especially if we remember what we
found out earlier and will comment on again presently about the role of
dawa and spirits.

dawa, miti, bizimba: materiality

As far as can be told from our conversation, Kahenga did all of his work
as a munganga with the help of dawa. The dictionary defines the term as
‘‘medicine, medicament, anything supplied by a doctor including ‘charm,
talisman &c,’ used by native medicine men.’’ Often pronounced lawa, it is in
Katanga Swahili also a general term for chemical products such as addi-
tives, solvents, lubricants, dyes, and the like. Taking into account these
extensions (of which Kahenga was of course aware), the basic meaning of
the term, therefore, may be said to be that of a material substance used as a
means to obtain certain results. ‘‘Used’’ presupposes a user, hence an agent;
substances become ‘‘means’’ when they are employed by persons. It has
become clear from Kahenga’s statements commented on earlier that he
thought of his work as involving other agents: God and mizimu, neither of
whom were ever said to provide or prepare dawa. ‘‘Materiality,’’ it seems, is a
characteristic of human agency.
Let us now look at what Kahenga told me about his knowledge of
substances and their applications. As a munganga ya miti he worked with
 Kahenga’s Thought

vegetal matters—roots, bark, leaves, seeds, and flowers—that he collected

himself in the ‘‘bush’’ (pori), as land that is neither settled nor cultivated is
called. To become ingredients of dawa these materials are usually processed;
they are pounded, ground up (13, 62), mixed (33, 40), and boiled or grilled
for external or internal use. Dawa are applied as ointments, given as enemas,
drunk as potions or infusions (4, 60), or eaten (23). Knowledge of plants
entails more than that of their active ingredients; a munganga must know
where to find and when to collect them. There are some he can take along
wherever he goes but others do not travel (33, 36). Nowhere in our conversa-
tion did Kahenga give so much as a hint that he thinks of his dawa as 101
commodities that can be sold and circulate; they are always part of the
services for which he is paid.
While Kahenga did not think of dawa as merchandise, some of his
remarks indicate that they could be regarded as a possession. When I asked
how people in the village called him, he said they would talk about him as
‘‘the one who has the dawa of his mama [his teacher]’’ (29). In a prayer to
mama Nyange he asked for strength and for the dawa to work with because
‘‘you gave me your dawa, I did not steal it’’ (41). In other words, he claims
legitimate ownership—but of what exactly? Dawa, as we have seen so far, is a
substance and substances come in kinds or portions, hence the plural
madawa. Remarkable about the instances just cited is that the term is used
in the singular and the context suggests a translation as the, not a medicine.
Is this a figure of speech in the reference to skills he learned from his mama
or was there something, perhaps some kind of material token (other than
an actual medicine), that was passed on along a line of transmission of
knowledge?
Kahenga called himself munganga wa miti and we had first consulted him
as a herbalist healer. This encouraged me to call ‘‘medicine’’ the substances
he used as well as the trade he plied. It was an inescapable categorization,
predestined and reinforced by the image of the ‘‘native medicine man’’ that is
deeply engrained in our collective popular imagination and does not simply
disappear when we relabel such practices in our scientific discourse as
traditional herbal medicine or therapy and call their practitioners healers.
At times Kahenga himself made statements that seemed to indicate that he
conceptualized miti as a clearly defined and separate domain. For instance,
 Kahenga’s Thought

when we reconstructed the closing of the house (40) I began asking him
about the holes he had prepared and about the ‘‘things’’ deposited there.
Guessing what I had in mind he interrupted me: [You think it was] bizimba,
charms? Not at all, ilikuwa tu miti, it was nothing but vegetal matter (see
also 9).
Later I came back to the question of ingredients in his dawa and he
responded again with a categorical ‘‘just plant matter,’’ adding, ‘‘I take away
from all those things we eat the one that is a dawa (59). I kept prodding him
with one of my ‘‘informed’’ questions: How about hair (which, like cuttings
102 of nails, is a well-known ingredient of charms)? He said nothing about hair
at that point (later he denied using it) but conceded that he would take the
saliva of a person to be mixed with soil from his house, presumably in a
healing ritual.
Another occasion to present himself as an herbalist came when I asked
toward the end of our conversation what he does when he gets sick back
home (72). I go get my miti, was his answer. Don’t you go to the dispensaire, I
pursued (every mission has a place where Western medication can bought
or is handed out). He didn’t but his children would sometimes. What for, I
continued to ask, he had dawa, could it be that sometimes his medicine
failed him? Not often, he assured me, but it could happen that he had to
resort to Western medication. He reminded me that I had given him some
quinine (here used as a general term for pills) when he had a bout of diarrhea
but no miti available.
Early in our conversation we had an exchange that added yet another
dimension to the complex meanings of dawa (9). We had concluded that the
ones he prepared consisted only of miti when he came up with an aside I
cannot translate exactly though the general sense seems to be clear enough.
Its background must have been that in Katanga Swahili it is often said that
the clients of a munganga look for bizimba. Originally this is a Luba term,
signifying generically, and usually translated in French as, fétiche. Kahenga
rejected this and was at pains to make it clear to me that this designation
was a misconception: He worked, as he had said in another context, only
with miti. ‘‘Miti are the buzima [life, health, literally: wholeness] of man.’’∞≤
We came back to bizimba when I asked Kahenga about his apprenticeship,
trying to get him to tell me as much detail as possible (33). He described how
 Kahenga’s Thought

he would accompany his grandmother when she went into the bush to
collect all sorts of miti and would write down the names she told him. I asked
whether this meant he had a ‘‘book’’ of dawa back home. He confirmed this
and then continued with a vivid account of the way they worked together.
His teacher, it turned out, was blind and Kahenga described how she would
sniff the samples he brought to her, take some time to reflect, and then name
them. But why did he, just after we had referred to them as miti and dawa,
suddenly call them bizimba? He introduced the term in a parenthesis—‘‘these
things they call bizimba’’—and that suggests that in this instance he may have
given a Hemba appellation whose meaning differs from that of the loanword 103
bizimba in local Swahili (more on this later).∞≥ In other words, discrepancies
in Kahenga’s use of bizimba do not necessarily reflect confusion or inconsis-
tency; they may simply be due to the fact that Kahenga ‘‘quoted’’ them from
different languages and communicative practices.
The passage that followed, incidentally, added another facet to the per-
sonality of his teacher. Here he told me that his grandmother had diffi-
culties walking. But it turned out that this was not the reason why she
refused to travel to this part of the country. It had nothing to do with her
infirmities but was due to prohibitions (bizila) imposed on her by her
tutelary spirit.
Bizimba showed up again when we began to discuss the closing ritual (40)
and Kahenga insisted that things he had brought along to be buried on our
lot were not bizimba but miti. When he used the word for a last time during
the conversation—he told me about answering a missionary’s ‘‘ethnographic’’
questions—it was again in its local Swahili meaning, that is, a ‘‘charm,’’ in
contrast to dawa (54).
Bizimba is a loan word from Luba. The dictionary by Van Avermaet and
Mbuya has one of those long entries that make this work an ethnographic
gold mine (1954: 824). For our purposes it may be summarized as follows:
Kizimba refers to any human, animal, or vegetal substance that may be used
to prepare a bwanga, the Luba equivalent of Katanga Swahili dawa, medicine
or charm. ‘‘The bizimba∞∂ constitute the element, the essential ingredient,
the ‘magic’ substance (‘‘produit’’) of bwanga; it is the kilumbu not the munganga
who procures them; one is convinced that by using these manga one appro-
priates for oneself the vital force of the being whose bizimba one possesses.’’∞∑
 Kahenga’s Thought

Could it be that there is a ‘‘temporal’’ distinction between dawa and

bizimba in that the former (the ‘‘medicine’’) faces the past, or a given state,
diagnosed as already existing whereas the latter (the ‘‘charm’’) is directed at
something to happen or to be accomplished? They seem to have in common
that they are addressed to ‘‘problems’’—but can this be said of good-luck
bizimba? Or is the notion of fortune, the common element in misfortune and
good fortune, part of some deep philosophy of the fortuitousness of every-
thing that happens?
Be that as it may, Kahenga’s ways with terminology should not be inter-
104 preted as operating fixed taxonomies but as context-specific acts of commu-
nication. The task of this chapter has been to extract from our document a
reasonably complete account of Kahenga’s thought as it was communicated
in our conversation. A lesson to be learned from the preceding and other
lexical excursions is that in Kahenga’s mind, as well as for many of his
clients, traditional and modern, rural and urban resources of reasoning are
co-present; they interpenetrate, or interact with, each other. The question
whether or not they merge to form something like a ‘‘symbolic system’’
cannot be answered on the basis of information given by Kahenga. Still,
that they inform a coherent practice, or practices, can hardly be doubted.
Of course, one might ask now, if Kahenga’s thought is as little coherent as
it appears to be at times, how can it guide a coherent practice? Two ways of
responding come to mind: First, in the examples that brought us to this
question his statements are incoherent only by strictly logical criteria (such
as those applied in the construction of taxonomies). Very little is incoherent
rhetorically in what Kahenga revealed about his thought and knowledge.
Second, in these attempts to understand Kahenga’s work and thought I let
myself be guided by insights gained from a long-standing preoccupation
with the critique of culturalism (a position that assumes that culture orients
action ‘‘as a system’’). In my view, culture as practice never simply ‘‘enacts’’ or
‘‘reflects’’ a system of beliefs or symbols. It consists of habits of acting, and
‘‘acting’’ entails working things out by matching intellectual resources with
practical tasks in ways that are not systematic or necessary but historically
contingent. That, I am convinced, applies even to the ‘‘methods’’ and ‘‘ritu-
als’’ in Kahenga’s bunganga. Routine is an aspect of practice, not its essence.
Already during the ‘‘making’’ of the text I noted down an observation that
 Kahenga’s Thought

bears on the question of coherence. It regards the fragmentary nature of

ethnographic information in this account. What we learn from the text is
fragmentary with respect to the urban context and even more so to the
Hemba background of Kahenga’s work. But what does it mean to recognize
a piece of information as a fragment? Fragmentary as opposed to what? To
the holistic ambitions for which anthropology prided itself in its ‘‘modern’’
phase and that became a target of critique more recently? If fragmentariness
is a characteristic of the object of inquiry (‘‘patchwork’’ is the metaphor en
vogue), how can ethnographic information be other than fragmentary? Not
only this, one could also point to specific, text-internal, topical fragmenta- 105
tion caused by the many turns and starts in our conversation.
When one sets aside, for a moment, holistic ambitions as well injunctions
against holism and thinks about the concept of fragment itself, it begins to
lose its negative aura. Fragments may be seen as the rubble that results from
the destruction of a whole; unless one can put the pieces together again
fragments remain meaningless. But ‘‘reconstruction’’ (what archaeologists
and paleontologists do with pot shards and bones) is not a good metaphor
for ethnography if one has given up holistic ideas. Reified or essentialist
holism should not be confused with dialectical approaches positing that
knowledge is produced and should be presented in a field of tension be-
tween particulars and a totality.

mizimu: ‘‘spirituality’’
After confronting challenges posed by the concepts of dawa and bizimba
we can now resume our commentary on the text and turn to another key
concept in Kahenga’s thought: mizimu, spirits. In the conversation this
notion came up more than once, first when I asked whether mizimu could
cause illness (18). Kahenga did not give a general answer to this general
question. Repeating the term in the singular and as a question (muzimu?) he
offered an example. The spirit of your (deceased) father may be angry and
send you an illness because you neglected to make required offerings to him.
You should then consult a diviner about the cause of your affliction and he
may tell you that the trouble was caused by a certain spirit, in this case your
father’s. You would then make the offering (‘‘cook’’ a goat or chicken) and be
cured. This example was followed by a passage about the ‘‘mobility’’ of
 Kahenga’s Thought

ancestor spirits across ethnic boundaries (commented on earlier). When

Kahenga later recited the prayer he had said during the closing ritual it
turned out that it was addressed to an ancestor of his, also referred to a
muzimu wetu, our spirit (41).
Mizumu were again mentioned when we continued with the list of dif-
ferent kinds of disease and their causation (20). Did people think that
Whites cause illness (21)? I had asked this with some vague knowledge of
connections between colonization and the spread of diseases. As we saw in
the preceding chapter, Kahenga did not want to speculate about this in
106 general and responded with a ‘‘case,’’ the story of how the Whites in his
country dealt with sleeping sickness (see above), in which a muzimu of a
different kind plays a role. This spirit lives in, or on, the mountain and he
or she is approached for help through bamizimu, people who are spirit-
attendants (see also an oblique reference to sorcerers who work with mizimu
in paragraph 31). Given the association with a feature of landscape and the
allusion to a cult, such a muzimu would probably be classified as a nature-
spirit. However, to Kahenga, it appears, this kind of spirit is of interest as a
historical rather than ‘‘natural’’ agent.
It was when we talked about Kahenga’s biography and apprenticeship
that we came to the first of two sequences I titled ‘‘Spirits and spirit
associations’’ (27–28, 34–35) in the outline of our text. They look like
intrusions into topics we were discussing and they could be regarded as
interludes, mere asides, until one realizes that they are highlights in Ka-
henga’s thinking. In the first passage he revealed that his teacher was not
only a munganga but a medium, ‘‘possessed’’ by a muzimu, a tutelary spirit (to
use an ethnographic label). He first called it bubira and then, correcting
himself, bugembe (27). It was to bugembe that she owed her (knowledge of)
dawa. Upon further questioning it turned out that bugembe was not her
personal spirit and that his teacher was a member of one of several spirit
associations practicing possession and performing dances. Others he named
were butembo, bulungu, bumbudi, nyambe, all of them ‘‘large groups’’ (28).
We came back to mizimu when we discussed the prayers Kahenga offered
during the closing of the house (41). They were in Hemba and I could not
understand them but I thought I recognized invocations of ancestors by
 Kahenga’s Thought

their names. I mentioned this and Kahenga confirmed my hunch and gave a
list of names, first those of his deceased father, Mukenge Mbuyi, and of his
mama Nyange. Then he named Kayembe, ‘‘our chief,’’ and Yagamino ‘‘our
big spirit.’’ I take the latter to refer to yet another category of mizimu, not his
personal ancestors but ‘‘ours,’’ that is, the collective spirits of his country. We
talked at some length about Yagamino, whom Kahenga described first as a
spirit who has his abode in a prominent rock (much like the spirit-on-the-
mountain in the story about the Whites and sleeping sickness) and then, in
political terms, as ‘‘the chief of all our spirits back home’’ (42). Thinking that
we had ascended in the hierarchy of spirits to a supreme being I asked 107
whether this Yagomino was what is called a vidye in Luba. That did not get
us very far; Kahenga knew the term gave but gave me the impression that it
was not used in Hemba. We had to stop the recording briefly at this point
and when we got back to it—he had had a moment to ponder my question—
we only caught part of his next statement, ending with ‘‘our vidye.’’ I wanted
to make sure that this ‘‘chief of spirits’’ was not Mungu, God. Certainly not,
was his reply, Mungu was Mungu. Yet, he then cited a short prayer to God in
Hemba in which he did use vidye in what sounded like a translation or
paraphrase of Mungu. It is possible, of course, that he just wanted to please
me since I had kept asking him about vidye (not an uncommon thing to
happen to ethnographers).
When we returned to our exchange after another short pause (I was
checking the cassette recorder) Kahenga clarified the issue: Yagamino was a
‘‘regional’’ spirit. Every inchi, country or region, in Hemba land had one and
he named two others, Muhona and Mulamba. Yagomino’s and Kayembe’s
territory (I had used territoire in my question) was called Nkuvu (43).
Finally, tucked away in another discussion of the name of God and easy to
overlook, Kahenga made a statement containing just the barest hint to yet
another kind of spirit: You pray to a muzimu, you address the prayer to a
muntu, short for human ancestor, or to a nyama, an animal (50). I let this
allusion to possible totemic ideas go without further questioning because I
still had the Hemba name for God on my mind, which turned out to be
vilinyambi.∞∏ In the theological discussion that followed (51) he made sure I
understood that making offerings and praying to mizimu did not exclude
 Kahenga’s Thought

Mungu (or relegate God to a remote place as deus otiosus, the technical term
used by historians of religion). After all, who had created the mizimu? The
power of mizimu was ‘‘God’s power.’’∞π
Given my aim to write ethnography in the genre of commentary, the
presentation of materiality and ‘‘spirituality’’ in Kahenga’s thought turned
out embarrassingly ‘‘monographic.’’ We encountered this problem earlier
and I can only reiterate that the choice I made is not a (re)lapse but the
inevitable outcome of working within a frame of tensions between modes of
representation. However, we should be able to do better than that. Another
108 look at both themes, dawa/bizimba and mizimu, can bring our commentary
back ‘‘on track’’ by reminding us to approach the text as the document of
a communicative event, not (only) as a depository of information. This
means that we should pay attention not only to what Kahenga imparts but
also to how he chose to present himself. More than once he insisted that, far
from simply enacting what he had learned about bunganga from his teacher
and Hemba ‘‘tradition,’’ he had made decisions and taken positions when he
formed his professional identity. Decisions and positions are historically
mediated (or context-specific) acts.
We spent some time discussing his choice to define himself as a munganga
ya miti, an herbal specialist. Admittedly, the result was ambiguous, reflecting
the complex meanings of each of the three concepts, dawa, miti, bizimba, as
well as the equally complex semantic relations between them. Eventually it
emerged that Kahenga had taken his distance from employing nonherbal
substances or objects and especially from getting drawn into affairs of bulozi,
sorcery. Though he provided protection he would, other than his teacher,
not get involved in finding out agents of bulozi (31).

kivuli: material spirituality

When I came to the end of the preceding section I discovered yet another
way in which the monograph may encroach on commentary: it may blot out
entire passages of the text. This happened when I almost forgot to com-
ment on one of the most intriguing exchanges we had during our conversa-
tion. I did report earlier on a statement of Kahenga’s regarding the use of
substances other than miti, such as human hair, but then neglected to
address what he told me in the remainder of that paragraph (59) because it
 Kahenga’s Thought

did not fit the schema of opposing ‘‘materiality’’ to ‘‘spirituality’’ I followed in

the two preceding sections.
Kahenga had told me that he did not ‘‘cut hair’’ to be used as an ingredient
of dawa but that he would ‘‘carry off ’’ a person’s kivuri. I was not familiar
with the word and it took a demonstration (Kahenga got up from the table
and took a few steps) before I caught its meaning: kivuri is your shadow that
appears to sit still until it follows you when you move. In asides to myself I
translated this to a Swahili word I did know, (n)giza, and then into French as
ombre, shadow or shade. Kahenga let this go without comment and gave as
another example the shadow cast by a tree. ‘‘That is what you carry away, its 109
shadow.’’ Of course, my next question had to be: How does one do this? His
answer began with ‘‘It is not just to carry,’’ which could mean ‘‘not literally
carry,’’ but then he elaborated: What one can take away, literally, is soil from
the ground on which a shadow has fallen.∞∫ The locative phrase mu bulongo,
on or in the soil, not only locates the shadow, it also puts it ‘‘inside’’ a
substance that can be picked up. When I repeated the phrase (trying to
understand it), Kahenga offered an explanation that took me by surprise.
Literally translated, he said: ‘‘A person must die: he/she changes into dust.’’
‘‘And the shadow . . .’’ I began to ask, whereupon he completed my question
or, rather, made it a statement, saying ‘‘it (the dust) makes it stay on.’’
Kahenga offered yet another ‘‘example’’ that must have left me baffled
then and continues to perplex me now. The elliptic story he told, here
further condensed, had me as a ‘‘dead man walking,’’ a muzungu, killed by
sorcerers in a car accident, *something they accomplished by carrying away,
not just the visible person but my vuli. He had given the key to understand-
ing this when he stated: ‘‘uzima wa muntu ni kivuri, the life of a person is
(his/her) shadow.’’ Fifty paragraphs earlier he had told me: ‘‘miti njo buzima
ya muntu,’’ the life of a person is miti, here best (but not adequately) trans-
lated as ‘‘plants’’ (9). What is one to make of the equivalence of shadow and
plant in these statements? Both are predicated on (b)uzima, which would
not pose a logical problem if we interpret shadow and plant as representa-
tions, as signifiers, as symbols, or as figures of speech. They could be
metaphors of uzima, life (a plant is ‘‘alive’’; the ‘‘tree of life’’ comes to mind),
or metonyms (a living person, his or her shadow, the ground on which the
shadow falls, and soil that can be collected from that ground would be links
 Kahenga’s Thought

in a chain of connections). Given the specific context in which Kahenga

made his statements—practices of healing and ‘‘magic’’ and the materiality
of both—the latter, metonymy, would seem more likely. But would this
mean that healing and magic are but figures of speech? Is to bespeak persons
and problems all a munganga does?∞Ω
This would hardly be satisfying as a comment on the specific statements
we are trying to understand. They were not pronounced (or recited) as
examples of magical formulae but as attempts on Kahenga’s part to help me
understand why or how he thought healing works. He formulated premises
110 of his reasoning. When that is recognized, it turns out that, in the phrase
uzima wa muntu ni kivuli/miti, it is not the predicates, kivuli and miti, shadow
and plant, that challenge understanding but the seemingly familiar subject,
(b)uzima, life, health. As far as I can see, nothing was said indicating that
(b)uzima wa muntu, the health/wholeness of a human being, could be ‘‘con-
tained,’’ like a material substance, in a plant or a person’s shadow. On the
other hand, miti and kivuli (much as dawa and bizimba) are not mere sym-
bols, perhaps metaphors (plants for vitality, shadow for wholeness). It is
almost as if thinking as practiced in bunganga took the inverse direction of
reasoning that ascends from sensual, material experience to the realm of
ideas. Bunganga works by objectifying, materializing thought; to prepare
dawa is to give to afflictions, conditions, or events, conceptualized as threat-
ening wholeness, a material presence.
Endings and Ends

6
One would think that to end a commentary
should be easy; you just stop when you feel
you have exhausted the text—or yourself.∞
A general conclusion is not needed because
there was no general argument and for a re-
sounding ending we lack a single, absorbing story. What is
it, then, that makes such an easy way out all but incon-
ceivable? It must be that, contrary to what I just glibly
stated, this particular commentary turned out to have
been full of arguments and stories that cry out, if not for
conclusions and endings (which most of them had when
they were made or told), then for assessing the claims
of this project. The most ambitious and encompassing
among them was that commentary is a viable genre of
ethnography based on virtual text archives. ‘‘Viable’’ means
 Endings and Ends

that this particular form is suitable for producing and communicating

ethnographic knowledge such that findings can be evaluated as contribu-
tions to anthropology. Whether or not our experiment has been successful
will be for critical readers to decide. Meanwhile I would like to offer some
points for consideration.

Once Again: Presence and Representation

Long before I began to work on this project I had conceived it, or one like
112 it, as an ‘‘experiment’’—without giving much thought to the meaning of the
term. Undoubtedly ‘‘experimentation with genres,’’ one of the proclaimed
outcomes of anthropology’s literary turn, was on my mind. Yet it was not a
playful and tentative choice of form, a wish to try something different, that
made me write the book. A desire for experimentation alone would not
have carried me through the travails of commentary. The undertaking was
triggered and its momentum sustained by the document of an event that
might have been forgotten or stayed buried among stacks of tapes and
papers had it not taken on a new kind of presence in a virtual archive on the
Internet. Even now, with memories that had worn off in the course of many
years refreshed, I cannot easily state what made me select the conversation
with Kahenga. For a long time our exchange had little interest or urgency
for me as ‘‘ethnographic material’’ because it did not seem to be directly
relevant to research topics I was working on. Yet as soon as I took the first
step and began to listen to the recording I was captivated by the immediacy
and nearness of our exchange. This experience of presence created by a
document from the past rather than by current concerns and theoretical
preoccupations was the encouragement I needed to put to a test thoughts I
had formulated in a programmatic essay on the ethnographic potential of
virtual archives (Fabian 2002b; reprinted in 2007: chap. 9)
There is something intensely personal about experiencing presence
through a document of past events. Listening to voices and sounds fills
one with the pleasure of recognition; it feels good to be able to un-
derstand the language, and one cannot wait to exercise old skills of tran-
scribing and translating. It does not take long, however, before delight
becomes mixed with pain, enthusiasm with strain, and play turns into work,
 Endings and Ends

perhaps not necessarily but whenever we want to re-present what we

experienced.
Personal and private as the first impetus may be, re-presentation that
counts as ethnography is a public undertaking; it must, in its form, be
recognizable and it must offer content and discuss questions that are on the
agenda of the discipline. Form—commentary as a genre—has been at the
center of attention in almost every part of this book. Its demands and possi-
bilities, as well as its limits, were discussed and demonstrated. With regard
to content—matters talked about, topics discussed, domains of knowledge
contributed to—my commentary stayed as close to the text as was feasible 113
and did little to relate information and insights systematically to past ethno-
graphic research and current theoretical questions.
Long before I came to this final chapter, I wrote a note to myself thinking
it could help to explain (if not excuse) a conspicuous absence of references
to ‘‘the literature’’ in this commentary:
When I wrote drafts of the preceding chapters I consulted dictionaries (al-
ways also sources of ethnographic information) and let myself be guided (as I
had done when I talked with Kahenga) by what I remembered from my
readings and many conversations on Luba culture. Rare visits to the excellent
library at Stanford University left me staring at shelves full of books on magic,
and witchcraft. Leafing through some of the anthropological classics, among
them ethnographies based on research in regions outside of Katanga and the
Congo but closely related culturally and linguistically, gave me, I must admit,
attacks of anxiety that reminded me of my student days. What could this
commentary possibly add to existing knowledge of these practices and their
attendant conceptual and linguistic apparatus? But then, I consoled myself,
such had never been its purpose. What was the purpose? To re-present the
document of an event in the past so as to make it possible to confront it in the
present. Ethnographic knowledge of magic and witchcraft may have been
cumulative to the point of diminishing returns from further research; our
understanding, not only of the practices we witnessed but also of the docu-
ments we made of them, is not cumulative. Like contexts of inquiry, contexts
of understanding and interpretation change and require (or justify) new
efforts every time we confront our archives.
 Endings and Ends

I see no reason to retract this statement now but it would be a disservice

to this project to maintain the stance I took without compromise. There-
fore, as much as it can be done without compromising the experiment by
burdening the commentary on a conversation with commentary on the
commentary (or reducing the commentary to a monographic presentation,
after all) I will, in the remainder of this final chapter, attempt to show the
significance of our findings about Kahenga’s world, work, and thought
before I end with reflections on the idea of ethnography as ‘‘confrontation.’’

114
Kahenga and the Contemporaneity of Luba Tradition

In the chapter on Kahenga’s work we asked whether he was idiosyncratic as

a munganga. Going on text-internal evidence alone we concluded that such a
characterization made little sense. True, there was much in his life history,
training, and current work that expressed an emphasis he put on personal
choices and cultural preferences. But—a point made before and worth
repeating now—that was precisely what made him a successful, modern,
and certainly nonmarginal practitioner of his trade in a part of contempo-
rary Africa.
Among the threads that kept the questions and topics of our conversation
together were Kahenga’s conscious identification with his background in
Hemba tradition and, as I noted more than once, the ‘‘foreknowledge’’ of
Luba ethnography I brought to our encounter. To start with the latter, I
should now state why and how familiarity with Luba culture came to be
part of my intellectual biography. The story began half a century ago when I
first read Placide Tempels’s Bantu Philosophy, a treatise in which this Belgian
friar summarized his understanding of the deep (‘‘ontological’’) principles
guiding the life and thought of the people he had come to convert to
Christianity, the Luba-Shankadi of Katanga, southwestern neighbors of the
Hemba.≤ By the time his book reached an international public in the 1950s
Tempels had turned from missionary ethnographer to charismatic prophet
and leader of the Jamaa, a Catholic movement that became the subject
research for my dissertation (Fabian 1971). Just before I left for the Congo in
1966 I had a memorable meeting with Tempels in Belgium and later became
acquainted with several of his confreres who were similarly steeped in Luba
 Endings and Ends

language and culture, which, they insisted, was the key to understanding
Tempels’s message and the success it had beyond Luba country in the
multiethnic settlements of miners and railway workers where the Jamaa
emerged. They also made me aware of the ethnographic literature that had
accumulated since the beginning of Belgian colonization.
During fieldwork I realized that the Luba past was present in the move-
ment I studied and, after returning to my university, I spent much time and
effort reading the sources. Most of what I learned I filed away as ‘‘back-
ground’’ material to be used as examples of the survival of traditional
African culture in the contemporary scene of religious enthusiasm. This 115
was a misunderstanding it took many years to overcome, quite likely be-
cause these writings seemed antiquated compared to, say, Victor Turner’s
magisterial works based on his research among the Ndembu of Northern
Zambia, a subgroup of the Lunda whose close kinship with the Luba was
somehow obscured by political borders between (former) British and Bel-
gian colonies and probably even more so by linguistic barriers between
English and French/Flemish ethnographic scholarship.
Work on this commentary made me remember all this. I went back to one
of the classics almost immediately when I started to transcribe the recording
(see the many notes citing Van Avermaet and Mbuya’s Dictionnaire Kiluba-
Français). Later I reread two important works, one a two-volume summary
of Luba ethnography published early in the colonial period (Colle’s Les
Baluba, 1913), the other an equally comprehensive monograph, De Luba-mens
by Theuws, published soon after Independence in 1962.≥ To consult once
again other authors I remembered as authorities cited by the authorities just
mentioned (Burton, De Clercq, Van Caeneghem and others, more or less
the bibliographic list in Theuws 1962) I would have needed library resources
I do not have in this final phase of writing. At any rate, for present purposes
such an effort would not be justified.∂
What are these present purposes? They concern answers to questions
about the representativity (1) of Kahenga as a munganga, (2) of his dis-
course and practices as Luba, and (3) of his work as part of modern,
contemporary culture in the region where we met. I am prepared to answer
the first and second questions indirectly by affirming that, checked against
written sources, what Kahenga did and later explained in our conversation,
 Endings and Ends

can be traced generally, and often quite specifically, to information given in

these classical ethnographies. At times it was tempting to demonstrate this
in the commentary by notes filled with quotations and references. With the
exception already mentioned (the Luba dictionary) I decided not to follow
this convention of displaying scholarship mainly because it would have
amounted to yet another relapse into the monograph, the genre for which
commentary should be an alternative. In other words, I am asking the
reader to accept an all-inclusive claim for representativity (risking being
proven wrong).
116 One could point out that the second question—how ‘‘Luba’’ was Ka-
henga’s work?—takes a shortcut because it skips an issue that should be
addressed first. Is it really justified to subsume his Hemba identity under
Luba? The claim I just made implies the short answer (yes, it is justified).
The long answer would be an involved and complicated account of confu-
sion between linguistic and sociopolitical categories that brought about the
colonial invention of Luba, first as one, later as two major and numerous
‘‘related’’ ethnic groups. An important (and politically fateful) distinction
between Luba-Kasai (speaking tshi-Luba) and Luba-Katanga (speaking ki-
Luba) was based on linguistic criteria defined at a level above local variation
that suited the missions and colonial administration. For at least a century
the people thus labeled have, on the one hand, accepted (and developed)
these colonial categorizations (at times with tragic consequences not unlike
the genocidal Hutu-Tutsi distinction in Rwanda) and, on the other, often
ignored them on the local level and in interpersonal relations, especially in
urban contexts, only to revert to them in situations of political and eco-
nomic strife.
The third question—how representative was Kahenga’s work as part of
modern, contemporary culture in the region where we met?—could also
have been answered by detailed annotations culled from studies based on
field research in Zaire/Congo conducted in the seventies during the same
decade when I met with Kahenga. I should mention two works on medical
anthropology, The Quest for Therapy in Lower Zaire (1978) by John Janzen and
Death in Abeyance: Illness and Therapy among the Batabwa of Central Africa
(2000) by Christopher O. Davis.∑
Finally, aside from noting the scant use of the sources of ethnographic
 Endings and Ends

information just mentioned, readers may wonder why I did not discuss
older classics and contemporary writings on the anthropology of magic and
witchcraft such as Marwick 1970 or Tambiah 1990.∏ The subject is a timely
one, to say the least, especially also in African studies (Moore and Sanders
2001, Pels and Meyer 2003), but our task has been to comment on a text,
not to mine it for evidence or arguments in current theoretical debates.

Commentary and Confrontation

While timeliness has not been my concern in this late ethnography, reflec- 117
tions on time—on temporal aspects of research and writing—got the project
started and moved it through every phase of its realization. To recapitulate
these steps will be a convenient way to pass our work in review. Before I get
to this, however, I should say more about ethnography as ‘‘confrontation,’’
an idea that has occupied me ever since my first object of research, the
Jamaa movement, made me think critically about a key concept of its
teachings: encounter.π My first African interlocutors left me little choice but
to assume a stance that was, at about the same time, emerging as ‘‘ethnogra-
phy of communication.’’ As a theoretical (re-)conception of practices of
research it offered anthropology an alternative to naive positivism, until
then the dominant epistemological position that had equated empirical
scientific inquiry with observation and data collection (a view obfuscated
rather than clarified by ‘‘participant observation,’’ a seemingly irrepressible
ghost of a concept that keeps haunting pronouncements about ethno-
graphic method).
This reorientation had political reasons and implications, responding, as
it did, to changed postcolonial conditions of research, and it prepared the
ground for, among others, dialogical and feminist approaches and a host of
other critical revisions of anthropological theory and discourse. About
some of them I have critical reservations, others I accepted or may have
helped to bring about; more thinking needs to be done, especially about
interactive, communicative, and interpretative conceptions of ethnography
that appear firmly established. One critical notion that has been guiding my
own work, the present project included, is the idea of confrontation. What I
have in mind is not only an intellectual mode or attitude that makes one
 Endings and Ends

inclined to be polemical in reasoning and arguing but an agonistic, dialecti-

cal view also of phases of ethnography that have been thought of as being
merely preparatory to theoretical debates.∫
Fieldwork is not only the subject of controversies once it is done; it should
be recognized as being confrontational while it is done and the same goes
for each of the steps we take when we document, interpret, and present
ethnographic knowledge. Descriptively or programmatically, ‘‘confronta-
tion’’ occurred in every chapter of this commentary. In the introduction I
announced that the conversation with Kahenga and the text will be con-
118 fronted; in chapter 1 I discussed reasons why; in chapter 2 transcript and
translation of our text were said to confront each other; chapter 3 prepared
the ground for confronting Kahenga’s thought, and examples were com-
mented on in chapter 4; and in chapter 5 confrontation was introduced as
an alternative to relativist bracketing as ‘‘beliefs’’ of other ways of thinking
and other kinds of knowledge.
In some academic cultures being confrontational does not count as a
social virtue; it is considered bad style if not actually unethical. Cultural
preferences, however, should not be allowed to determine epistemological
positions we need to take when we reflect on the nature of ethnographic
knowledge based on intersubjective exchanges, on meeting those whom we
study ‘‘at eye level.’’ Furthermore, and this is the reason for invoking episte-
mology, what counts is not only factual encounters we have in our work
(those we have come to recognize as dialogical) but our understanding of
their ‘‘conditions of possibility.’’ Pointing to confrontation as such a condi-
tion, however, will serve its purpose only when its static connotation, that of
a face-off, as it were, is not allowed to dominate. We can avoid this if we
think of confrontation as action and event.
Before I summarize how this view informed the current project I would
like to turn one more time to the text, as a relief from these abstract
reflections and as a last fling at commentary. It occurred to me to search
the conversation with Kahenga for linguistic evidence of a confrontational
mode (or mood) by looking at the incidence of negation in this generally
relaxed, amiable, and seemingly nonconfrontational exchange. To simplify
matters I decided to concentrate on (h)apana, no.Ω The first and most
remarkable finding was that hapana occurred most often and most emphat-
 Endings and Ends

ically when Kahenga took his turn in the conversation (ninety times as
opposed to about a dozen occurrences when I spoke). Granted, higher
frequency may in part reflect that Kahenga’s speech occupies more space in
this text than mine but a glance at just a few examples will show what I
mean by emphatical negation: He disagreed when I suggested his ancestors
could not have known about labor problems (7). He denied use of bizimba,
magic charms, and insisted that he only works with miti, plant material (9,
40). He contradicted my suggestion that Whites may have brought illness
to the country (21). Negation as resistance was expressed when he spoke of
his ‘‘mother’s’’ opposition to the plans the missionaries had for his education 119
(25). He told me that he refused to work with a tutelary spirit and spirit
associations (27, 28, 34) and denied practicing sorcery (31). He rejected my
suggestions that his knowledge was purely oral (33) and that the mission-
aries might be against his work (54, 56).
In what way is this significant? First, these and many other statements
suggest a distinction to be made between the mood of a conversation (I
called ours amiable) and its mode, which attention to negation reveals as
confrontational. Second, it should be remembered that negation and con-
frontation are not the same. Negation may appear to be one-sided (as it
does in this case) but it takes two to confront each other and that condition
was fulfilled by the stance I took in my contributions to the exchange.
Grammatically, most of them may have been questions, quite a few were
statements, yet all of them were ways of confronting my interlocutor. Per-
haps this is what makes me think of commentary as a continuation of our
conversation.

Co-presence and Contemporaneity

A key issue in all this is presence; there is no confrontation without pres-

ence. Recalling how this idea informed the writing of commentary from
start to finish should allow me, if not to formulate a definitive conclusion
(that was already ruled out) then to extricate myself from the infinite
possibilities of this experiment.
It all began with the thesis that a technological development, the virtual
archives of texts that can be made available on the Internet, changes condi-
 Endings and Ends

tions of writing because it gives a new kind of presence to documents on

which we can base our ethnographies. The amenities of word processing
and document storage, however, only make us keenly aware that the pres-
ence of a text becomes more, not less, problematic (or demanding) when it
is easily retrievable.∞≠ Ethnographers who have conducted research, re-
corded speech, labored over representing recorded sound graphically and
translating their transcripts into another language remember too much to
think of the documents they produced as being simply there to be deposited
(and disappear) in archives. All that work would have been wasted if it had
120 not served the purpose of making these texts yield ‘‘ethnography.’’ Hence the
question that must be asked: What is it a document makes present?
Our way of answering that question was to retrace the genesis of our
document to something that happened: a conversation, a speech event that was
occasioned by another event, the performance of a ritual. With that the
problem was no longer one of presence vs. absence but of presence vs. past,
and from then on we tried to show how remembering and re-cognition are
involved at all stages, including those (chapters 1 and 2) that may seem
merely preparatory but are in fact at the core of commentary as a genre of
ethnography.
Both events, ritual and conversation, required copresence of the partici-
pants. This is not the truism it appears to be when we recall the idea of
coevalness. Other than factual synchronicity, which may be predicated on
actions or events occurring in physical time, coevalness means sharing of
time as a condition of intersubjective relations without which communica-
tion and joint performances could not occur. In intercultural even more
than in ordinary situations where participants already share knowledge and
habits we are made aware that sharing of time does not ‘‘come natural’’ and
is always a precarious undertaking, depending on linguistic and other cul-
tural skills. Without them copresence in action, or communication as per-
formance, could not happen—not to forget the element of serendipity, sheer
luck, that most ethnographers have learned to appreciate in their work.
Had I not taken up Kahenga on his casual offer to ‘‘close’’ our house, had he
not been willing to talk about the ritual after it was performed, and had
I not thought of recording our conversation—a series of ‘‘accidents’’—we
would have had nothing to comment on.
 Endings and Ends

Apart from presenting, making present, what the text contains about
Kahenga’s work, world, and thought, the chapters that followed addressed
above all issues of copresence. We commented on the coexistence of ele-
ments that made up the context of practicing bunganga: Kahenga’s clientele
was rural, urban, and multiethnic; the problems for which he was consulted
included illness, fertility, marital and labor relations, security, and power
politics, which he tackled with a gamut of ‘‘therapies’’ running from herbal
medicine to magic, ritual, and practices that (despite his denial) came at
least close to sorcery. We repeatedly denounced these terms and the distinc-
tions they posit as imposed and inappropriate but do not seem to be able to 121
avoid them entirely. In my view, this is yet another way in which the
essentially confrontational nature of ethnography comes to the fore. Playing
games of classification is one thing, letting what we classify disturb our
peace of mind is another. A disturbed mind is a mind alive. Likewise, the
purpose of commenting on text is to show that it is alive. In anthropology,
to extract explanations from texts or use them for analytical exercises,
the business of scientific inquiry, should not depend on treating them
as corpses.

Ends of Commentaries and Archives

Let me close chapter and book with attempts to answer two simple ques-
tions: Who needs commentaries on texts deposited in archives? Who needs
virtual text archives? It may be surprising that I ask them in that order—
after all, logically, texts come first, commentaries ‘‘are extra,’’ to paraphrase a
statement about symbols attributed to Woody Allen. But then, foremost on
my mind has been a practical issue: writing ethnography in the form of
commentary. An obvious reply to the first question could be that readers
need help if they are to understand the form and content, the social and
political aspects, and the historical significance of the text the ethnographer
puts before them. However, this would miss the specific purpose of our
study, which was to respond to changed conditions for representing knowl-
edge due to the virtual presence of texts on the Internet. Above all, such a
commonsense answer remains vague as long as it is not clear who those
readers who need help are. One thing is certain, the potential readership of
 Endings and Ends

texts in virtual archives is not the same as the one that conventional eth-
nographic publications address and reach. The ethnographer who has relin-
quished control of his ‘‘data’’ by conveying them to an ‘‘open source’’ cannot
be sure what—in the sense of topics treated, problems posed, solutions
offered—readers, who may have found a text just by surfing the Internet,
were looking for, much less what kind of help they need (or want). It is safe
to assume that they are more interested in the content of virtual documents
than in questions of form and representation. Conversely, the audience the
ethnographic commentator can anticipate, most of them colleagues and
122 students from his or related disciplines, may read a commentary such as this
one without consulting the text on the Internet. All this can put a damper
on the enthusiasm for the ‘‘new kind of presence’’ of ethnographic texts that
made me conduct this experiment.
Whether or not ethnography as commentary is worth the effort will
depend on our answer to the second question: Who needs virtual archives?
What such archives are good for, how they enable the ethnographer to
make research documents widely available, as single texts or as part of
a corpus, and how they provide evidence the reader can check—all this
should be clear by now. One could also point out that setting up virtual
archives can be a step toward meeting not only demands and expectations
to ‘‘return’’ our research results to the people we study but to initiate
discussion of our work as well as additions to the corpus. That documents
created by blogs and chat groups devoted to themes anthropology is inter-
ested in deserve our attention is by now widely recognized; Internet-based
ethnography has become accepted as a legitimate alternative for, or comple-
ment of, traditional fieldwork and the concomitant literature on research
methods is bound to grow. At the same time it becomes more urgent to
engage in reflections on epistemology and theory. I should like to conclude
by formulating a thesis for debate: What ‘‘writing from the virtual archive’’
could entail for anthropology may become clarified if we consider the
difference between a database and an archive.
Databases, conceived and established long before the advent of the com-
puter and the Internet, belong to the conceptual arsenal of a positivist and
essentially ahistorical (some would call it ‘‘modernist’’) view of anthropology
as a science that, to put it mildly, is no longer generally accepted. Any
 Endings and Ends

depository of information may of course be called a database—metaphori-

cally. Strictly speaking, the term should be reserved for stores containing
discreet, quantifiable bits that are amenable, at least ideally, to statistical
operations requiring random sampling. There is no need further to elabo-
rate on these characteristics to make the point of my thesis: A virtual
archive of the kind envisaged here is never a neutral storage device. Given
the communicative practices involved in producing ethnographic texts, it is
inevitably a partial, personal, and, again ideally, communal creation. An
archive’s historical contingency is its strength if its purpose is to mediate
between (recorded or otherwise documented) events, the presentation of 123
evidence, and our aims and claims to produce anthropological knowledge.
Notes u

1: An Event
1 This is a possible translation I chose for the phrase in the text
on the website. He said niko napita nabo, literally: ‘‘I (regularly)
pass, go along with, them,’’ or ‘‘I keep their company.’’

2: A Text
1 How this position emerged and how it relates to develop-
ments in anthropology (and related disciplines) is discussed
in an essay on ethnographic objectivity (Fabian 1991a, re-
printed in Megill 1994:81–108 and Fabian 2001: chap. 1).
2 The text shows not a single instance of code switching as well
as an extraordinary absence of French grammatical words and
fillers that infiltrate the language of most Swahili speakers in
this region. As far as I can see, Kahenga made copious use
only of one such filler, French bon (fine, all right), though
others occur (for instance, chose, 17). He was not influenced by
my constantly using mais (but). Most of the French loans
 Notes to Chapter 2

he did use were numerals for dates and grade levels in school, terms like
certificat, examen, and pères for missionaries. On the ‘‘poetics of lexical borrow-
ing’’ in Shaba Swahili, see Fabian 1982 and de Rooj 1996 on stylistic functions
of code switching.
3 Rereading this I feel I should add a qualifying note. It must have occurred to
me at the time that this exchange contributed to my project ‘‘Language and
Labor’’ (see Introduction) as did the many conversations I had with painters
and actors. We always touched on work and conceptions of work. Later on I
will comment on work and labor relations among the problems treated by
Kahenga. Still, the text shows that I did not follow a schedule of topics and
126
questions.
4 See, for example, Cook (1990) and Ashmore, MacMillan, and Brown (2004).
5 When I searched the Internet for ‘‘ethnography of listening,’’ most of the
roughly seventy results were related to Bendix’s article, but see the online essay
Hill n.d. with further references. A search for ‘‘ethnography of hearing’’ did
not bring up a single entry (!) in spite of the obvious connection between
‘‘listening and hearing,’’ which resulted in more than 25,000 hits.
6 As far as recording goes, this has by now been recognized and the same goes
for taking field notes. Excessive claims have been made for translation as the
essence of ethnography. Transcribing still seems to rank lowest in critical
examinations of ethnographic knowledge production.
7 To give just one example, Kahenga often pronounces the negative prefix ha–
without initial {h} (but initial {h} may—because of hypercorrection?—also be
omitted in other cases, such as in the demonstratives (h)ii, (h)uyu). Since this is
not consistent I opted for using the standard form ha. This also goes for
aspiration and palatalization of {s}, {z}, and {j}.
8 This view of transcription contradicts claims structural linguists made when
they took pride in being able to decode, without recourse to semantics, the
phonology and morphology of a language from a small sample of text.
9 On the document, see Fabian 1990a; on reading such texts, 1993. The original
of the Vocabulary in facsimile as well as the oralized version and the English
translation can also be found online in Archives of Popular Swahili.
10 Recognition as re-cognition has been a theme I explored in several essays
which may be consulted as background to these remarks (for instance, Fabian
1999, reprinted in 2001: chap. 9). The thoughts on transcribing and remem-
bering that follow were first formulated in a paper, ‘‘Ethnography and Mem-
ory’’ (Fabian 2007: chap. 10).
11 Strictly speaking: not marked morphologically. In baba, father, vs. mama,
mother, gender is of course marked semantically, but this is not the case with
many other kinship terms, such as mukubwa, elder sibling, or nkambo, grand-
parent or ancestor.
 Notes to Chapter 3

12 In standard Swahili the present marker is –a–; –na– is usually called ‘‘progres-
sive,’’ the tense of continued action. In Shaba/Katanga Swahili the two seem
to have merged.
13 Though much progress toward such an aim was made by the work of a
Congolese linguist who grew up in the region; see Kapanga 1991.
14 For a social history of Shaba Swahili up to the mid-1980s, see Fabian 1991b.
15 I found little evidence in this exchange of two other kinds of avoidance or
omission that may result in elliptic speech: politically sensitive or dangerous
subjects (see the remarks on ‘‘vociferous silence’’ in the conversation with
Tshibumba, Fabian 1996:306–9) and the effects of forgetting (see the conversa-
tion with Baba Ngoie, Fabian 2003). Both texts can be consulted online in 127
Archives of Popular Swahili.
16 I make a similar argument with respect to the literature on Luba culture and
medical anthropology in the concluding chapter.

3: Kahenga’s Work
1 This term and concept is crucial in Kahenga’s work and will be illustrated
by examples that follow; for a more comprehensive account see chapter 5,
‘‘Kahenga’s Thought.’’
2 According to Sacleux (1939:497), the term malari, alternate maladi, now generic
for illness, originally referred to maladie du sommeil, sleeping sickness.
3 Kahenga must have thought of his practice at home in his village. The
question never came up, but it is unlikely that his wife accompanied him on
his travels. On the road he most likely was assisted by a female relative.
4 The Swahili dictionary confirms this (under dudu, dim. kidudu): ‘‘Also used for
various diseases caused by, or attributed by the natives to, parasites and other
insects on the body.’’ Unless another source is cited, the ‘‘Swahili dictionary’’
refers to my edition of Johnson’s classic (1963, orig. 1939).
5 He said naweza kufanya, lit. I can do, which is elliptic for ‘‘I can treat diseases
with remedies I prepare.’’
6 There is an unexplained discrepancy between ‘‘Mukonkwa,’’ taken from the
inscription on the tape recording, and ‘‘Mukonkole,’’ transcribed from the
tape.
7 A person may have secret names, may take or drop additional names or
altogether change public names, and names one is called by may differ from
those that show up on documents.
8 In my translation I have them down as belonging to the order of the White
Fathers, officially Missionnaires d’Afrique, usually Pères Blancs for short (‘‘White
Fathers’’ in English). I checked this because pères blancs could also mean ‘‘white
missionaries’’ but it proved correct for the time specified.
 Notes to Chapter 4

9 Achille Mutombo, a friend (and contributor to Archives of Popular Swahili),

confirms that mufumu is a current term in Katanga Swahili. However, its
principal meaning is ‘‘diviner.’’ Only in contexts where mufumu is opposed to
munganga can mufumu mean sorcerer, i.e. a practitioner whose power is based
on means other than herbal medicine.
10 The term ‘‘idiosyncratic’’ was used by Janzen to designate banganga who did
not fit distinctions of specialists recognized in what he calls the ‘‘traditional
system’’ (1978: 23, 24, 45–46).

4: Kahenga’s World
128
1 Katanga Swahili mukini, or mugini, is a locative form derived from East Coast
Swahili mji (analogous to mutoni, course of water, from mto). Mji, according to
the Oxford English Dictionary, means ‘‘village, hamlet, town, city, i.e. a collection
of human dwellings irrespective of number,’’ and this apparent lack of preci-
sion may well be an accurate terminological reflection of coastal Swahili
culture and society.
2 The latter depends on whether there is matrilinear or patrilinear reckoning of
the relationship, something that kinship terms in Katanga Swahili do not
specify.
3 In fact, I don’t recall any occurrence in Katanga Swahili of such a term. In East
Coast Swahili, the meaning of mbegu, lit. seed, may be extended to include
breed or race but, as far as I can tell from documented speech, that was not the
case in Katanga. Of course, had race become a topic, French loans, such as race
or even racisme, would have been available to us.
4 One gets tired of repeating this, but muzungu is not defined by race if race
means skin color. A ‘‘black’’ African American or African European will be
called muzungu without hesitation while a ‘‘white’’ Greek, Italian, or Por-
tuguese local merchant is understood to belong to a category apart. Neverthe-
less, it is difficult to shed the habit of translating muzungu as ‘‘white European’’
(see on this Fabian 1995) and it may be said that I gave up trying when, in the
following and elsewhere in this commentary, I use ‘‘white,’’ ‘‘White,’’ and
‘‘European’’ rather indiscriminately.
5 The prefix ki– usually marks the noun class of things. When it is used with
terms signifying persons it can take on all sorts of meaning, among them
contempt, sometimes mixed with respect as in kimama, usually applied to a
corpulent, powerful woman.
6 Katanga Swahili also has namna and ginsi. In my understanding the former
denotes different modes while the latter refers to something like different
instances; neither seemed appropriate for eliciting a taxonomy of diseases.
7 There is other evidence in our text of his work crossing ethnic boundaries.
 Notes to Chapter 5

When he took us on as expatriate clients this was upon recommendation from

Mwenze Kibwanga, ethnically a Luba-Katanga and also a client (13), and G.
Munongo, a prominent Katanga politician of the Yeke group, had sought his
teacher’s services (33).
8 Was this a hint for me? If it was I would not have taken it; I never paid for
information (although in some cases—painters who had brought along pic-
tures I bought and Kahenga, who performed a service—interlocutors were
remunerated indirectly).
9 The Jamaa had been the subject of my dissertation fieldwork in southern
Katanga eight years earlier (Fabian 1971 and numerous later publications).
10 Kasongo is a town north of Kahenga’s home country. It always had historical
129
significance as one of the places where Arab and Belgian colonization clashed
during the so-called antislavery campaigns of the 1890s.

5: Kahenga’s Thought
1 See for instance Kirsch 2004. In fairness to our predecessors it should be
acknowledged that not all of them found comfort in the Durkheimian posi-
tion; see Engelke 2002 (on Evans-Pritchard and Turner). An essay by R. Firth
(1959) may still be the clearest statement of the problem anthropology has had
with the study of religion.
2 A phrase coined by Bruno Latour in an essay critical of accepted oppositions
between science and religion (2001).
3 For a comprehensive appraisal of the ‘‘rationality debate,’’ see Buchowski 1997.
4 A brief comprehensive statement on thought and popular culture in Shaba/
Katanga may be found in Fabian 1998, chapter 4.
5 Remember, the task I set myself is commentary, not detailed lexical/semantic
analysis. Therefore, when I searched this text it was always only with the help
of the search function of my word-processing program, not with any of the
specialized linguistic software that is available.
6 The marked Swahili terms for remembering, the verb –kumbuka, with a
causative form, –umbusha, and the noun ukumbusho, all of which were promi-
nent in the examples of discourse mentioned earlier, are also absent from this
text.
7 I have no explanation for the fact that Kahenga, who came from a region
known for ‘‘good’’ Swahili, used the ‘‘low’’ –yua form.
8 It would seem that ‘‘knowledge of names’’ consists above all of being able to
use taxonomic labels. But even the fragmentary information about naming in
our text suggests that nomenclature serves identification rather than classifica-
tion. See van Avermaet and Mbuya 1954:827 about the name as essence (of
plants, for instance).
 Notes to Chapter 5

9 It is not listed in Swahili dictionaries. Most likely it is one of the few loans
from Lingala (or other languages from the lower Congo). If so, it would be
interesting to speculate about mayele as belonging to a colonial vocabulary.
10 I briefly mentioned this at the beginning of the section on thought and
knowledge; for more thorough treatments, see an early interpretation of ma-
wazo in the discourse of the Jamaa movement as a gnosis (Fabian 1969)
and comments on metahistory (or historiology) in the work of the painter-
historian Tshibumba Kanda Matulu (Fabian 1996:309–16).
11 From East Coast Swahili –fanza, cause to make but also to put in order, to
mend. In this and other texts I transcribed the applicative as –fansia rather
130
than –fanzia because that was how it sounded. I don’t remember ever hearing
–fanza in Katanga Swahili.
12 Kahenga used this abstract term for health only three times in this conversa-
tion in which health was a central theme. After the example just noted it
occurred again as adjective, muzima, alive, and then again as a noun, in a
context to be commented on later.
13 Swahili dictionaries have ki/bizimba for birdcage, coop, dovecote, hutch, wit-
ness stand—nothing approaching the meaning of the term in local Swahili
unless one wants to speculate on a metaphorical sense: something in which
power is captured.
14 Van Avermaet and Mbuya use phonetic symbols to mark palatalization in
Luba; perhaps a more accurate transcription in our text would have been
bizhimba (or bijimba in a French text). In local Swahili both pronunciations, zh
and z, can occur.
15 As regards vital force, force vitale, the dictionary refers to Placide Tempels’s
‘‘Phil. Bantoue . . . p. 58–59.’’ The passage can be consulted on the Tempels
website, see note 1 in the following chapter. What van Avermaet and Mbuya
have to say about bizimba may clarify its meaning to some extent, until one
consults the cross-references to kilumbu (381–83) and nganga, under anga (24–
27) and comes upon semantic fields of stunning complexity that make the
difficulties we encounter in our text pale beyond comparison.
16 I take this term to be a compound noun made up of vili, the Hemba equiva-
lent of Luba-Katanga vidye, and nyambi, the Hemba term for God. On vidye,
see the long entry in van Avermaet and Mbuya (1954:783–86).
17 Toward the end of our conversation, when I showed Kahenga a Luba diviner’s
gourd and its contents, he said about several figurines that they were mizimu
(70). These objects were miniature carvings of well-known ancestor statues
(of the kind that would be called fétiches).
18 In East Coast Swahili the pronunciation is kivuli, for which the dictionary has
this entry: ‘‘(1) a shadow, a shady place; (2) sometimes used to mean a ghost,
apparition.’’ In the mu/mi class, it becomes mvuli: ‘‘a shady place, shade of a
 Notes to Chapter 6

tree.’’ See also Sacleux (1939:634), who has ‘‘photographie, portrait’’ as possible
meanings of mvuli.
19 The wordplay speech/bespeak is irresistible, especially when one also hears
the German cognate of ‘‘to bespeak,’’ besprechen, which means something like to
treat illness by reciting formulae ‘‘over’’ the patient or the diseased part of his
body. A classic comes to mind: B. Malinowski’s pragmatic speech-theory of
magic developed in the second volume of his Coral Gardens (1935).

6: Endings and Ends

1 Observant readers, I expect, will have noticed that two substantial parts of our 131
text, ‘‘Comments and explanations on specimens of plants’’ and ‘‘Looking at a
diviner’s calabash from Luba country,’’ did not get the attention they deserved.
The former (60–69) prompted observations on indexical communication and
ellipsis in chapter 2 and examples of Kahenga’s work as a healer in chapter 3; a
few references to the latter (70–71) were made here and there in comments on
ethnic distinctions (chapter 4) and practices of divination (chapter 5). Neither
did justice to the wealth of ethnographic information in these parts of our
conversation, especially the kind that Kahenga provided when we looked at
specimens, photographs taken of trees from which they were collected, and at
the Luba divining gourd with its contents of miniaturized statuary. Ideally, the
visual material we talked about in these two sections could have been pre-
sented together with the text and would have enriched the commentary, but
this was not a practical option. The eighteen photographs taken by Ilona
Szombati in September 1974 were shown to a botanist at the University of
Lubumbashi. On a handwritten list he identified nine plants by their botani-
cal names (three of them by species only). Because this information is not
likely to add anything significant to specialist knowledge, to reproduce the
photographs and the list of identifications is hardly worth the effort and cost.
As to the objects we had before us, the plant specimens were not preserved
and the divining gourd is no longer in my possession.
2 Bantu Philosophy was first written as a series of articles in Flemish and then
published as a book in French (1945); it appeared in an English translation in
1959. For comprehensive background information a web site devoted to Tem-
pels’s work, including an archive of his writings, may be consulted at http://
aequatoria.be/tempels.
3 Colle was a White Father missionary stationed in Hemba country. His work
(1913) was published in a series intended to set up what we would today call a
data bank for the colonial government (incidentally, the long introductions by
Cyriel Van Overbergh to each of the two volumes are excellent but to my
knowledge unused sources for the history of our discipline). The work was a
 Notes to Chapter 6

monograph in the strict sense in that it presented the material in rubrics

previously established by standard questionnaires. Theuws, like Tempels a
Franciscan missionary but also an Oxford-trained anthropologist, worked
among the Luba-Shankadi and wrote a veritable summa of Luba ethnography.
This monograph was published by the Royal Museum of Central Africa in
1962, two years after the end of Belgian colonial rule. Colle had compiled his
own observations and those by other authors who had preceded him; Theuws
took a similar approach but his own contributions were based mainly on a vast
corpus of texts collected by himself and his confreres, a small portion of which
he later published in English (1983).
132
4 The most interesting result of revisiting the literature was the outlines I began
to see of a history of colonial ethnography and the simultaneous construction
of its object (in this case the ‘‘Luba’’), the role of government sponsorship of
publications by missionaries (Catholic, with the exception of Burton), and
structures of authority among Luba ethnographers, all of this with politi-
cal consequences up to the present. For a contemporary reading of Van
Caeneghem, see Mudimbe (1991).
5 Both could be complemented by Hunt’s impressive attempt to document the
history of medicalization (of childbirth in this case) in the context of everyday
life in the colony (1999).
6 See also the interesting work on contemporary magic in India and Europe
(Glucklich 1997; Greenwood 2000, 2005).
7 See Fabian 1971 and a recent essay, ‘‘Inquiry as Encounter’’ (2007, chapter 12).
8 That this is not an idiosyncratic idea may be taken from a collection of essays
on ‘‘counterworks’’ edited by R. Fardon (1995).
9 Used as a particle, hapana is defined in the Swahili dictionary as ‘‘a verb-form,
there is not there, there is none . . . [c]ommonly used as a simple negation.’’
The prefix ha– generally marks negative verb forms, which are not considered
in the following.
10 Some time ago Don Handelman had thoughtful though rather pessimistic
things to say about the presence of ethnographic texts (1993).
Works Cited u

Ashmore, Malcolm, Katie MacMillan, and Steven D. Brown.

2004. ‘‘It’s a Scream: Professional Hearing and Tape Fetishism.’’
Journal of Pragmatics 36:349–74.
Bendix, Regina. 2000. ‘‘The Pleasures of the Ear: Toward an Eth-
nography of Listening.’’ Cultural Analysis 1.
Blakely, Pamela A. R., and Thomas D. Blakely. 1994. ‘‘Ancestors,
‘Witchcraft,’ and Foregrounding the Poetic: Men’s Oratory and
Women’s Song-Dance in Hêmbá Funerary Performance.’’ In
Religion in Africa: Experience and Expression, ed. Thomas D.
Blakely, Walter E. A. van Beek, and Dennis L. Thomson, 399–
442. London: James Currey.
Boyarin, Jonathan, ed. 1993. The Ethnography of Reading. Berkeley:
University of California Press.
Buchowski, Michal. 1997. The Rational Other. Poznán: Wydawictwo
Fundacji Humaniora.
Colle, R. P. 1913. Les Baluba. Collection de Monographies eth-
 Works Cited

nographiques publiée par Cyr. Van Overbergh. Brussels: Albert Dewit / Institut
International de Bibliographie.
Cook, Guy. 1990. ‘‘Transcribing Infinity: Problems of Context Presentation.’’ Journal
of Pragmatics 14:1–24.
Davis, Christopher O. 2000. Death in Abeyance: Illness and Therapy among the Batabwa
of Central Africa. Edinburgh: Edinburgh University Press.
Engelke, Matthew. 2002. ‘‘The Problem of Belief: Evans-Pritchard and Victor
Turner on ‘the Inner Life.’ ’’ Anthropology Today 18, no. 6:3–8.
Fabian, Johannes. 1969. ‘‘An African Gnosis: For a Reconsideration of an Authorita-
tive Definition.’’ History of Religions 9:42–58.
134
———. 1971. Jamaa: A Charismatic Movement in Katanga. Evanston, Ill.: Northwestern
University Press.
———. 1982. ‘‘Scratching the Surface: Observations on the Poetics of Lexical Borrow-
ing in Shaba Swahili.’’ Anthropological Linguistics 24:14–50.
———. 1990a. History from Below: The ‘‘Vocabulary of Elisabethville’’ by André Yav: Texts,
Translation, and Interpretive Essay. Amsterdam: John Benjamins.
———. 1990b. Power and Performance: Ethnographic Explorations through Proverbial Wis-
dom and Theater in Shaba (Zaire). Madison: University of Wisconsin Press.
———. 1991a. ‘‘Ethnographic Objectivity: From Rigor to Vigor.’’ Annals of Scholarship
8:381–408.
———. 1991b. Language and Colonial Power. Berkeley: University of California Press.
———. 1993. ‘‘Keep Listening: Ethnography and Reading.’’ In The Ethnography of
Reading, ed. Jonathan Boyarin, 81–97. Berkeley: University of California Press.
———. 1995. ‘‘On Ethnographic Misunderstanding and the Perils of Context.’’ Ameri-
can Anthropologist 97:1–10.
———. 1996. Remembering the Present: Painting and Popular History in Zaire. Berkeley:
University of California Press.
———. 1998. Moments of Freedom: Anthropology and Popular Culture. Charlottesville:
University Press of Virginia.
———. 1999. ‘‘Remembering the Other: Knowledge and Recognition in the Explora-
tion of Central Africa.’’ Critical Inquiry 26:49–69.
———. 2001. Anthropology with an Attitude: Critical Essays. Stanford: Stanford Univer-
sity Press.
———. 2002a. Time and the Other: How Anthropology Makes its Object. 2nd edn. With an
introduction by Matti Bunzl. New York: Columbia University Press.
———. 2002b. ‘‘Virtual Archives and Ethnographic Writing: Commentary as a New
Genre?’’ Current Anthropology 43:775–86.
———. 2003. ‘‘Forgetful Remembering: A Colonial Life in the Congo.’’ Africa 73:489–
504.
———. 2007. Memory against Culture: Arguments and Reminders. Durham: Duke Univer-
sity Press.
 Works Cited

Fabian, Johannes, and Vincent de Rooij. In press. Ethnography and Memory.

Fardon, Richard, ed. 1995. Counterwork: Managing the Diversity of Knowledge. London:
Routledge.
Firth, Raymond. 1959. ‘‘Problem and Assumption in an Anthropological Study of
Religion.’’ Journal of the Royal Anthropological Institute 89:129–48.
Foucault, Michel. 1973. The Order of Things: An Archeology of the Human Science. New
York: Vintage.
Glucklich, Ariel. 1997. The End of Magic. Oxford: Oxford University Press.
Greenwood, Susan. 2000. Magic, Witchcraft, and the Otherworld: An Anthropology.
Oxford: Berg.
———. 2005. The Nature of Magic: An Anthropology of Consciousness. Oxford: Berg. 135
Handelman, Don. 1993. ‘‘The Absence of Others, the Presence of Texts.’’ In Cre-
ativity/Anthropology, ed. Smader Lavie, Kirin Narayan, and Renato Rosaldo.
Ithaca: Cornell University Press.
Hill, S. Arden. n.d. ‘‘The Performance of Listening in Actual and Constructed
Sonic Environments—An Ethnography of Listening.’’ http://www.phonography
.org/listening.htm.
Hunt, Nancy Rose. 1999. A Colonial Lexicon: Of Birth Ritual, Medicalization, and
Mobility in the Congo. Durham: Duke University Press.
Hymes, Dell. 1974. Foundations in Sociolinguistics. Philadelphia: University of Penn-
sylvania Press.
Janzen, John M. 1978.The Quest for Therapy in Lower Zaire. Berkeley: University of
California Press.
Johnson, Frederick. 1963. A Standard Swahili-English Dictionary. Oxford: Oxford Uni-
versity Press.
Kapanga, A. M. 1991. ‘‘Variation and Change: A Case Study of Shaba Swahili.’’
Ph.D. dissertation, University of Illinois.
Kirsch, Thomas G. 2004. ‘‘Restating the Will to Believe: Religious Pluralism,
Anti-syncretism, and the Problem of Belief.’’ American Anthropologist 106:699–
709.
Latour, Bruno. 2001. ‘‘ ‘Thou Shalt Not Take the Lord’s Name in Vain’—Being a
Sort of Sermon on the Hesitations of Religious Speech.’’ Res 39:215–34.
Lyons, Maryinez. 1992. The Colonial Disease: A Social History of Sleeping Sickness in
Northern Zaire, 1900–1940. Cambridge: Cambridge University Press.
Malinowski, Bronislaw. 1935. Coral Gardens and Their Magic. London: Allen and
Unwin.
Marwick, Max, ed. 1970. Witchcraft and Sorcery. Harmondsworth: Penguin.
Megill, Allan, ed. 1994. Rethinking Objectivity. Durham: Duke University Press.
Moore, Henrietta L., and Todd Sanders, eds. 2001. Magical Interpretations, Mate-
rial Realities: Modernity, Witchcraft, and the Occult in Postcolonial Africa. London:
Routledge.
 Works Cited

Mudimbe, V. Y. 1991. Parables and Fables: Exegesis, Textuality, and Politics in Central
Africa. Madison: University of Wisconsin Press.
Nietzsche, Friedrich. 1976. On Truth and Lie in an Extra-Moral Sense. Ed. and trans.
Walter Kaufmann. New York: Viking Penguin.
Pels, Peter, and Birgit Meyer, eds. 2003. Magic and Modernity: Interfaces of Revelation
and Concealment. Stanford: Stanford University Press.
Ricoeur, Paul. 1971. ‘‘The Model of the Text: Meaningful Action Considered as a
Text.’’ Social Research 38:529–62.
Rooij, Vincent A. de. 1996. Cohesion through Contrast: Discourse Structure in Shaba
Swahili/French Conversations. Amsterdam: Ifott.
136
Rorty, Richard. 1980. Philosophy and the Mirror of Nature. Princeton: Princeton Uni-
versity Press.
Sacleux, Charles. 1939. Dictionnaire Swahili-Français. Paris: Institut d’Ethnologie.
Tambiah, Stanley. 1990. Magic, Science, Religion and the Scope of Rationality. Cam-
bridge: Cambridge University Press.
Tedlock, Dennis. 1979. ‘‘The Analogical Tradition and the Emergence of a Dia-
logical Anthropology.’’ Journal of Anthropological Research 35:387–400.
———. 1983. The Spoken Word and the Work of Interpretation. Philadelphia:University of
Pennsylvania Press.
Theuws, Théodore. 1962. De Luba-mens. Annales, Sciences Humaines, no. 36.
Brussels: Musée Royal de l’Afrique Centrale.
———. 1983. Word and World: Luba Thought and Literature. Studia Instituti Anthropos,
vol. 32. St. Augustin: Anthropos Institut.
Van Avermaet, E., and B. Mbuya. 1954. Dictionnaire Kiluba-Français. Tervuren: Mu-
sée Royal du Congo Belge.
Webster, S. 1986. ‘‘Realism and Reification in the Ethnographic Genre.’’ Critique of
Anthropology 6:39–62.
Young, Crawford, and Thomas Turner. 1985. The Rise and Fall of the Zairean State.
Madison: University of Wisconsin Press.
 Index u

Allochronic discourse, 77–78, 93 13, 55–58, 99, 108, 121; ethics of, 16–18;
memoranda and, 12
‘‘Belief in belief,’’ 91–93
Confrontation: of ‘‘beliefs,’’ 92; as mode
Boas, Franz, 14
and mood, 117–19; presentation and,
Closing the house, 6, 24, 31, 32; incanta- 54, 74
tions and, 26–27, 33; meaning of, 27, Context, 15
32, 59, 65; medicines and, 25, 28–29; Co-presence and contemporaneity, 119–21
parts of, 35; prayer and, 27, 33–34;
Davis, Christopher O., 116
remembered by ethnographer, 21–25;
Dawa, 25, 28–29, 35, 61–63, 88–89, 96,
remembered by ethnographer and his
100–104
interlocutor, 25–30, 37; requirements
Dialogue, 13, 36–38
of performance, 35
Durkheim, Emile, 18
Collage, 9
Colle, P., 115, 131n3 Epistemology, 6, 9, 37, 38
Colonization, 88–90 Ethnicity, 83–84
Commentary: comments and, 11–12; as Ethnographer: as client 5; duplicity and, 6
genre of ethnographic writing, 9–10, Ethnographic literature, 113–17
 Index

Ethnographic texts: ‘‘based on,’’ 3, 7; col- Kahenga Mukonkwa Michel, 3, 6, 27,

lecting of, 14; corpus of, 14–15; as doc- 65–70; clients, problems, and reme-
uments of events, 11, 41, 120; presence dies of, 58–60; coherence of, 104–5;
of, 10, 11; presentation of, 36, 45–46, diagnoses, treatments, and etiologies
54, 88 and, 60–65; ‘‘doing’’ medicine and,
Ethnography: autobiography and, 4, 8; 99–100; kinship and family of, 78–80;
of communication, 41; genre and, 8; on kivuri, 108–10; on knowledge and
historiography and, 4; late, 4, 8; lis- power, 96–99; knowledge in action
tening and, 44, 126n5; popularity of, and, 99–110; on landscape and mem-
9; presentation of knowledge and, ory, 87–90; race, ethnicity, and poli-
138
36, 37; remembering and, 12; text- tics of, 81–84; religion of, 84–87; on
centered, 14, 16; as transgression, 5–6, thought and knowledge, 94–96; in
40, 82; writing and, 3–4 village and city, 74–78
Europeans, 65, 68, 81–83, 88–90, 106, Kizimba, 26, 33, 87, 100–194
128n4 Kolwezi, 1, 71
Fardon, Richard, 132n8 Luba/Hemba culture, 114
Foreknowledge, 31, 69 Lubumbashi, 1, 6, 22, 49, 71
Fundamentalism, textual, 7, 13–14, 38 Lyons, M., 89
Genre. See Commentary; Writing Magic, 5, 28, 31, 32, 65. See also
Genre painting, 2, 41 Kizimba
God: ideas of, 85, 97, 107–8; illness Malinowski, Bronislaw , 131n19
caused by, 64–65; spirits and, 107 Marwick, Max, 117
Greenwood, Susan, 132n6 Medical anthropology, 15
Medicine, 25, 28–29, 35, 61–63, 88–
Handelman, Don, 132n10
89, 96, 100–104
Healer, 26, 66, 74–75, 101–2; associa-
Meyer, Birgit, 117
tions of, 69
Missionaries, 66–67, 75, 84–87,
Hunt, Nancy Rose, 132n5
127n8
Hymes, Dell, 41
Moore, Henrietta L. 117
Illness: causes of, 61–65; classification of, Miti, 63, 100–103, 109
62; Europeans causing, 65, 88–90, Mizimu, 97, 105–8. See also Spirits
106; God causing, 64–65; spirits caus- Mobutu Sese Seko, 22, 84
ing, 64, 83, 88. See also Kahenga Monograph, 8
Mukonkwa Michel Mudimbe, V. Y., 132n4
Informants, 16–17 Mufwankolo, troupe, 2
Internet, 3–5, 121–23; presence of texts Mungu: ideas of, 85, 97, 107–8; ill-
and, 4–5, 112–13, 119–20, 122 ness caused by, 64–65; spirits and,
107
Jamaa, 2, 87, 129n9, 130n10
Mwenze Kibwanga, 3, 41, 59, 129n7
Janzen, John, 56, 116
 Index

Naming: knowledge and, 30–31, 33, 96; code switching in, 125n2; social
129n8; practices of, 65–66, 80 history of, 51, 97–98
Nganga, 26, 66, 74–75, 101–2; associa- Symbolism, 26
tions of, 69
Tedlock, Dennis, 36
Nietzsche, Friedrich, 6
Tempels, Placide, 115, 130n15, 131n2. See
Nyange ya Kahenga, 27, 66, 101, 107
also Jamaa
Objectivity, 40 Texts: features of, 41–43; making of, 40,
50; as mediator, 39 as metaphor, 37; as
Pels, Peter, 117
model of culture, 37; presence of, 119–
Popular culture, 2, 15, 94
20; recording of, 41–43; transcribing
Representation and presence, 112–14 of, 43–49; translating of, 45, 49–54. 139
Ritual, 31–32. See also Closing the See also Ethnographic texts
house Theuws, T., 115, 132n3
Tradition, modernity vs., 74, 76–77, 86,
Sanders, Todd, 117
104
Shaba/Katanga, 2, 22, 59
Tshibumba Kanda Matulu, 2–3, 17, 94,
Sorcery, 5, 24, 34, 68, 96, 108
130n10
Speech events, 41–43, 45
Turner, Victor, 115
Spirits, 27–28, 31–32, 67, 85, 105–8; asso-
ciations of, 69–70, 106; ethnic iden- Virtual archive, 3, 121, 123; presence of
tity and, 83–84, 105–6 texts and, 4–5, 112–13, 119–20, 122
Swahili, Shaba/Katanga: labor and, 1–2,
Writing, 8–9. See also Commentary
126n3; as linguistic medium, 42, 43; as
medium of popular culture 2, 52, 94– Zaire, 1, 74–75
johannes fabian is an emeritus professor of cultural
anthropology at the University of Amsterdam and a member of
the Amsterdam School of Social Research.

Library of Congress Cataloging-in-Publication Data

Fabian, Johannes.
Ethnography as commentary : writing from the virtual
archive / Johannes Fabian.
p. cm.
Includes bibliographical references and index.
isbn 978-0-8223-4261-8 (cloth : alk. paper) —
isbn 978-0-8223-4283-0 (pbk. : alk. paper)
1. Ethnology—Computer network resources.
2. Ethnology—Authorship.
3. Communication in ethnology.
4. Communication and culture.
I. Title.
gn307.65f33 2008
306—dc22
2008007703