You are on page 1of 234

How Do We Know?

How Do We Know?
Evidence, Ethnography, and the Making
of Anthropological Knowledge

Edited by

Liana Chua, Casey High, and Timm Lau

Cambridge Scholars Publishing

How Do We Know? Evidence, Ethnography, and the Making of Anthropological Knowledge,
Edited by Liana Chua, Casey High, and Timm Lau

This book first published 2008 by

Cambridge Scholars Publishing

15 Angerton Gardens, Newcastle, NE5 2JA, UK

British Library Cataloguing in Publication Data

A catalogue record for this book is available from the British Library

Copyright © 2008 by Liana Chua, Casey High, and Timm Lau and contributors

All rights for this book reserved. No part of this book may be reproduced, stored in a retrieval system,
or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or
otherwise, without the prior permission of the copyright owner.

ISBN (10): 1-84718-581-9, ISBN (13): 9781847185815


Acknowledgements .................................................................................. vii

Introduction: Questions of Evidence

Liana Chua, Casey High, and Timm Lau .................................................... 1

Old and New Reflections

Marilyn Strathern ..................................................................................... 20

“We Sing About What We Cannot Talk About”: Music as

Anthropological Evidence in the Venda Region of South Africa
Fraser McNeill ......................................................................................... 36

Audio-Visual Evidence and Anthropological Knowledge

Carlo A. Cubero ....................................................................................... 58

End of the Spear: Re-imagining Amazonian Anthropology and History

through Film
Casey High ................................................................................................ 76

Pragmatic Evidence and the Politics of Everyday Practice

Ann Kelly .................................................................................................. 97

The Vodou Priest Who Lost His Spirit: Reflections

on “Modernity” and “Tradition” in a Beninese Village
Anna-Maija Aguilera Calderòn.............................................................. 118

Enmities and Introspection: Fieldwork Entanglements and Ethnographic

Emma Varley .......................................................................................... 133

Understanding Tibetan Shame and Hierarchy through Emotional

Experience in Fieldwork
Timm Lau................................................................................................ 157
vi Table of Contents

Virtual Returns: Fieldwork Recollected in Tranquillity

Lisette Josephides.................................................................................... 179

Afterword: Malinowski’s Heirs

Keith Hart............................................................................................... 201

List of Contributors ................................................................................ 210

Index ...................................................................................................... 213


This volume is based on a series of papers delivered at the conference

Questions of Evidence: Ethnography and Anthropological Forms of
Knowledge in January 2007 at Cambridge University, organised by the
editors of this volume. We would like to thank the Centre for Research in
the Arts, Humanities and Social Sciences (CRASSH), notably Michelle
Maciejewska, and King’s College for funding and hosting the conference
proceedings, as well as C-SAP for providing financial support. We are
deeply grateful to all the speakers and participants for their lively and
challenging contributions. Not all the papers given at the conference are
represented in the present volume, and we would like to thank Rita Astuti
for a keynote speech, as well as Laurence Douny, Shaylih Muehlmann,
Xiaoxiao Yan, and Andrew Babson for their contributions. We also thank
the conference discussants, Matthew Engelke, Christina Toren, Barbara
Bodenhorn, Helen Lambert, and Peter Loizos for their valuable comments
and insights on their respective panels. We are grateful to Marilyn
Strathern for allowing us to reproduce the text of her keynote speech here,
and to Keith Hart for reprising his summation role and writing the
afterword. Gratitude is also due to Jon Mair for his early involvement in
helping Timm Lau initiate the conference, and to Matthew Engelke for his
support of the conference and this volume. Finally, we are indebted to
Carol Koulikourdi, Amanda Millar, and Andy Nercessian at Cambridge
Scholars Publishing for their interest and support, without which this
volume would not have materialised.


Questions of evidence have become increasingly prominent within
anthropology in the last few years. Once an invisible pillar of the
anthropological enterprise, “evidence” is now becoming an object of
scrutiny in its own right. In the inescapably reflexive climate of much
contemporary anthropology, such a move may seem surprisingly belated,
for the accrual of evidence through fieldwork and its presentation through
written ethnography have long been instrumental to the discipline.
Regardless of their theoretical persuasion, anthropologists have invariably
relied implicitly on the quality of their evidence as the basis of their
epistemological and rhetorical authority. That evidence is central to the
production of ethnography should thus be beyond debate. Yet, its virtually
unquestioned centrality has arguably also been its veil. As we suggest
below, evidence was for a long time treated simply as the factual basis of
ethnographic analysis: the “object” that stood outside the “argument”
(Hastrup 2004), the means to a more prominent end. Even as the
postmodernist debates of the 1980s paved the way for a more reflexive
approach to ethnography, and rubrics like “writing culture” and
“knowledge practices” directed anthropologists’ attention towards the
conditions in which anthropological knowledge arose, evidence itself
remained a non-topic.
In recent years, however, the situation has altered considerably, with
the contours of a new discussion taking evidence as its central theme being
shaped by a growing number of anthropologists. During this time, a
number of articles (e.g. Csordas 2004; Hastrup 2004) and a special issue of
the Journal of the Royal Anthropological Institute (Engelke et al. 2008)
have yielded analytical, theoretical, and methodological insights into its
place and significance within anthropology. Evidence, in these works,
takes multiple forms: more than merely “facts” or “data”, it can equally be
a trope (Csordas 2004) or an experience (Hastrup 2004), a tool or a
condition (Engelke 2008, S5). The intellectual genealogies of these
2 Liana Chua, Casey High, and Timm Lau

approaches extend much further back than their thematic novelty suggests,
however, for the questions they raise strike at the very heart of
anthropological practice. Interrogating the concept of “anthropological
evidence” concomitantly demands interrogating what anthropology itself
is all about: its objectives, methods, parameters, ambitions, doubts. If
evidence, as Csordas notes, has to be “of and for something” (2004, 475),
then those who would discuss it must inevitably work out what that
“something” is.
That process of working things out is what interests the contributors to
this volume, as well as the participants in the conference from which it
derives. “Questions of Evidence: Ethnography and Anthropological Forms
of Knowledge” was held over two days in January 2007 at the Centre for
Research in the Social Sciences and Humanities (CRASSH) and King’s
College at the University of Cambridge. Focusing on ethnography and the
processes by which anthropological evidence and forms of knowledge are
created, it featured a diverse range of papers that explored “evidence” in
and through the British broadband industry, films of the Caribbean,
dreams and emotions, Cucapá indigenous curse-words, vodou goddesses,
and pragmatic clinical trials, to name but a few. Inherent to them all was a
strong reflexive impulse, whether personal or disciplinary; an ambition to
not simply talk about evidence, but to examine the circumstances that give
rise to it and make it recognisable as such.
The wide-ranging response to our call for papers was, serendipitously,
unanticipated. Our initial intention had been to hold a one-day workshop
for PhD and postdoctoral anthropologists to critically evaluate the use of
encompassing explanatory models, and the role of small-scale
ethnography in engaging with them. From this we planned to meander
down various evidence-related paths, asking where it might be located,
what forms it might take in an age of multimedia and globalisation, and
how it might be rethought with the aid of interdisciplinary approaches. The
proposals and comments we received, however, markedly widened the
conference’s scope and interests.
Significantly, the participants came from both ends of the anthropology
career spectrum: while most were young anthropologists and graduate
students, many others were experienced academics with long-standing
reflexive interests in the discipline. The paper-presenters were connected
by the fact that they had either recently done fieldwork or were writing
about fieldwork and processes of anthropological knowledge-making.
Interpreting “ethnography” in its broadest sense, these participants—many
of whom feature in this volume—were grappling with issues of immediate
import to the practice and ethos of anthropology. Their papers are
Introduction: Questions of Evidence 3

accordingly contemplative, but also tinged with rawness; a sense of having

only just worked through the ethnographic process, with all its flukes and
Placed in dialogue—as they were during the conference—with Marilyn
Strathern’s keynote address and Keith Hart’s afterword, these papers
comprise a volume which is grounded in anthropological experience but
also methodologically, analytically, and theoretically ambitious in
exploring evidence in anthropology. While they share the epistemological
and philosophical concerns of recent writings on “how evidence works in
and for the discipline in its generation of knowledge” (Engelke 2008, S3),
they address them through specific case studies, exploring how evidence is
located, elicited, and utilised in ethnography. In the process, they remind
us that more than being a topic for philosophical-scientific discussion,
evidence fundamentally emerges in and through what ethnographers do. In
this respect, the composition of our contributors is particularly apt: the
majority have just begun to feel their way through the “doing” process,
and the others have “done” and reflected on it for much longer. The result
is a volume which extends the nascent debate on evidence in anthropology
in the direction of ethnographic practice and the generation of specifically
anthropological forms of knowledge.

Evidence and the changing object of anthropology

Before moving on to some key themes in the volume, it is worth
addressing what we mean when we talk about evidence. To unequivocally
define it as an object with specific parameters and features would be futile:
as the following papers reveal, evidence is highly contingent on the
contexts in which it is generated. Our point in this section is thus to trace
the outline, however faint, of the concept of evidence vis-à-vis existing
debates on the topic and in anthropology more generally. It is important to
note that although evidence is our explicit focus, the questions we raise
here are not new. In exploring them, we encounter our disciplinary
ancestors who, by interrogating and redefining anthropology’s methods,
epistemologies, and interests, also reshaped our modes of apprehending
evidence. This has resulted in a proliferation of evidentiary conventions
both at the core of the discipline and along its borders, rather than a clear
consensus on what they should entail. Studying these multiple and
sometimes competing conventions—with or against which anthropologists
have had to work—provokes questions of great reflexive import. These do
not merely centre on what evidence is, but on how its very nature and
guise have evolved along with the objects of anthropology.
4 Liana Chua, Casey High, and Timm Lau

A glance at the history of anthropology is instructive in this regard.

During the formative years of the discipline, “primitive societies” became
the objects of anthropological study; whether as evidence of earlier stages
in human evolution, or of social and cultural variability around the world.
As various forms of functionalism came to dominate early-twentieth
century anthropology, seemingly exotic practices discovered through long-
term fieldwork in far-off places became evidence of how particular
societies operated as distinct wholes (Kuper 1988). Over time,
anthropology moved away from both its original object of study
(“primitive society”) and its founding theories (evolutionism and
functionalism), while retaining its dominant methodology of participant-
observation, advocated most prominently by Malinowski and Rivers
(Grimshaw and Hart 1995). But while “fieldwork” has lingered at the core
of contemporary anthropological research and disciplinary identity,
anthropology’s objects, theories, and scope have undergone multitudinous
changes. In the last century, anthropologists have greatly expanded and
diversified their objects of study, which have come to include the
structuring tendencies of the human mind, symbolic meanings, historical
processes, and ethnographic knowledge production itself. These
developments, as well as the repertoire of methods accompanying them,
have significantly (if sometimes silently) reshaped our understandings of
evidence as both a tool and a concept, while also influencing our evidence-
gathering practices.
One theoretical development which has had a particularly significant
impact on anthropological engagements with evidence is the emergence of
reflexivity as an intrinsic component of ethnography. Its entrance into
mainstream anthropology in the 1980s took place simultaneously with the
rise of postmodernist approaches, which issued significant challenges to
the authority with which previous generations of anthropologists had
appeared to speak. Such challenges have highlighted the role of
ethnography as both producer and form of anthropological evidence,
thereby turning ethnographers themselves into part of the object of study.
The reflexive nature of much postmodernist anthropology has on more
than one occasion come under fire for turning “self-conscious authorial
positioning [into] an authenticating device” at the expense of “scientific
standards” (Roth 1989, 557). In the context of this discussion, however,
some circumspection is necessary.
The growing prominence of reflexivity within anthropology has in no
way diminished the conceptual importance of evidence to contemporary
anthropological practice (Engelke 2008, S7). Regardless of the prevailing
theoretical and methodological climate, the anthropologist’s ability to
Introduction: Questions of Evidence 5

marshal convincing arguments depends upon the construction of a body of

evidence to support them. The idea of evidence thus remains foundational
to the organisation and dissemination of anthropological knowledge as a
crucial and expected feature of good analysis. What has changed, or at
least widened, is an understanding of the possible forms that evidence can
assume—including those which blur the ostensible distinction between
“self” and “science”. As some of the chapters in this volume reveal,
personal engagements in ethnography may become forms of evidence no
less rigorous than those produced by a supposedly “detached” observer.
Instead of collapsing the distinction between self and other or subject and
object, they show that we may take epistemology and the production of
anthropological knowledge themselves as objects of enquiry and evidence
In light of such developments, it is unsurprising that the nature of
evidence within the discipline has been undergoing increasing reflexive
scrutiny. As mentioned earlier, recent anthropological writings have
approached the topic from different directions, variously exploring
evidence as an epistemological phenomenon inseparable from
anthropological experience, a rhetorical and discursive device, or a tool.
All, however, are united in emphasising its relational quality, the fact that
it does not stand alone as an objective phenomenon, but must be
recognised, generated, and evoked within a larger network of connections.
And yet this network does not branch out infinitely; at some point it has to
be “cut” (Hastrup 2004, 456, after Strathern 1996) in order to separate
potentially limitless “data” from what counts as “evidence” (Csordas 2004,
477). But if not all data become evidence, not all evidence merely derives
from data. As the following chapters reveal, evidence of and for
ethnographic analysis does not only arise from the ostensibly “objective”
information amassed in the physical and temporal space of fieldwork, but
also from memories, reflections, emotions, writing, reading, creative and
documentary processes, and institutional demands. Ethnography, in this
sense, is multi-sited (Marcus 1995) in more ways than one.
It is also in this sense that we suggest the papers in this volume make
their strongest contributions to discussions of evidence in anthropology.
While most voices in the “evidence debate” have thus far addressed the
philosophical, analytical, and methodological implications that a focus on
evidence might have for anthropology, the contributors to this volume
concentrate on the processes through which they found, created, and used
evidence during their fieldwork and writing. To claim that evidence is
what they make of it is not a statement of impudent ambiguity, but a
reminder of the situatedness of their accounts. “Evidence” was not simply
an abstract ideal that guided our contributors’ ethnographic explorations,
6 Liana Chua, Casey High, and Timm Lau

but something that emerged amid the exigencies and contingencies of

fieldwork and post-fieldwork experiences. In the process, many of them
came face-to-face with evidentiary conventions and expectations within
and without anthropology. The papers they have produced are, in that
regard, explorations of the role of evidence in their attempts to “get on”
with anthropology (Metcalf 2002). In the following pages, we examine
these contributions more closely through three key themes around which
the volume coalesces: the evidentiary forms that lie “beyond text”, the
relation between ethnography and theoretical models, and the role of
emotions and reflexivity.

What counts as evidence?

In anthropology, the question of what counts as evidence has been
significantly complicated by the expanding variety of methods and sources
on which anthropologists might draw. The chapters in this volume attest to
this point, engaging directly with the epistemological implications of the
often diverse and contrasting forms of evidence that they bring to their
ethnography. Rather than treating music, film, history, and personal
experiences as raw data, they critically examine how these sources can be
used to advance particular ethnographic claims that may both contradict
and corroborate our informants’ everyday discourses as well as dominant
anthropological models.
Carlo Cubero’s chapter on ethnographic filmmaking on the Caribbean
island of Culebra explores how film and the process of filming provide
distinctive forms of ethnographic evidence. On the one hand, film footage
as an end product can be seen as material evidence of social relations
observed in fieldwork—an argument advanced by Margaret Mead and
other advocates of ethnographic film who focused on audiovisual media’s
capacity to create an objective and durable record of “facts”. On the other
hand, Cubero examines how the actual experience of filming during
fieldwork became an important “purveyor” of ethnographic knowledge.
He describes how the personal, even physical, process of recording
musical performances challenged his ideas about Culebra in important
ways that informed his subsequent ethnographic writing: even footage
which he initially shot as a “visual record … for future historical
reference” eventually led him to understand an unfolding controversy
about development on the island in new ways. His paper reveals the dual
role of film in recording events and processes as well as allowing
anthropologists to reformulate and rearticulate the questions they bring to
ethnographic writing.
Introduction: Questions of Evidence 7

Cubero’s paper raises questions, fundamental to visual anthropology,

about the relationship between visual and textual forms of representation
(Hastrup 1992; MacDougall 1998). In particular, he explores how these
forms of ethnography constitute, rather than simply represent, different
kinds of evidence in his research. By focusing particularly on the
subjective experience of his dual role as filmmaker and ethnographer,
Cubero demonstrates how visual and other non-textual media allow
anthropologists to redefine what they take as evidence and how they use it.
Most importantly, he suggests that audiovisuals do not merely constitute
evidence of “historical facts” in the positivistic sense implied by previous
generations of ethnographic filmmakers. Rather, his experience of filming
during fieldwork created an equally if not more important form of
ethnographic evidence. As with other chapters in the volume (Josephides,
Lau, Varley), we see that such experiences and the knowledge they create
are no less important or rigorous types of evidence than formal interviews,
fieldnotes, or video recordings.
Whereas Cubero explores the experiential and material value of film as
evidence, Fraser McNeill’s paper suggests that performances themselves
are also valuable forms of ethnographic evidence which enable the
anthropologist to deal with practices and knowledge normally kept hidden
from the public sphere. In this context, music becomes a mode of
evidence, revealing knowledge otherwise unavailable to anthropologists
who rely solely on discursive and textual sources. McNeill examines the
role of music as an important medium through which HIV/AIDS
educational messages are communicated in the Venda region of South
Africa, where AIDS is “bound to webs of secrecy and suspicion that
prohibit its inclusion in open conversation”. Since the disease’s
transmission is enmeshed within witchcraft accusations, silence and
avoidance of the topic in everyday life protect individuals from threats of
guilt by association with a death. Music, however, has become a key site
of HIV/AIDS education due to its wider cultural role as a legitimate and
safe medium for the communication of ritual knowledge.
McNeill’s paper illustrates that music and other performances beyond
everyday interactions and speech can become fruitful objects of
anthropological enquiry precisely where other evidence is inaccessible.
For McNeill, music allows a new perspective on local attitudes about the
disease, revealing “the hidden dimensions that remain elusive in public,
and often private, arenas”. More than a supplementary source of
ethnographic evidence that allows anthropologists to get closer to the
“truth”, music in the Venda context reveals a distinct form of knowledge
production not normally accessible to local people or ethnographers. Like
8 Liana Chua, Casey High, and Timm Lau

Cubero’s exploration of film and filmmaking as ethnographic evidence,

McNeill’s paper suggests that diverse media beyond everyday discourse
may provide evidence to question previous ethnographic assumptions and
contribute to new anthropological knowledge. However, the paper also
cautions researchers not to assume music-based evidence to be
representative of a particular group. This raises the issue, also addressed
by High and Calderòn, of how encompassing models generated by
anthropological research are as much a result of what the researcher adopts
as evidence as they are representative of “social reality” among the group
Casey High’s paper broaches the question of how expanding the scope
of ethnographic evidence allows us to rethink dominant analytical models
in regional studies. In particular, his paper draws from historical sources
and popular cinema as evidence for reassessing anthropological claims
about contemporary indigenous Amazonian ontologies and regional
history. By juxtaposing his ethnographic research on historical narrative in
Huaorani communities in Ecuador with a recent North American feature
film about past Huaorani violence, High suggests that while indigenous
Amazonian ideas and practices bear certain elements of a generic model of
“Amazonian” cosmology, they are equally constituted within a wider
social and historical context that extends well beyond indigenous
High describes how past violence is central to both indigenous oral
histories and Western characterisations of the Huaorani during much of the
twentieth century. Partly because of their famous history of spear-killing
raids, the Huaorani have attracted much attention from Christian
missionaries since the 1950s. This culminated in the widely-publicised
killings of five North American missionaries while they were attempting
to establish a mission in Huaorani territory in 1956. Fifty years after this
event, and in the aftermath of an intensive missionary campaign in the
region, a Hollywood film was released based on the history of Huaorani
violence. High explores how the imagery of Christian martyrdom
suggested in the film and in missionary texts can also be found in
historical narratives told by Huaorani people. His paper engages
missionary discourses and images as a form of evidence that allows him to
place Huaorani notions of victimhood within wider historical and inter-
cultural relations. He argues that the dominant Huaorani self-identification
as essential victims not only reflects the generalised Amazonian model of
“ontological predation” advanced in the regional literature (Viveiros de
Castro 1996), but must at the same time be seen as the product of a
particular history of interaction between indigenous people and missionaries.
Introduction: Questions of Evidence 9

High concludes that combining evidence generated in fieldwork and

the archives with less localised sorts of evidence, such as popular cinema,
can benefit anthropologists who seek to place their ethnographic accounts
within a wider social and historical frame. As with McNeill’s chapter, we
see that engagement with multiple forms of evidence holds significant
potential for questioning dominant analytical models within anthropology.
The contributions of Cubero and High illustrate in different ways the
important potential of visual media in ethnographic and anthropological
projects. Both papers point beyond antecedent conceptions of film as an
objective representation of “reality” and demonstrate that visual media can
be combined with more conventional sources of ethnography to produce
new anthropological knowledge. In their own ways, they illustrate that
film and other visual media are more than mere aids to ethnographic texts
(MacDougall 1998).
More generally, these papers reveal how anthropological arguments,
whether they concern claims about ethnographic “realities” or abstract
theoretical models, are pivotal in determining what counts as ethnographic
evidence. As Hastrup has argued, anthropological knowledge is
“relational” insofar as it is “defined as much by epistemology as by the
ontology of its object of interest” (2000, 456). Evidence thus does not exist
as an objective “substance” or “raw data” independent of the
ethnographer’s questions and arguments. New kinds of argument
necessarily require new kinds of evidence—but also, as Josephides (this
volume) reminds us, different ways of (re)interpreting it. The ethnographic
cases explored in this volume attest to this relational quality of evidence,
recognizing how new and multiple forms of evidence consequently lead to
new forms of anthropological knowledge.

Evidentiary conventions and encompassing models

As Engelke notes, evidence is not an easy topic for discussion amongst
anthropologists, not least because of the lack of a “well-developed
language” with which to do so (2008, S3). Allied to this is the fact that
most “evidentiary protocol[s]” (ibid., passim) within the discipline have
remained implicit despite their decisive role in defining and generating
what is recognised as evidence. The work of Engelke and others
mentioned earlier, however, reveal a growing disciplinary reflexivity
which has not only focused our analytical attention on the concept of
anthropological evidence, but also on the conventions by which it is
framed, validated, and articulated. More than asking what constitutes
evidence, we must also consider how it is generated and recognised as
10 Liana Chua, Casey High, and Timm Lau

such—and more intriguingly, how its very definition might shift over time
and space.
Marilyn Strathern’s keynote address deals with these questions by
exploring two comparative methods through which anthropologists and
others “reduce, digest and otherwise summarise information in such a way
as to yield a yardstick or measure by which other information can be
judged, proven or verified.” On the one hand, she describes a
“mathematics of encompassment” frequently found in Euro-American
(and anthropological) settings, whereby persons, examples, and
arguments, among other things, are drawn together and aligned as
equivalent (comparable) entities within a single framework. On the other,
expanding on her earlier work (2006), she explores “analogic reasoning”
as a mode of comparison which involves “elucidat[ing] one thing by
reference to another”, thereby maintaining each element’s distinctiveness.
She illustrates the difference between these two modes through an array of
case studies, from British university promotions exercises to courts of law
in Papua New Guinea, using these particular instances—themselves
analogies of each other and of anthropological practice—to elucidate our
understanding of how anthropological evidence may be constituted.
The resonance of Strathern’s chapter derives partly from the fact that
like the analogy, it offers no resolution to the incommensurability of
different evidentiary conventions. For her, such divergences are better to
think with than to solve problems with. For other contributors, however,
working with or against certain evidentiary conventions engendered
epistemological problems which sometimes demanded difficult solutions.
In some cases, this involved balancing the demands of different
representational forms, such as film and text (Cubero, High). Others
encountered dilemmas over how to reconcile pressing but relatively
unconventional forms of evidence that emerged through—and as—
personal experience, with mainstream organisational and rhetorical tropes
in ethnographic writing (Lau, Varley). In explaining how they worked
through these dilemmas, each contributor simultaneously lays bare the
often unspoken evidentiary allowances and expectations underpinning
anthropological writing and practice.
These examples reveal the sheer difficulty of defining a coherent set of
anthropological evidentiary conventions. Yet in a world of increasing
interdisciplinarity and “applied” research, anthropology, like many other
disciplines, is coming under mounting pressure to articulate, if not define
precisely, the epistemological and methodological qualities that set it
apart. Whether or not anthropological responses to these recent demands
have been productive—and indeed whether the boundary-crossing
Introduction: Questions of Evidence 11

impulses behind them have entirely desirable consequences (cf. Strathern

2006)—any discussion of evidentiary conventions within anthropology
would be incomplete without some reference to their status vis-à-vis other
parties. Indeed, Engelke (2008, S12) posits that anthropology’s growing
interdisciplinary and political connections may be causally related to the
intensification of anthropological interest in evidence. No longer
concerned exclusively, if it ever was, with studying the differences
between “us” and “them”, anthropology has now had to extend its
attention to the “commons and borderlands” (Strathern 2004) of
disciplines, institutions, and other groupings.
Almost inevitably, questions and controversies over evidence surface
at these junctures. Evidentiary conventions and expectations are not
everywhere the same, even if they often overlap and intersect. This is
especially true in many Western knowledge economies, of which
anthropology is historically part, and in which the concept of evidence
possesses significant, if varied, import. In an age of “evidence-based”
policy and medicine, for example, what counts as evidence can influence
anything from the allocation of public funds to new environmental laws
for reducing carbon emissions. Squaring such conceptions of evidence
with anthropological ones can be both trying and illuminating. While
anthropologists have long been used to dealing with and even eliciting
difference in their engagements with “other” cultures, working with(in)
disciplinary or institutional settings which have their own distinctive
evidence-making structures can generate other sorts of tensions.
Ann Kelly’s chapter, which derives from her research with a UK-based
arthritis research charity on a trial osteoporosis screening programme,
brings some of these issues to the fore. Reflecting on her fieldwork, she
admits to feeling “woefully inadequate” upon realising that her
ethnographic findings would not produce the “rich empirical detail” that
she expected would benefit clinical practice. Instead, “the phenomena on
which [she] was reporting belonged to a different order than routine clinic
research procedure and its practical effects.” The evidence she produced as
an anthropologist and the evidentiary conventions of the clinical trials she
studied, in other words, simply lay on different planes. One could not feed
constructively into the other in a “real-life” clinical setting.
Kelly’s response to such apparent epistemological and methodological
incommensurability is instructive when juxtaposed with Strathern’s
keynote address, which recounts several similar instances of mismatching
conventions. Like Strathern, she turns to analogy rather than
encompassment as a means of addressing the tension between divergent
sets of conventions. Clinical trials and ethnography, she suggests, might be
12 Liana Chua, Casey High, and Timm Lau

regarded as “analogous methodological strategies”, each reliant in

different ways on concepts of “everyday practice”. Rather than leaving
them thus as mutual illuminators, however, she further argues that “the
efforts made by the designers of clinical experiments to transform
evidence of ‘everyday practice’ into evidence for ‘everyday practice’”
hold analytical and methodological lessons for anthropologists working in
any field. By adopting this perspective, she adds a new dimension to
understandings of both anthropological evidence and anthropology itself,
thereby provoking a pressing question: what if we were to admit a form of
ethnographic evidence that does not simply represent reality, but actively
shapes and becomes part of it? Put differently, what happens when model
and reality are collapsed?
Kelly’s ruminations occupy familiar territory for many anthropologists.
The relation between anthropological models and frameworks and the
experiences and discourses of the people we work with has variously been
a source of consternation, intrigue, and inspiration. Every “gift” (Mauss
1990) or “exemplary centre” (Geertz 1980) has its ethnographic roots; but
it only takes a glimpse at their numerous offshoots to realise how quickly
they too become the theoretical basis for further ethnography (see
Josephides, this volume). On this point, we return to a central theme of our
original call for papers: the relation between encompassing explanatory
models and ethnographic evidence from “small places” (Eriksen 1995).
As suggested earlier, evidence is only recognised as such in relation to
other things: questions, theories, disciplines, experiences, persons, and
epistemological frameworks, all of which provide the conditions for
transforming data into evidence. In the process, these conditions are
shaped and reshaped—even created anew—by what is identified and
harnessed as evidence. But if ethnographic evidence generates analytical
and theoretical models, the reverse is also true, for it may also be used to
legitimate and perpetuate them in a variety of contexts quite dissimilar to
the ones from which they stemmed. Underpinning this, to use Strathern’s
phrase, is a certain logic of encompassment, by which specific persons and
events become incorporated and flattened within a broader analytical
framework. To a degree, this reflects anthropology’s perennial predilection
for comparison, which prevents the discipline from getting mired in
innumerable vignettes too particular to be of analytical use. But how far
can we go before the opposite occurs, and our ethnographic data only gain
salience as evidence for the validity of broader analytical models?
This is the question with which Calderòn grapples in her paper, which
provides an in-depth account of the rise and fall of an ambitious vodou
priest in a village in Southern Benin. Eschewing prevailing models of
Introduction: Questions of Evidence 13

“tradition” and “modernity” which have generally depicted vodou worship

as “a convenient study of the effects of globalisation on a local level”,
Calderòn presents the story of Sofo as a situated counterpoint to what she
sees as an anthropological approach too generically tied to a “meta-
narrative” of modernity (Englund and Leach 2000). In the process, she
attempts to “render analytical” the categories and discourses used by her
informants to discuss the practice of vodou, and more specifically,
relations with its spirits. Rather than being “subjective and partial
representations” (Viveiros de Castro 1998, 478) of what some
anthropologists might see as the objectively “real” phenomena of
globalisation or modernity, such categories become the very basis of her
ethnography: they are the real thing. Calderòn thus elides the difficulties
inherent in balancing models with ethnographic realities, not by recourse
to analogy, but by creating her own micro-narrative which places her
ethnography in constructive and critical dialogue with anthropology’s
broader concerns and models.

Emotions, experience, and anthropological knowledge

As noted above, one may be justified in saying that reflexivity has
grown on anthropologists: reflexive approaches have played an important
part in redefining the discipline’s methods, strategies of representation,
and theories. We also note that the very concern with evidence constitutes
a reflexive moment in both anthropological endeavours and the discipline
as a whole. Amidst these wider processes stands old-fashioned social-
scientific reflexivity as defined by The Oxford English Dictionary: “that
which turns back upon, or takes account of, itself or a person’s self, esp.
methods that take into consideration the effect of the personality or
presence of the researcher on the investigation”. In this volume, three
chapters in particular make significant contributions to discussions of
reflexivity in anthropology, which we might define more narrowly as
recognition of the ethnographer’s subjectivity in the formation of
anthropological knowledge. In combination, they address the question of
what may constitute evidence. Rather than only envisioning evidence in
the form of “fieldwork data” recorded in fieldnotes, these contributions
demonstrate that evidence is created relationally and contextually through
subjective processes that constitute fieldwork and ethnographic writing.
Emma Varley’s contribution argues that much avowedly reflexive
anthropology has remained “constrained by a quiet political correctness”
which obscures the more problematic aspects of our selves in fieldwork.
Varley argues that if our biographical background, personal interests, and
14 Liana Chua, Casey High, and Timm Lau

values influence our ethnographic production, then “our work inevitably

reflects or integrates not only positive aspects of our selfhood but also the
more problematic or morally uncertain”. Yet, anthropological tropes of
reflexivity have been purposefully incomplete, “smoothing away many of
our own rough edges”, stressing instead early misunderstandings and faux
pas which in the long run lead to more enlightened ethnographers and
ethnography. Her account of her own fieldwork experience serves as a
contrast to what she critically describes as “being generous to ourselves”
in only partially reflexive accounts.
Varley carried out fieldwork in Gilgit, Northern Pakistan, where she
researched the impact of Shia-Sunni conflicts on the provision of health
services to local women. Her research took place amidst two simultaneous
zones of strife: communal fighting between local Sunni and Shia factions
which escalated heavily during her fieldwork, and communal infighting
between herself and a neighbouring Gilgiti Sunni family. Varley’s account
does not shy away from those moments and dynamics of fieldwork that
include “personal wrongdoing or malintent”, and reveals how her
involvement in interpersonal battles shaped relationships with her
informants during fieldwork.
Significantly, these circumstances and her actions partially formed her
ethnographic perspective and the position from which she was able to
produce insights into Gilgiti sociality. On the one hand, her own conflict
with members of the Sunni community, centring on her locally contested
marital position, elicited insights into how antagonistic relationships are
conducted within the community. On the other hand, in the context of
Varley’s doctoral research, her own story as an embattled wife brought out
related stories from her informants. These provided evidence to question
her initial assumption about the communal conflict’s centrality,
emphasising instead the importance of domestic relationships in
determining whether and how Gilgiti women took up local health services.
Varley demonstrates that her ethnographic insights were centrally enabled
by her personal tribulations, and thus shows that the challenging events of
her fieldwork carried highly significant evidentiary potential for
anthropological analysis. She stresses that, if allowed their space as
potential evidentiary platforms, such dilemmas within the experience of
fieldwork can be important sites of cultural understanding. Her chapter,
then, makes an important and original contribution to discussions of
reflexivity in anthropology by urging us to include its—and hence our
own—more problematic sides.
Where Varley’s chapter suggests the importance of personal fieldwork
experiences, however emotionally troubling, to the generation of
Introduction: Questions of Evidence 15

ethnographic evidence, Timm Lau’s contribution treats emotions as

specifically sensitising experiences in the making of anthropological
knowledge. Lau presents a key episode during his fieldwork with Tibetans
in the diaspora in Northern India, during which he crossed local
hierarchical boundaries. He shows that this event engendered an intense
emotional experience produced by the gaze of others which emphasised
relative social positions. Lau’s argument could be understood as an
exercise in analogic reasoning (Strathern, this volume), in which his own
emotional experience provided a bridge or heuristic relation that made it
possible to understand the importance of shame (ngo tsha) in his
informants’ lives. The resulting awareness of local contexts of shame
facilitated his analysis of the importance of public visibility and ngo tsha
for the socialisation of children, behaviour in public gatherings, and the
constraints on romantic relationships inside Tibetan refugee settlements.
Lau demonstrates that Tibetan ngo tsha is centrally connected to the gaze
of others, leading to a powerful “emotional aesthetic of shame” for his
Tibetan informants. His resulting analysis suggests that “the processes by
which Tibetan social actors draw upon ngo tsha in social interaction at the
same time reproduce the structural features of the social hierarchy”.
The intrinsic connection of Lau’s emotional experience during a key
fieldwork episode to awareness of his Tibetan environment and
hierarchical social position is reflected in his analysis of the importance of
the gaze of others, and the way in which meaning is attached to notions of
shame vis-à-vis hierarchical superiority. Lau’s exploration of Tibetan
concepts of shame addresses how Tibetan social hierarchy is ingrained on
the ground, in Tibetan sociality. His chapter demonstrates that new kinds
of evidence in ethnography, in this case elicited by the sensitisation of
emotional experience, can be highly productive and significant in
answering lingering anthropological questions. As with Varley, the
starting point of Lau’s exploration is his own emotional experience, which
is structurally produced through its local setting, thus making it possible to
“see and connect pieces of ethnographic evidence in new ways”. Lau
advances the notion that in anthropological analysis, the ethnographer’s
own emotional experience can be of significant evidentiary value through
its sensitising capacity, as a portal for evidence of the meaning carried by
local emotions.
Both Varley’s and Lau’s chapters demonstrate that attention to
reflexivity and emotional experience is not about “studying ourselves”, but
about how anthropologists’ own subjectivity counts as one of many
sources of evidence upon which they may draw in fieldwork. Lisette
Josephides’ chapter adds to this a longer term perspective on ethnographic
16 Liana Chua, Casey High, and Timm Lau

fieldwork and evidence. First, she discusses what happens after fieldwork,
by way of what she calls “virtual returns”, when evidence is produced
away from the field through processes of contemplation and recollection
that involve fieldnotes, emotions, and other elements. Secondly, she
reflects on her own long-term trajectory as a social anthropologist,
demonstrating shifts in how she collected and used evidence in two books
from distinct stages in her career. Her paper gives a nuanced and in-depth
discussion of reflexivity and subjectivity which reveals how her individual
research prerogatives influenced or determined what she took as evidence.
Josephides introduces the concept of virtual returns through
Wordsworth’s notion of “recollect[ions] in tranquillity”, in which
memories of events are contemplated in such a way that the essence of the
experience remains in one’s consciousness as other aspects of it fall away.
She demonstrates that recollection in times away from the field can be
enriched in numerous ways—whether through memories and dreams of
emotional importance, or books read after fieldwork. Josephides makes an
important contribution in showing that these influences “in fallow periods”
may lead to different ways of conceptualising evidence, long after the
initial fieldwork is done. In this context, her own material suggests with
great openness that fieldnotes are an edifice we construct in fieldwork
which has the potential to contain evidence outside and beyond our initial
reading. Like the ethnographic film rushes described by Cubero, such
entities contain more information than that which they were originally
intended to convey. For Josephides, fieldnotes, while being constructed in
the phenomenological context of the “world of the ethnography” and its
expanding horizons, can transcend as well as contain the emotional
context in which they are created.
The chapters by Varley, Lau, and Josephides share common ground in
working to integrate the ethnographer’s subjective experience into the
evidentiary realm—the realm of that which is admissible to “think with” in
conventional anthropological analysis. All three are also highly reflexive
pieces for anthropology itself. Varley admonishes anthropologists to own
up to their shortcomings and in turn realise the latter’s evidentiary
potential. Lau ends his chapter on an analogy between anthropological and
Tibetan social practice, pointing out that in both cases the treatment of
shame secures present hierarchies. Josephides’ contribution goes even
further by turning the epistemological itself into an object of
anthropological enquiry.
Introduction: Questions of Evidence 17

The contributions to this volume collectively raise the question of how
we might conceptualise evidence in fieldwork, ethnographic writing, and
the processes of knowledge production within anthropology. The chapters
provide ethnographically grounded examples of how multiple forms of
evidence—including film, music, emotions, and the very processes of
creating ethnography—allow and indeed demand a rethinking of
disciplinary conventions and analytical models within anthropology. The
inevitable corollary of this process, however, is the need to rethink our
understandings of anthropological knowledge and practice. The often
unspoken evidentiary protocols which determine what and how
anthropologists know (or claim to know) are inextricably tied to shifting
conceptions of the very object and scope of anthropology. As discussed
earlier, what counts as evidence has been the very locus of change within
the discipline. We suggest that by the same token, future shifts in
anthropological theory and methodology will be informed by changing
conceptions of the nature of anthropological evidence. In making visible
and giving analytical purchase to the shifting evidentiary basis of the
discipline, the papers that follow offer glimpses of the mutable and
multifarious shapes that anthropological knowledge takes.
A fundamental contention of this introduction is that discussions of
evidence are inexorably linked to the ethnographer: “how we know” is
deeply entrenched in “who we are”. By this we do not mean to (re)invoke
a navel-gazing reflexivity that turns the anthropologist into the object of
study. Instead, we make the case that ethnography is ultimately the
product of a historical, social, and personal assemblage which includes not
only the ethnographer’s person, but also one’s intellectual background,
institutional demands, conceptual genealogies, and relational quirks within
and beyond the field, to name but a few. Exploring the importance of
evidence in current and future anthropologies thus means critically
considering the role played by anthropologists in creating rather than
simply uncovering it.
This notion is not alien to contemporary anthropology. Indeed, to the
extent that anthropology is now widely understood as a written mode of
representation (Clifford and Marcus 1986), it has become somewhat
axiomatic. In critically thinking through the concept of evidence, however,
the chapters in this volume provoke new insights into other aspects of
anthropological knowledge-making in which the anthropologist is
invariably entangled. Such insights can both supplement and nuance
contemporary debates on the nature and practice of anthropology, by
18 Liana Chua, Casey High, and Timm Lau

highlighting the ways in which ethnography holds new promise for

questioning the potentially procrustean power of abstract models in
anthropology and beyond.

Clifford, James, and George E. Marcus. 1986. Writing culture: The poetics
and politics of ethnography. Berkeley, Los Angeles, and London:
University of California Press.
Csordas, Thomas. 2004. Evidence of and for what? Anthropological
Theory 4 (4): 473-480.
Engelke, Matthew. 2008. The objects of evidence. In The objects of
evidence: Anthropological approaches to the production of knowledge,
ed. Matthew Engelke. Journal of the Royal Anthropological Institute
Special issue:S1-S21.
Englund, Harri, and James Leach. 2000. Ethnography and the meta-
narratives of modernity. Current Anthropology 41:225-248.
Eriksen, Thomas H. 1995. Small places, large issues: An introduction to
social and cultural anthropology. London: Pluto Press.
Geertz, Clifford. 1980. Negara: The theatre state in nineteenth-century
Bali. Princeton: Princeton University Press.
Grimshaw, Anna, and Keith Hart. 1995. The rise and fall of scientific
ethnography. In The future of anthropology: Its relevance to the
contemporary world, ed. Akbar S. Ahmed and Cris Shore. London:
Hastrup, Kirsten. 1992. Anthropological visions: Some notes on visual and
textual authority. In Film as ethnography, ed. Peter I. Crawford and
David Turton. Manchester: Manchester University Press.
—. 2004. Getting it right: Knowledge and evidence in anthropology.
Anthropological Theory 4 (4): 455-472.
Henare, Amiria, Martin Holbraad, and Sari Wastell. 2007. Introduction to
Thinking through things: Theorising artefacts ethnographically, ed.
Amiria Henare, Martin Holbraad, and Sari Wastell. London:
Kuper, Adam. 1988. The invention of primitive society: Transformations
of an illusion. London and New York: Routledge.
Marcus, George E. 1995. Ethnography in/of the world system: The
emergence of multi-sited ethnography. Annual Review of Anthropology
Mauss, Marcel. 1990 [1925]. The gift: Forms and functions of exchange in
archaic societies. Trans. W.D. Halls. London: Routledge.
Introduction: Questions of Evidence 19

MacDougall, David. 1998. Transcultural cinema. Ed. Lucien Taylor.

Princeton, NJ and Chichester: Princeton University Press.
Metcalf, Peter. 2002. They lie, we lie: Getting on with anthropology.
London and New York: Routledge.
Roth, Paul A. 1989. Ethnography without tears. Current Anthropology 30
(5): 555-569.
Simpson, John and Edmund Weiner. 1989. Oxford English Dictionary. 2nd
ed. Oxford: Clarendon Press.
Strathern, Marilyn. 1996. Cutting the network. Journal of the Royal
Anthropological Institute, n.s., 2:517-535.
—. 2004. Commons and borderlands: Working papers on
interdisciplinarity, accountability and the flow of knowledge. Oxford:
Sean Kingston Publishing.
—. 2006. A community of critics: Thoughts on new knowledge. Journal
of the Royal Anthropological Institute, n.s., 12:191-209.
Viveiros de Castro, Eduardo. 1996. Images of nature and society in
Amazonian ethnology. Annual Review of Anthropology. 25:179-200.
—. 1998. Cosmological deixis and Amerindian perspectivism. Journal of
the Royal Anthropological Institute, n.s., 4: 469-488.


Let me begin with an anthropologically obscure corner of the world,

and with one of its practices. It is a practice that summons a model of
encompassment, for it purports to give a measure of a person by adding up
the elements of the person's accomplishment (cf. Hoskin 1996); indeed,
literally added, since the result is a score at the end of the day, I want to
suggest that ways of measuring persons yield insights into how
anthropologists might take the measure of the societies and cultures they
study, that is, arrive at some kind of summative or summary typification of
the nature of social life. Summaries may be global, totalising, definitive,
while variably encompassing or not encompassing in their effect.
My example comes from the University of Cambridge and its
procedures for academic promotion. What is interesting from our point of
view is the role given to evidence. The adjudicators have to score the
academic’s research performance—originality, reputation, and so on.
However, what they score is not what they think of the candidate’s
originality, reputation, and such, but whether or not there is evidence in the
paperwork. That is, there needs to be mention of these qualities in the
references and other supporting documents. It makes sense—the
adjudicators may not be specialists, and attention on the paperwork
eliminates anecdote and makes other kinds of knowledge extraneous.
An observer—an ethnographer in the anthropological sense perhaps—
might remark that this notion of evidence embodies procedure in a
documentary form (Riles 2006) with its own codes of presentation. Like
an oracle, it gives the procedures an authority—you might not know if the
scholar has done original work, but should one of the referees mention the
term, there is the evidence, strong or weak as it may be. Moreover, its own
scale—C, S, D: “Clear evidence”, “Satisfactory evidence” and “Doubt”
[about whether criteria are met]—offers the adjudicators a formula
through which to express their opinions without having to elaborate them.
In effect their opinions take statements in the paperwork as the evidence.
“Evidence” as an explicit focus of attention substitutes for other kinds of
factors that might contribute to a decision.
Old and New Reflections 21

Is this not a rather odd use of the term? We live at a time when
“evidence” carries huge politico-cultural freight—“evidence-based”
medicine is not the only mantra that rings in our ears. Much of it is to be
welcomed; that is not the point. The point is that however anthropologists
might wish to deploy the term in research settings, it has other lives
elsewhere, including no doubt for the university academic elsewhere in the
university, and takes diverse forms. On my desk, weighing far too much
from its thick paper, is a mammoth volume of glossy coloured
photographs of the natural world, presenting the arguments for
Creationism in a form heightened into medieval splendour with all the
lasciviousness of modern imaging technologies. The form is that of
scientific evidence.

Dividing evidence from evidence

Our observer has reminded us of the world in which anthropologists
operate. Relying on his or her language of exposition as a source of
analytical terms, the anthropologist’s refinement is to draw a guiding
distinction between different contexts. On the one hand is a specific
cultural form: in English “evidence” as a special mode of presentation is
also a technique for sifting data from data, the contours of a file into which
some but not all information will fit. There is an old insight here, namely
that only certain presentations of information will count as evidence. For
English speakers this is familiar from the law and is encountered wherever
you hear the term “evidence-based”. On the other hand, evidence can be
taken by observers for their own purposes as the grounds for knowledge,
the unfabricated, pre-constructed facts of a case always available for
further scrutiny. Here the relative rawness of what is a greater or lesser
fact is of course open to infinite regress. In either case, to talk about
evidence will be to establish a line, create a division, between what is/is
not recognisable or admissible as such.
So what is going to qualify for analytical usage? The English speaker
has no problem in managing multiple connotations in his or her own
language but to make the term do analytical work is another matter. How
explicitly, for example, does there have to be a corresponding concept in a
local repertoire? In his classic comparison of Merina and Zafimaniry in
Madagascar, Bloch (1975) found that he could analyse the economics of
both in terms of “labour”, but for only one of them could he say that they
made labour an explicit grounding principle for themselves. What, by
analogy, might we wish to make of the concept of evidence?
22 Marilyn Strathern

One of anthropology’s preoccupations is about just where the

ethnographer is going to make divisions between information presented as
having undergone minimal processing (“data”) and ethnographic
knowledge pressed into modes amenable to exegesis in order to specify
the context of explanation (“analysis”). This is a working problematic
rather than a problem to be solved, for the Euro-American scholar shifting
between contexts is ubiquitous to knowledge-making. The construction of
an explanation involves distinguishing contexts by typifying or
summarising sets of information. In short, “evidence” requires separating
kinds of information from one another. Leaving aside the notion of
“context”, let me adapt this formulation to at least make it possible to see
corresponding forms elsewhere.
I propose taking “evidence” as a construct pointing to practices—
whether undertaken by the anthropologist or his or her subjects—that
imply the ability to reduce, digest and otherwise summarise information in
such a way as to yield a yardstick or a measure by which other information
can be judged, proved or verified. It thus merges the two senses sketched
above in the general process by which matching pieces of information to
one another selects or privileges such information as is amenable to being
summarised. This draws attention to the mathematics entailed in the
Those promotions protocols deploy evidence in a specific way through
a mathematics of encompassment—the notion that a person’s worth [for
these purposes] can be added up as the sum of his or her parts, that lots of
little scores taken together will yield a bigger score, as one might imagine
adding together 3, 4 and 5 making a sum of 12, or 3*, 4* and 5*, coming
out with a summative 4*. I use “encompassment” for its connotations of
an embrace, of everything (of importance) gathered together.1 It used to be
thought that societies and cultures were amenable to such gathering, that
the more bits one described the more one knew, in the sense of having
enough material to arrive at some synthetic typification of their
fundamental characteristics. Indeed this was one way in which the very
idea of “societies” or “cultures” as singular, definable entities was created.
With this in mind I want to see how far our anthropologist would get in
situations where people do not add up their accomplishments in quite the
same way, where there is no sum of parts. Local mathematics is relevant.
For example, if “one” is the largest number there is, all multiples being
fractions of one, then it is division that creates magnitude, while each
specification of magnitude will always fall short of the totality (one).2 In
that kind of situation one has to ask what it means to typify or summarise,
and whether people look for anything like “evidence” in one another’s
Old and New Reflections 23

dealings. How does one match information with information? Here a

method already familiar to social anthropology’s comparative repertoire is
worth rediscovering. I refer to analogic reasoning. If indeed ways of
measuring persons yield insights into how we might take the measure of
social life, arrive at some kind of summary typification of what is going
on, then this non-encompassing measurement could hold some interest.

Admissibility: evidence of one thing

as evidence for another
Here is an instance of people giving information that they seemingly
want to be taken for evidence, in this case grounds for their grievances.
Their interlocutors make it clear that there is no way of adding up all the
different grounds or items presented to them. It is not that the information
does not count as evidence but that they take it as evidence of something
other than the grievance put in front of them, and they come up with their
own measurement or summary identification of the issue. The summary is
global in its effect.
I go back to Mt Hagen, Papua New Guinea, as it was in the 1960s.
People had somewhat eagerly embraced the then Administration’s
injunction that they should “sit down” under the law, and they regularly
held unofficial courts to deal with local matters whose novelty (“Now we
live in a new time”) was part of their popularity. These offered an arena to
which women could bring complaints, and to which men could bring
women. A surprising number of hearings concerned marital relations, not
so surprising perhaps in a situation where marriages sustained alliances
between men. Yet, from the observer’s point of view, women seemingly
found it hard to get their grievances taken as grievances. Indeed, the
(unofficial) courts appeared to be exacerbating rather than remedying the
problem. That was because of how they dealt with women as “judicial”
subjects: it was their feelings that commanded attention. What the women
said was taken as evidence not of their grievances but of their state of
mind or disposition, and that could often be summarised very simply. This
reaction had a globalising effect insofar as it displaced anything else the
women wanted to present. Evidence for one thing [grievance] was taken as
evidence for another [state of mind]. The airing of the grievances, and thus
the motivation to do so, rather than their substance, became the focus.3
And however many little complaints there were, they did not add up to
anything bigger: each pointed as much to evidence of the woman’s state of
mind as the next.
24 Marilyn Strathern

Men’s and women’s expressions of emotion had, and still have, public
effect. Feelings offered a currency to male politics, massive warfare
compensations being made on the grounds of assuaging (men’s) anger or
grief. A complainant’s interlocutors—those hearing a case—would
construe their summary explanation on the further conviction that
emotional state was in turn evidence of disposition or intention. Indeed
venting anger (popokl) could lead a person being construed, by him/herself
or by others, as a suffering subject. The state of being popokl (Strathern
1968), that is, being angry or upset or frustrated because of a grievance
against another, evoked the sympathy (kaemb) of the ancestors, who put a
person’s body into a suffering condition.4 In the 1960s example, how a
woman felt, in the double sense of the state of her emotions and her state
of mind, was what the court tried to find out. If court adjudicators knew
what her feelings were, they did not need to dwell on the reasons for the
way she had acted or the grievances she aired. So, when complaints were
against the husband, a frequent question was whether she was intent on
divorce anyway. This summed up the point at issue.
Now at the time I attempted to analyse these court cases in a quasi-
judicial frame. This set the Hagen observations into an arena of dispute
settlement procedures, informal courts and judicial processes.5 The
anthropologist’s modelling here supposed that the particular local case (the
example from Hagen) is a “small” instance of something that exists more
broadly (across an ethnographic range). And, in this model, the more cases
come into view, the more adequate the characterisation or summary of the
broader phenomenon becomes, each fresh example adding information to
the previous ones. We might read the same scaling effect into Hagen
men’s treatment of women’s complaints as a matter of disposition. That is,
one woman’s attitude might be taken as a single example of what you will
find in all women, and more is learnt about that the more cases there are.
However, I do not think this is quite how Hagen people would reckon it.
Insofar as one woman—and there are circumstances where the same
would apply to men too—embodies the characteristics of women in
general, she indexes or contains “all women” within in her. “One” is as big
so to speak as “all” or “many”. In this register it is meaningless to add
cases to one another—rather, each woman’s behaviour is like the next’s.
One woman is both the same and not the same as another. In terms of clan
and other affiliations, they are socially more radically dissimilar than the
English language generic (“woman”) conveys; before the court, however,
they are in analogous positions. I take the summarising conclusion about
states of mind as an example of analogic reasoning.
Old and New Reflections 25

Analogy can take the place of (other kinds of) evidence-gathering. On

the one hand, when a woman brought her grievances to court, by analogy
with other instances the court knew what kind of (other) evidence it
needed; on the other hand, in measuring one woman by analogy with
others certain outcomes were also already assumed: what did the woman
intend? how did she feel? was what the court knew it needed to find out in
this case too.
Now if there can be good or bad evidence, there can be good or bad
analogies. My turn to the concept of analogy is made interesting by the
fact that there are also occasions where analogies are refused or not
recognised or do not apply. If evidence may or may not be admissible,
analogies may or may not be appropriate. To prepare the ground for this
further point, I set the scene with an analogy of my own, a situation that
has nothing to do with Hagen but is illuminating for it.

It is taken from events that unfolded in the UK in 2005-6. In
conserving anonymity, I give unattributed quotations from documents and
disguise some of the facts. The reader who might find it difficult to take
the ensuing results as an adequate or satisfactory description, might or
might not accept its value as an analogy itself.
An Institution, and it could be a medical establishment or a university
or a museum, announced its intention to return the remains of some half
dozen individuals that had been in its collections of human material for
many years. This was in response to Activists from their country of origin,
for whom the significance of these relics could not be reduced to a bare
description of bits of bone. The Institution’s authorities were sensitive to
this. In fact, given the need to respond to the Activists they turned
themselves into adjudicators in trying to weigh up or measure the
respective “significance” of the remains to different parties. In line with
best practice, and to ensure that their decision would be evidence-based,
they called for dispositions from both the Activist claimants and the
Institution’s Department that had been keeping the remains for scientific
investigation. It is that weighing up that is of interest here.
The relics had been acquisitioned on the Institution’s side legally.
However, everyone knew that they had been extracted under horrific
circumstances. The Institution recognised the distress that its continuing
possession posed: the manner in which they had been torn from their
original location could only elicit sympathy for those claiming to be
descendants, as the Activists did. The Institution avoided questioning the
26 Marilyn Strathern

status of the Activists’ claims; it was moved by the feelings that they
expressed, by their obvious pain as suffering subjects.6
The issue became how to balance such feelings with the demands of
the international scientific community. The authorities in the Institution
took a lead in speaking of the need to weigh up the requirements of all
those with an interest in the remains, and spoke of the importance “of
respect for the different views and values of the various parties involved”.
The aim of dialogue between what was construed principally as two
parties (claimants and scientists) was first of all to identify the values they
shared. Their (the authorities’) own sincerity was expressed in the
acknowledgement that securing common ground would not be easy. Each
side was thus enjoined to extend “respect” for the values of the other side;7
both had claims on the Institution’s sympathy. The two parties were thus
put in the same scale: different weights would tip the balance in favour of
one or the other.8 So procedure entailed first a summation of the number
and gravity of arguments on each side and then matching each onto what
the Institution saw as its responsibilities.
However, those speaking on behalf of the Activist community made it
very clear that the relationship between themselves and other so-called
claimants could never be co-eval. The notion of a balance between
“parties” was rejected outright. Their claims simply could not be measured
on the same scale as those of the scientists—they were of a different order
altogether. If the Institution was trying to extend its field of ethical
judgement to encompass everyone, the Activists would have none of it.
One justification for retention on the scientists’ part was that these
remains were potentially “evidence” that could contribute to all kinds of
research. Scientific interest was legitimated by the medical information the
remains might yield as well as by promise of information about human
evolution, a sense much enlarged by new interest in the materials afforded
by molecular biology. It was feared that knowledge would be lost with
repatriation, for example, through subsequent interment or cremation. In
addition, it was argued, preserving the remains would allow their
descendants to know about their ancestors and thus more about
In the Activist camp, the appeal to enhanced knowledge seems to have
fallen flat. The difference between the two parties could not be reduced to
the idea that with discussion and information each would appreciate the
context from which the other was operating, and shift their own
viewpoint.9 The arguments could not be weighed together. The Activists
were related to their ancestors, the scientists were not. To borrow a phrase
from Astuti (1995), claimants and scientists were different “kinds” of
Old and New Reflections 27

people,10 and as different kinds of people their relationships to the human

remains were of different orders. Indeed, there was nothing more the
Activists needed to know about themselves in order to press their case,
only a matter of proving how they were related, through dance, song and
other evidence of entitlement. Knowing the conduct and meaning of such
performances would, in turn, only be effective when deployed by those
with the right to use them. There was nothing outsiders could offer by way
of evidence.
Where the Institution, then, strove for procedural parity, for an arena in
which claims could be balanced against one another, the Activist case
refused the invitation. The Institution wanted to weigh up or measure the
more or less / greater or fewer elements that would contribute to a final
(and encompassing11) summation, in the same way as each specimen itself
was enhanced by being part of a collection to which it added a quantum of
value. By contrast I infer that the significance that the Activists accorded
the remains created a field in which each element instantiates the whole. A
bit of bone is no more nor less than the whole, not the whole skeleton
because that is not the referent, but the ancestor in its integrity—and an
ancestor analogous to all ancestors. Whether “one” or “many”, each relic
is like all relics in coming from a specific place that ties it to its
descendants, while significance is constant whatever the size or quantity of
the elements. Another contrast: where the Institution drew an analogy
between claimants and scientists in supposing that each side had claims on
its sympathy, the Activists appeared not to acknowledge any grounds on
which such an analogy might appear.

Drawing and refusing analogies

These arguments came to my aid in thinking through some field
materials, and in thinking about someone who both drew analogies
between persons and refused or ignored or did not recognise them. Again I
disguise some of the details.
At roughly the same time as the Institution was debating these issues,
the BBC reported on child psychologists being sent to several regions in
the UK to help develop people’s parenting skills. British parents
apparently need help in learning how to control their children and to feel
comfortable dealing with them. Ideally, parents guide a child’s personal
development. Perhaps it was a pale reflection of this notion of parental
responsibility that led me on a visit to Mt Hagen in 2006 to complain to
the parents of two young Hagen men who—through letter and email—had
successfully conned me out of large sums of money. The response of the
28 Marilyn Strathern

parents took me some time to work through. In both cases the reaction was
eager agreement with me that their sons were cheats—and had cheated
them too. We were meant to expostulate together on their nefarious
characters, from which we had all suffered, rather than rehearse in the
parental generation (as parents and child psychologists would rehearse)
issues that in the Hagen view lay with the children themselves.
I am afraid that for quite some time I could only think that the Hagen
parents were avoiding facing the facts, the evidence of their sons’
behaviour staring them in the face. Moreover, rather like the Institution’s
desire to create an open dialogue, I hankered after some “talking out” of
what had happened. The young men came from two families I knew well,
and I couldn’t just push my grievance away. I would have settled for a
private admission that my relationship with the parents was in some way
affected by what their sons did. But that admission never came; vaguely
disappointed, I instead embarrassed myself talking about it all too much.
And of course the more I talked about it the more the seniors showed me
sympathy by saying they had been treated like that too! I think the seniors
axiomatically assumed I would have the same feelings on the matter as
them, and that if I was angry it was because of the loss of money, not
(which is what I thought was my grievance) through being hurt that I had
been duped by close acquaintances. In any event, I hadn’t learnt from my
experience forty years earlier that the more people gave public vent to
their grievances, the more they were examined on their feelings and
In many other ways, the visit reminded me that I already knew that
parents do not construe children as needing to be controlled, however
irritating or worrying their refusing to help or not coming home at night or,
indeed, being caught in petty thieving is. And I already knew that it would
be absurd to say the seniors were avoiding responsibility. For that is
conventionally defined in other terms: it is the parents’ responsibility to
keep the child in good health, and not themselves do stupid things that
might endanger it, including incurring ancestral wrath that can result in a
child falling sick.12 So a mother can trick a baby into yielding up the
kitchen knife; she is less likely to risk its anger by taking the implement
It is generally very important to Hagen people to assess how others are
feeling. And this is partly because of the effect those feelings will have,
both on the feeling person and on those with (as in sympathy) or against
(as in anger) whom those feelings are expressed.13 People in Hagen do not
invite themselves to extend “respect” for the emotions and dispositions of
others, gathering a range of attitudes together as though virtue lay in
Old and New Reflections 29

encompassing diverse viewpoints. Rather, they create fields in which

persons appear as analogues of one another. My friends rushed in to tell
me they had had just the same experience with their children.
Now one of the parents, so anxious to empathise with me on this score,
on a quite different occasion reacted not at all to a proffered analogy. That
was when he had visited me in the UK and hoped that he would finally
meet the important person behind my endeavours, who would really know
what reward would be commensurate to all the work he had contributed to
mine. I disappointed him I think, since I could not produce such a
character, although I could produce a senior and much respected fellow
anthropologist. Of course I nodded in approval when the anthropologist
explained to both of us that he had great sympathy with the Hagen man’s
quest, since he had encountered exactly such expectations in his own field
area, and knew how much he owed his own field friends, just as I owed so
much to mine. This was offered as a sign of respect. But the attempt at
analogy or symmetry (“I understand how you [the Hagen visitor] feel
because I have had just this experience myself with my own field
friends”)14 fell flat on its face. The Hagener professed to be completely
puzzled, making neither head nor tail of the allusion. What was this
personage doing talking about these other people? That wasn’t the subject
in hand; the subject in hand was he himself, an important man from Mt
Hagen at Marilyn’s University. “I am here [not all these others he is
talking about]!”15
I have introduced analogic reasoning as a practice of summation or
typification—this bone is [equivalent to] everything I know about my
ancestor; this woman's behaviour is [equivalent to] the behaviour expected
of all women; we can sympathise with one another because we have [in an
equivalent manner] both been cheated—and it is a practice of a non-
encompassing kind. It is without merging their identities that analogy
elucidates one thing by reference to another. It offers a yardstick or
measure that allows one to match (in these examples) person to person in
summary fashion.
However, referring to one thing by reference to another can also work
to displacing effect. One explanation or artefact or person is foregrounded,
filling space in a global sense. It is seen where evidence of one thing is
taken as evidence for another, as when the specifics of a grievance are
taken as evidence for a state of mind, or being related to an ancestor
obliterates any other consideration. One probably would not want to call
these analogies. Or perhaps one could recognise a collapsed analogy,
where the grounds for considering equivalence are merged, and one entity
substitutes for another.16 This is not encompassment in the sense I am
30 Marilyn Strathern

using the term here: the effect is rather that of eclipse. Importantly, the
possibility of refusing the analogy—claiming non-equivalence—is pre-
This brings me to a reflection. For elsewhere (see note xvii) I have
talked of analogic reasoning as one of social anthropology’s longstanding
comparative modes. Yet when anthropologists draw analogies between
social forms, they are most likely dealing neither with phenomena known
in advance to be similar to one another (as in the first instance above) nor
(as in the case of the collapsed analogy) with just eclipsing one kind of
knowledge by another. So what is being put forward under this rubric?

Analogies in social anthropology

At the outset I referred to taking measures of persons as one takes
measures of societies and cultures, or however one describes the segments
of social life that become subjects of study. There is no need to make too
much of the concept of measurement. The point was to draw attention to
ways in which people assess one another, and I took my examples from Mt
Hagen, that might be illuminating for thinking about anthropological
The discipline’s long-standing investment in the comparative method
has now largely turned away from accounting for differentiation and
similarity in principles of social organisation.17 Thus Gingrich and Fox
attack the idea of comparison “as a ‘hard science’ methodology employed
to support some universal theory or meta-narrative” (2002, 1).18 The
deployment of analogy, by contrast, is still very much part of the repertoire
of comparative methods; it allows comparison through a different
mathematic. Analogic reasoning is non-encompassing in that it
presupposes neither a common field against which comparisons are made
nor an effort to bring disparate “voices” into some kind of alignment.
Indeed when it comes to putting accounts of social life alongside one
another, it is possible to be agnostic as to whether any of the elements so
described have an interest in any other simply because they are together in
the narrative.
Social anthropologists constantly juxtapose materials from domains
they construct as distinct. They carry around in their heads examples from
“elsewhere”, using the images and languages of one thing to talk about
another. We are familiar with the way models may be drawn from Euro-
American institutions such as economy or law in order to create a
language through which to describe economic or legal practices in non-
Euro-American situations. Sometimes the parallels between distinct
Old and New Reflections 31

systems remain implicit. Where they are explicit, the non-reductive nature
of this kind of comparison becomes most apparent. For the genre, analogic
reasoning, offers insights into both the terms being brought into relation
with each other—illumination goes both ways.
Analogies at once conserve and extend. What makes this type of
relation non-encompassing is thus the fact that the two elements to an
analogy are not merged through appeal to some higher order / fundamental
similarity. The analogies in this present account align elements that
according to Euro-American notions of identity have a prima facie
similarity (between women bringing grievances to court / between people's
ancestors); but they also align those that on the surface appear dissimilar
(university promotions procedures / unofficial courts in Hagen forty years
ago / an Institution trying to foster dialogue), and in the case of my field
friend (who both sympathised with me and saw no connection between
myself and a fellow anthropologist) an invitation to discriminate between
what is appropriate and what is inappropriate. Comparison is thus allowed
to emerge in specific contexts—there is no need for an encompassing
agreement of similarity before the examples are drawn together. In sum,
the power of thinking one thing through another lies in conserving the
mutual distinctiveness of each. At the same time, in accord with the
anthropological project of comprehension, the understanding of each is
extended by introducing the other into its description. We may note that
neither prior equivalence nor collapsing the one into the other is at issue.
What does seem at issue in the anthropologist’s method is the
juxtaposition of disparate forms. The old problem with that, a familiar
criticism of anthropologist’s techniques, was that anything could
seemingly be juxtaposed to anything else. The material presented here
offers a small comment. (Indeed if it is illuminating then it serves as
analogy for some of anthropology’s practices.) In showing specific
instances where analogies are judged apt and those where they are not, the
material leads us back to the question of evidence. For it reminds the
anthropological user of analogic reasoning that not all analogies may be
equally appropriate. The very concept of analogy forces one to specify the
conditions under which bringing disparate items into one frame might or
might not be illuminating. The process will then create “evidence” for the
analytical choice.

I have talked about diverse indigenous or vernacular practices that lead
to people making summary judgements, much as one might say that the
32 Marilyn Strathern

anthropologist derives summary observations that can be matched with

material from elsewhere. It creates a form in which material can be
presented as evidence. The account touched on both encompassing and
non-encompassing modes of doing so.
The examples of encompassment draw attention to a common
mathematic that elides many examples into one summary of them, as when
scores are added up, or authorities taking a decision ask for opinions from
all sides and then try to weigh them up. No single instance should take
precedence; each is a unit contributing to an overall schema. A
contributing element is recognised as such, however, because it is
amenable to, recognisably part of, the concluding summation. Thus a
submission to authorities “collecting evidence” from interest groups does
best when it is in a form that fits the documentary record of a fair
adjudication. The final decision is also a final summary, encompassing
everything that went into it.
There are also processes of typification and summation I have called
non-encompassing. Here I have used the concept of analogy, and in three
different ways. One entity may summon a similar other, or may be
eclipsed by a different other, or else diverse entities may be juxtaposed to
one another. In all three cases the purpose is to illuminate an adjudication
or decision (whether in a broad sense political, or judicial or analytic). So I
have tried to understand my recent encounters in Hagen by referring to
earlier experience (court cases of forty years ago), what I had just heard on
the news (parental responsibility in the UK), and aborted attempts to
extend sympathy and even empathy (in the case of the Institution’s
authorities). What becomes crucial is the initial choice to use one set of
materials rather than another to illuminate the matter in hand. For the
anthropologist struggling with methods of comparison, there is no point in
saying the connections are spurious; there is every point in judging just
how illuminating or not they turn out to be. Of course such a judgement
can only be done after the fact. One does not know in advance what a bad
analogy will be; but bad analogies have to be made before they can be
thrown away.

I am very grateful to Karen Sykes for the invitation to the conference
“Living paradoxes: Moral reasoning and social change”, Manchester
University, December 2006, at which parts of the present paper were first
aired. Parts of that were in turn drawn from others given elsewhere.
However the principal inspiration remains a short period of fieldwork,
Old and New Reflections 33

undertaken in 2006 at two locations in the Western Highlands of Papua

New Guinea in collaboration with Almut Schneider; I owe much to her
several comments.
My warm thanks to the National Research Institute (PNG), the office
of the Administrator, Western Highlands Province, and the British
Academy for their support for the work in Papua New Guinea, and to Maja
Petrovic-Steger for conversations about the repatriation of human remains
and observations on this chapter.

Text of one of two keynote speeches at the “Questions of Evidence: Ethnography
and Anthropological Forms of Knowledge” conference, University of Cambridge,
27 January 2007.
Kindly note the particularity of this usage. I have tried to think of a substitute for
the term encompassment, since it has so many connotations in anthropology (and I
have elsewhere used it quite differently myself) but have so far failed.
The notorious example of Iqwaye (Mimica 1988).
For the sake of argument, I am summarising from numerous cases, and
innumerable possible positions that people could take. From the court’s point of
view, emotions were often visible and did not require questioning; however,
although everyone might jump to conclusions about disposition or intention,
extracting an “admission” to that effect might take a long time.
Their sympathy could turn sour if persons let their feelings get out of hand or
were angry for no good cause, with potentially lethal consequences.
This, along with a concern with male-female power relations in the idiom of the
time, no doubt contributed to my own stance, interpreting the men’s summaries as
dismissive, and assuming that women would want to have their grievances
considered one by one.
From a document submitted in evidence: “We are moved by the situation of the
[Activist] community in terms of its history and modern condition. The return of
the human remains can have an important role for the living community in moving
forward from a negative historical experience.”
Cf. the opening to Strathern 2006a.
The assumption that there were just two “sides” was also the guiding structure of
a report published by Human Remains Working Group (DCMS 2003), a group set
up by the UK Department of Culture, Media and Sport to reconsider the long
established responsibilities of, in this case, Museum Trustees for material entrusted
to them.
This is an inference, which draws both on the reported Activists' response to the
Institution's statements and on earlier representations made by other claimants to
the DCMS Working Group (see note viii). I use the term “context” advisedly here,
as a likely part of the authorities’ way of thinking.
34 Marilyn Strathern

The DCMS report gives several verbatim comments from claimants in similar
situations who take this viewpoint: e.g. “It is our direct ancestors that are being
experimented on [original emphasis]”; [in reference to an attempt to press for
repatriation] “We went ... to see our ancestors and we were told that we cannot see
them. For us it is like going to see somebody in hospital”; “How can research
possibly compare? We’re tired of other people interpreting us to ourselves”
(DCMS 2003, 55).
Minimally encompassing of all the facts that had been taken into consideration
(see note i).
A child can bring a similar fate on itself: very little children become popokl
when they are thwarted, and invariably against those parental figures for whom
they also feel “sympathy” and whose “sympathy” in turn they want back (cf.
Strathern 1968, 556-7).
Feelings invariably have other persons as their object, so we can also say that it
is (their) relationship that is the subject of attention. On situations where desire for
or greed over acquiring relationships can lead to people taking all kinds of auto-
destructive actions, see Strathern 1997.
A double analogy, between myself and the anthropologist, and between my
expressed sympathy for my Hagen friend and the anthropologist’s expressed
sympathy for his friends, so the “other” anthropologist could thus express common
feeling with “my” friend.
There were other issues here as well. For example, he was being asked to put
himself into the shoes of people he did not know; it may also be that he did not
recognise the sign of “respect” because, after finally coming to Cambridge, he was
put outside again in being compared to unknown people of non-Euro-American
origin (Almut Schneider, pers. comm.).
The reasoning here follows that of The Gender of the Gift where under one
rubric (“exchange”) I call both mediated and unmediated forms exchange, and
where I contrast two relational modes as replication and substitution (1988, 298).
The following is drawn from Strathern 2006b.
They put in its place three dimensions of comparison: the general point that
human cognitive development rests on the capacity to compare entities with one
another; the necessary translation across cultures involved in any anthropological
enquiry, simply because the anthropologist’s interests derive from another context;
the strong sense in which a research programme is driven by an interest in regional
or temporal variation, encompassing an emergent “broad new pluralism of
qualitative methods” (Gingrich and Fox 2002, 20).

Astuti, Rita. 1995. People of the sea: Identity and descent among the Vezo
of Madagascar. Cambridge: Cambridge University Press.
Old and New Reflections 35

Bloch, Maurice. 1975. Property and the end of affinity. In Marxist

analyses and social anthropology, ed. Maurice Bloch. London: Malaby
DCMS. 2003. The report of the working group on human remains.
London: Department of Culture, Media and Sport.
Gingrich, Andre, and Richard Fox, eds. 2002. Anthropology, by
comparison. London: Routledge.
Hoskin, Keith. 1996. The “awful idea of accountability”: inscribing people
into the measurement of objects. In Accountability: Power, ethos and
the technologies of managing, ed. Roland Munro and Jan Mouritsen.
London: International Thomson Business Press.
Mimica, Jadran. 1988. Intimations of infinity: The cultural meanings of the
Iqwaye counting and number systems. Oxford: Berg.
Riles, Annelise, ed. 2006. Documents: Artifacts of modern knowledge.
Michigan: University of Michigan Press.
Strathern, Marilyn. 1968. Popokl: The question of morality. Mankind,
Journal of the Anthropological Societies of Australia 6:553-562.
—. 1988. The gender of the gift: Problems with women and problems with
society in Melanesia. Berkeley and London: University of California
—. 1997. Double standards. In The ethnography of moralities, ed. Signe
Howell. London: Routledge.
—. 2006a. A community of critics: Thoughts on new knowledge. Journal
of the Royal Anthropological Institute, n.s., 12:191-209.
—. 2006b. Useful Knowledge. Proceedings of the British Academy


This chapter seeks to address a question that is central to the
anthropological analysis of HIV/AIDS in sub-Saharan Africa: what kinds
of ethnographic evidence are available to anthropologists who endeavour
to develop a nuanced understanding of highly stigmatised topics that
informants avoid consistently, and intentionally, in open conversation? I
suggest here that a potentially fruitful approach to such a problem may be
located in musical responses to HIV/AIDS. Ethnographic evidence in
support of this position is presented in the form of two performative genres
taken from the Venda region of South Africa.1 The music of female
HIV/AIDS peer group educators—who claim to “sing about what they
cannot talk about”—is discussed in relation to a recent song about AIDS
written and performed by Mr Solomon Mathase, a well-known Venda
Tshilombe (plural: Zwilombe) guitarist. The peer educators’ performances
are funded and designed by international donor agencies, and thus reflect a
biomedical understanding of the pandemic. Conversely, in his song,
Solomon Mathase reflects the “folk model” (Good 1994) of sexual illness
through which peer education, and biomedical discourse, have been re-
interpreted in line with conservative, patriarchal aetiology—headed
ultimately by networks of ancestral authority. As this example serves to
illustrate, biomedicine is often defined through the folk model as both a
proximate and ultimate source of physical illness at the heart of a
perceived crisis of social reproduction in the region (cf. Comaroff and
Comaroff 2004; Liddle et al. 2005).
Building on Fabian’s insight that ethnographic accounts of performance
may reveal forms of knowledge that are not otherwise expressed in
“We Sing About What We Cannot Talk About” 37

discursive statements (Fabian 1990), I argue in this chapter that the

ethnographic analysis of AIDS songs can reveal important, hidden
dimensions that remain elusive in public, and often private, arenas. Of
course, “performance” is a fluid and polyvalent category of analysis;
praxis-orientated linguistic anthropologists have argued, for example, that
the dynamics of conversation can be framed in performative terms. In the
discussion that follows, however, I take musical performance as quite
distinct from conversation—not only in aural aesthetics but, as I hope to
demonstrate, in the social and political implications they elicit when
applied to causes of death such as HIV/AIDS in post-Apartheid rural
South Africa.
The focus on musical interpretations here exposes the extent to which
AIDS educators are framed as vectors of the virus. In the Venda region—
as I argue elsewhere (McNeill 2007)—anyone who displays conspicuous,
public knowledge of a death may be equated with potential implication in
the fatality. Similar patterns of blame have been reported in southern
Africa more widely, and as such HIV/AIDS is bound to complex webs of
secrecy and suspicion that prohibit its inclusion in open, public
conversation. It would therefore seem apparent that the general reluctance
to discuss AIDS openly in this region is rooted not in a widespread
“collective denial” of its existence (see, for example, Campbell 2003;
Lwanda 2003; Reid and Walker 2003) but rather in the complex
relationship between AIDS and causes of death—enmeshed in networks of
witchcraft accusations—through which AIDS has been framed as a “sent
sickness” (Farmer 1990; cf. Ashforth 2004, 2005). The avoidance of open
conversation in this context is a safety precaution, collectively undertaken
by individuals to protect themselves from the constant threat of guilt by
association that connects knowledge of a death with experience of (read
implication in) the fatality. By maintaining a public silence, degrees of
separation are created between an individual and the cause of another’s
death, and thus the refusal to discuss death—or causes thereof—without at
least some form of obfuscation, is a direct protest of innocence.
As a highly stigmatised and potentially dangerous topic of conversation,
then, ethnographic investigation into AIDS in southern Africa presents the
researcher with a complex of difficulties. Of course, one ought not to rely
exclusively on methodological approaches that privilege the role of
conversation in the construction of social realities—and researchers have
made fruitful use of non-verbal forms of communication for many years
(see, for example, Gell 1999 or Blacking 1985 for a discussion of
symbolic codes in distinctive patterns of walking and dancing). Moreover,
what is not revealed in conversation should provoke the ethnographic
38 Fraser McNeill

imagination just as much as what people actually talk about. Both must be
considered as evidence from which ethnographers can attempt to mould an
interpretation, and tell their story. The task of gathering anthropological
evidence—in this context possibly more than others—becomes a case of
balancing divergent models of the world: interpreting what is said in
public or private conversation against what remains unspoken, and
reckoning this with what is sung in very public performances against
apparent realities that the ethnographer seeks to identify and piece together
“on the ground”. And yet every ethnographic account remains a partial
one: as much a representation of the ethnographer’s choices and
constraints as it is of those under scrutiny. Whilst this is important to
recognise, the case studies in this chapter point towards issues that demand
an elaboration upon conventional ethnographic methods by anthropologists
in search of evidence that may help us to understand localised responses to
HIV/AIDS in a region where one in five people is estimated to be HIV
positive (UNAIDS 2007).

“We sing about what we cannot talk about”

The constant claim that peer educators sing about what they cannot
talk about struck me as familiar. To be sure, it resonates with
anthropological approaches that suggest performance may allow people
to say things they would not otherwise articulate, and thus provide
access to what Vail and White have called ethnography from “within”
and history “from below” (1983, 887; Vail and White 1990). Following
this line of investigation, the analysis of musical performance can grant
the anthropologist “privileged access to layers of consciousness that are
normally not available to scholarly examination” (Erlmann 1996, 12). In
this vein, Fabian (1990, 12-13) has suggested that performance is “the
tip of the iceberg of culture floating in a sea of history and
consciousness”.2 Fabian’s assertions were inspired by Victor Turner’s
earlier theoretical concern with plays, ritual, and festivals, which he
argued must be read as more than mere enactment, but rather analysed
in terms of how they contribute to the (re)generation of new symbols
and meanings (Turner 1982). Fabian built on this to construct a notion
of “performative ethnography” through which anthropologists can
understand culture not only through the interpretation of a set of
symbols but, more essentially, through examining the enactment and
performance of plays, dramas, and songs (1990, 6). Such an approach is
intended to facilitate the reading of forms of knowledge that are not
“We Sing About What We Cannot Talk About” 39

expressed in discursive statements, but which find their way into the
public domain through alternative forms of expression.
What is it about music, then, that facilitates relatively extensive public
comment on AIDS? In a context in which most people choose to operate
in the registers of gossip, rumour, or what Stadler (2003) has termed
“public silence”, why is there an abundance of songs on the topic?
Fundamental to answering this question is the fact that some people in
some rural South Africa contexts can—and do—talk about death and its
causes in a variety of social situations.3 In Venda, it is those with
significant political power (traditional leaders or elected politicians) and
authority through spiritual sanction (such as healers, prophets, ritual
experts or Zwilombe musicians) who are more likely to speak openly
without fear of implication, although they may choose not to. Music is
often the medium through which these powerful actors frame such social
commentary, and this often constitutes part of a wider ritual process in
dialogue with ancestral spirits.
Conversely, as young, mostly unmarried women, HIV/AIDS peer
educators in Venda characteristically lack recourse to such tools for
negotiation, and by displaying their expert biomedical knowledge about
AIDS on a daily basis, they risk being caught up in webs of suspicion and
blame. I argue that, despite the inherent danger involved, when peer
educators in Venda sing AIDS songs, they are engaging in attempts to re-
appropriate symbolic relationships with political power and ancestral
authority through which they seek to legitimise their attempts to promote
safer sex. They engage in attempts to avoid implication in AIDS-related
deaths by transferring their bio-scientific knowledge through the medium
of well-known, often sacred, communal songs that are associated with
powerful, spiritually sanctioned institutions. It is through Solomon
Mathase’s musical contribution, in the second half of this chapter, that the
futility of this strategy is revealed.

HIV/AIDS peer group education

For the international donor agencies that fund HIV/AIDS peer
education projects, however, the incorporation of songs into public
performances takes on quite different significance. Connected to the
educators’ pervasive claim that they sing about what they cannot talk
about—and under the misguided assumption that “communal denial”
currently inhibits open discussion—policy makers have stuck to a rigid
programme of so-called “participatory interventions” in which the
composition and performance of pedagogic songs plays a central role. For
40 Fraser McNeill

the policy makers, this is intended not only to encourage safer sex in the
general public, but to “empower” educators by:

transferring health related knowledge from the hands of outside experts to

the hands of ordinary people, increasing their sense of perceived control
over their health … [and] … deriving respect and recognition for their role
in promoting health. (Campbell 2000, 492)

As young, unmarried women, peer educators lack socially recognised

access to status, and thus the “respect and recognition” to which Campbell
alludes is indeed an important motivational factor in their decision to
undertake voluntary work. In the Venda region, given the strong
connections between some NGOs and those influencing the distribution of
government contracts, there is also a good chance that becoming a peer
educator may be the first step on a ladder to public sector employment. In
relation to this, HIV/AIDS peer education is deemed successful by project
designers (or “experts” who conduct short evaluation visits) if it instigates
the collective renegotiation of social and sexual roles among target
groups—in this case young women. In its ideal context and under ideal
conditions, peer education is thus based on the premise that

people evaluate changes not on the basis of scientific evidence or

authoritative testimony, but by subjective judgements of close, trusted
peers who have adopted changes and provide persuasive role models for
change. (Wilson 2000, 45)

Although considerable resources have been channelled into peer

education, there remains a distinct absence of research on or knowledge of
“the means by which, and the effectiveness with which the programmes
change sexual behaviour” (James 2002, 180; cf. Campbell 2003; Turner
and Shepherd 1999). Our understanding of the social processes underlying
the alleged efficacy of such “participatory” approaches to health
promotion thus remains in its infancy. However, what is clear is that
singing AIDS songs at weekly public performances is intended to facilitate
participation, promote safer sexual behaviour, and advertise the
employability of the educators.
As part of this process, peer educators are trained in government
funded “workshops”, in collaboration with donor agencies, to refashion
pre-existing songs into what Alfred Gell may have termed an artistic
“trap” (Gell 1999). In doing so, they make simple lyrical and stylistic
adjustments to popular/relevant songs and dances, with the intention of
drawing in the audience who, if the trap is well-set, will have pre-empted
“We Sing About What We Cannot Talk About” 41

the well known lyrics. As the audience attempt to sing along with the
educators, they should notice the new lyrics which, in turn, should
theoretically encourage open dialogue as to the meaning of the new words.
The educators are trained to step in as biomedical experts and furnish the
audience with authoritative information about HIV/AIDS, before
distributing free condoms. As we see below, in combining “traditional”
melodies with those from church and the anti-Apartheid struggle, peer
educators select songs for symbolic value, lyrical compatibility and
general popularity.4 A crucial ethnographic component here is of course
the composition of the audience. The public performances take place once
a week, generally on a Friday afternoon, in beer halls, clinics, market
places, taxi ranks, or at any event that may have attracted groups of
people. In addition to the faithful core of friends and family that follow the
educators around, providing moral support, their performances are
watched by anyone who happens to be there. This can range from school
children, unemployed men in a beer hall, married wives at a market place,
to the elderly at a pension point. Sometimes the audience includes more
than fifty people, and at other times it can be fewer than five.

AIDS songs in peer education

In southern Africa, as in other regions of the world with a history of
missionisation, religious affiliation is an important indicator of historically
constituted social classes (De Gruchy and De Gruchy 2004). The church a
peer educator attends thus also determines the salient “peer group” from
which AIDS educators are constituted. Church songs are used throughout
weekly training sessions and public performances for opening and closing
prayers and for the blessing of food. Members of various denominations
bring songs and melodies from their Sunday worship into the peer
education arena, and they work in a group to find alternative lyrics that
encapsulate succinctly biomedical notions of HIV, AIDS, treatment and
prevention. The song below, for example, has been adapted from a very
popular Lutheran gospel song, performed during church services as a
short, upbeat “chorus”.

Condom is Number One!

Church Lyrics: (sung in English)

Leader: Jesus, Jesus, Jesus is number one!
Chorus: No matter what the people say, but Jesus is number one.
42 Fraser McNeill

Peer Education Lyrics:

Leader: Condom, Condom, Condom is number one!
Chorus: No matter how can (sic) people say, but condom is number one.

The peer educator version exemplifies the attempt to ensure efficacy

by keeping lyrical changes to a minimum over a well-known melody.
Audience reactions range from shock or amusement to discomfort and,
occasionally, anger. Many people associate church songs with sacred
spaces and their secularisation as blasphemous. For the peer educators—
who are acutely aware of this—this is exactly the point:

When we sing the songs from church, some people, especially the old men
[laughing] can get very cross with us. Even my auntie, she told me she was
not happy about this. It’s like we are stealing from them, or laughing at
Jesus, but we say No! They should be blessing us because the Bible says
“God is Love”, and how can you love if there is no life?! [laughing]
Really, if people are getting shocked then we think that is good! They must
just get tested [for HIV] and use condoms.

Thus, although church music seems an obvious choice for them, by

replacing “Jesus” with “condom”, peer educators make a flagrant attempt
at shock tactics in a generally conservative religious environment.
Condoms become sacred, worthy of worship and sacrifice with the
eventual gift of life over death. As I indicated above, however, peer
educators’ musical repertoire is not limited to or defined by church groups.
Peer educators also modify songs that are associated with political activity
against Apartheid.5 Struggle or “freedom songs” have recently been
described as

short slogans … set to simple melodic phrases, sung a cappella

(unaccompanied) and repeated over and over in a call and response style.
They are created and sung collectively, and frequently modified as politics,
attitudes and circumstances change, and are almost exclusively non-
commercial. (Gilbert 2007, 426)

Moreover, they are “ubiquitous but largely informal and un-

professionalised” (ibid., 423) and currently constitute a dominant medium
of popular political expression throughout South Africa. Although
reference to these songs is scarce in the existing literature on South
African music, they are clearly related to the originally Zulu Isicathamiya
genre (cf. Erlmann 1996) and, like it, are descended from pre-struggle
church music. They are also strongly derivative of, and arguably an
example of, the makwaya (choir) genre (James 1999, 2006, 155) that has
“We Sing About What We Cannot Talk About” 43

stylistic origins in the combination of “southern African singing traditions

with Christian hymnody” (Gilbert 2007, 426). As Gilbert has
demonstrated, a considerable number of these songs made reference to
African National Congress (ANC) leaders like Albert Luthuli, Oliver
Thambo, Walter Sisulu, and Nelson Mandela, “honouring their leadership
and calling for political guidance” (ibid., 435).
The sole purpose of this musical genre was to send out a political
message, and encourage collective political action (Nolwin 1996).
Although the purveyors of many genres such as jazz, reggae, gospel,
choral, and so-called traditional music were actively engaged in
challenging the policies of segregation, the struggle songs in question were
distinct in several ways. Chief among these was the accompanying “toyi
toyi” dance that grew to symbolise the preparation for warfare (McGregor
2005). The lyrical content of struggle songs was highly politicised and
often aggressive. The example below has a threatening message for the
political leaders of the Apartheid era homelands. The chanting comrades
exchange Mangosuthu (“Gatsha”) Buthelezi, previous head of the
KwaZulu homeland, for other leaders of the puppet regimes such as
(Patrick) “Mphephu” in Venda and (Lennox) “Sebe” in Ciskei, and
prophesise their rendezvous thus:

Hey wena Gatsha Hey you, Gatsha

Amasi abekwe elangeni It’s only a matter of time
Sodibanda we shall meet
Ngebazuka with a bazooka.6

Songs of the struggle were illegal under strict state regulations over
what could, and could not, become popular music, and they accompanied
illegal political rallies and demonstrations. Although they were illegal,
they were—and are—hugely popular. However as a result of the lack of
recordings, they circulated largely in formats that were not standardised,
and the lyrics changed depending on the leader. The versions aired on
Radio Freedom were re-invented in different languages throughout the
country. After 1990, with the un-banning of the ANC, this began to
change. Record companies such as BMG started making commercial
recordings, and artists such as Amaqabane, Blondie Makhene, and various
choral ensembles began to produce standardised versions with synthesised
backing instrumentation.
Given that contemporary Venda remains an ANC stronghold, such
songs often become part of the entertainment at public gatherings during
and after local ANC ward meetings, when they are sung in praise of old
and new comrades. They regularly form part of beer drinking or meat
44 Fraser McNeill

eating repertoires through which men remember the past and comment on
the present. Nonetheless, they are also widely associated today with
general protests against the ANC government, and the considerable
influence that struggle songs had in the past has to some extent been
brought into the post-Apartheid political arena. It is in this context that
HIV/AIDS peer educators attempt to harness the power associated with
them by utilising the genre in public performances. As one said:

[T]he struggle is not yet over. We are still struggling, can’t you see?! It is
not enough that we can vote, but voting is important. My house remains
without water and I still cook on a fire at night. Anyway, now AIDS is the
enemy, it kills us secretly just like the Boer used to, and we should fight it
[AIDS] as we did before with the Boer.

The song below is an example of the way in which peer educators have
adapted a very famous struggle song in praise of ANC comrade and
leader, Joe “Ntate” (father) Modise. A standardised recording was released
in 1990 by the band Amaqabane. The four beats of the opening line
“Nta/te/ Mod/ise/” (father Modise) are replaced by the peer educators, in
exactly the same space of rhythmic punctuation with “/condomu/ ndi
/bosso/” (condom is the boss). In doing so, the educators seek—very
consciously—to invest their own enterprise with the power and authority
of ANC leadership, and lobby for the support of their audiences, who are
familiar with the melody but not necessarily with the new lyrical content.
It is performed with variations on the themes presented below, in verses,
with the call and response differing depending on the peer educator chosen
to be the leader. Such variations, however, are always rooted in the
conviction that condoms are invincible barriers against HIV infection.

“Condomu Ndi Bosso”: Condom is the Boss!

Leader: Condomu ndi bosso! Condom is the boss!

Chorus: Ndi bo… It is…
Ndi bosso It is the Boss
Ndi bosso, condom ndi bosso. The boss, condom is the boss.(x2)
Leader: i thivhela malwadze. It shields us from sickness.
Chorus: Malwa… sick…
Malwadze. sickness.
Malwadze a vhudzekani. sexually transmitted sickness.
Leader: Condomu nga shume. Condoms are here to be used.
Chorus: Nga shu… to be…
Nga shume, to be used
“We Sing About What We Cannot Talk About” 45

Nga shume, Condom nga to be used, condoms to be used.

Leader: Condomu ndi ngwena. Condoms are crocodiles.7
Chorus: Ndi ngwe… it’s a croc…
Ndi ngwena it’s a crocodile
Ndi ngwena, condom ndi A crocodile, condoms are
ngwena. crocodiles.

In a manner akin to the attempted incorporation of political elements

through struggle songs, the peer educators also invoke “tradition” as a
means to further legitimate the promotion of safer sex. Most strikingly,
they open every public performance with an imitation of the Domba
“python dance”—the most emblematic symbol of female initiation—in
which they shuffle towards the chosen arena imitating a group of initiates
(vhatei) undergoing the rites of adulthood. In addition to this, when
performing “traditional” songs, peer educators use the murumba drum that
is central to the processes of Vhusha and Domba girls’ initiation (cf.
Blacking 1969).
But why should a ritual context be perceived as secure for the
transmission of knowledge pertaining to sexual health? As I discuss at
length elsewhere (McNeill 2007), the intricate processes of female
initiation in Venda involve the incorporation of milayo (laws) through
which young girls learn about the dangers of menstrual pollution and the
necessity of monthly ablutions (with the strict observance of blood-related
taboos) in preventing the onset of sexual illnesses that can threaten fertility
and lead, in some cases, to death. In this controlled ritual context, stratified
by hierarchies of ritual knowledge, the ancestors (through the authority of
ritual elders) actively facilitate healthy reproduction; mostly through the
performance of initiation songs. Music, in this context, acts as a medium
for the complete, safe transference of ritual knowledge whilst the songs
and dances of initiation represent the desire for continuity in healthy social
and sexual reproduction (McNeill 2007; cf. Blacking 1969, 1973; La
Fontaine 1985).
This highly formulaic pedagogic strategy can be seen in stark contrast
to everyday public conversation, which would not normally include sexual
illness (as a threat to fertility and potential cause of death). As I have
already suggested, the explanation for this avoidance is to be found in the
close relationship that exists between knowledge of death and potential
implication in the fatality. For this reason, the power associated with such
ritual knowledge is intrinsically ambivalent; it is dangerous and can
threaten life precisely because of its life-giving qualities, and those who
acquire or teach it are both “respected and suspected” (Lambek 1993, 7).
46 Fraser McNeill

Indeed, as Michael Lambek has acknowledged of ritual knowledge in

Madagascar (1993), and as we see below in Solomon Mathase’s lyrics, it
is often bad to know too much.
Despite this flirtation with blame (and also because of it), peer
educators continue to use “traditional” songs in the promotion of safer sex.
The potential repertoire available to them is connected to and restricted by
the classification of communal Venda music into distinct genres
performed by men, women, adults, and children (see Blacking 1965; Stayt
1931). As mostly young women, peer educators use a combination of
Malende (sung by men and women at beer drinks, especially during
harvest time and at the homecoming parades of initiates), Vhusha and
Domba (both from the cycle of female initiation schools), Tshifase (a
courting dance for teenagers closely related to initiation school songs), and
Tshigombela (young girls’/women’s entertainment). The fact that
Malende, Tshigombela, and Tshifase have historically been employed
extensively to construct and articulate social commentary on contemporary
affairs (Kruger 1999) does not exclude them from being associated with an
idealised, conservative, long-lost past. I suggest that it is the power
associated with this historical connotation and the connection with various
ritual contexts that peer educators try to capture when using traditional
songs, and it is precisely because of their historical role in channelling
social critique behind ancestral protection that they choose to transform
them into part of the AIDS education repertoire.

Jojo Tshilangano (Tshifase)

Tshifase is a dance in which the sexual and social tensions between
young men and young women nearing the end of their initiation cycles are
brought into the open and, at least theoretically, resolved. For Van
Warmelo (1989, 397) Tshifase is:

[S]ong and dance of young folk … a diversion for moonlit evenings. Boys
and girls stand in two groups apart, singing. A girl comes across to the
boys, hooks her choice by the arm and leads him to her own group. A lad
now does likewise, and so on until the groups are sorted out again.

Although the only lyric in Jojo Tshilangano is a repetition of the song

title, the young performers are made aware of a story behind it. Jojo is said
to be a migrant labourer (mugaraba) who has returned from Johannesburg
in characteristic style with money to spend, and is in demand with the
local girls. As performed outside of the peer education context, the girls
stage a competition to impress Jojo and he subsequently chooses the best
“We Sing About What We Cannot Talk About” 47

dancer. Jojo and his new girlfriend then meet in the middle of the two
groups, until another young man is chosen to play Jojo, and so it
The dramatic presentation of Jojo Tshilangano, and other dances like
it, places the performance stylistically between song and drama in the peer
education repertoire. In the version they stage, an educator dresses in drag
as Jojo, donning dungarees, a hat, boots, and whatever else she can find. In
an elaborate and exaggerated mime played out against rhythmic
accompaniment from the murumba drum, (s)he approaches a girl in the
other group and tempts her to dance in a sexually explicit, but comical,
routine. With the rest of the peer educators now forming a circle around
the pair, she accepts his proposition and they frolic in the middle, until
eventually he offers her money for sex. She is tempted, but refuses.
Dramatically, Jojo then produces a condom, thrusts it high into the air as
the peer educators ululate and the girl falls into his arms—generally to the
great amusement of the audience.
I suggest here that the manner in which peer educators use music to
deliver biomedical discourse can be seen as an attempt to transform AIDS
education into a medley of messages and, by extension, themselves into a
motley crew of warriors against social injustice with sacrosanct
associations, legitimated by a play on “tradition” and spiritual authority. In
doing so, they have cobbled together a bricolage of historically constituted
meanings and seek to present them in one coherent performance in which
each musical tradition takes on, and generates, new significance.
Taken as examples of ethnographic evidence in a socio-cultural
environment where people claim to sing about what they cannot talk
about, the songs above demonstrate the processes through which peer
educators promote safer sexual behaviour whilst engaging in performative
contexts intended to evade implication in unnatural death. They also point
towards the ways in which peer education provides spaces for disadvantaged
women to construct and project a distinctly positive identity through which
they strive to negotiate positions of power and influence in the adverse
circumstances that frame the current perceived crisis in social
reproduction. Barz (2006) has made a congruous argument based on the
analysis of a similar set of women’s AIDS songs in Uganda. He argues
that not only have such songs contributed to the well publicised reduction
in HIV prevalence there, but that they may even operate as “therapeutic”
interventions for the performers, and he posits this data as the beginnings
of a “medical ethnomusicology” (2006, 59).
The examples from Venda clearly highlight a distinct role played by
women’s songs in AIDS prevention but, as I demonstrate below, there is a
48 Fraser McNeill

flip side to this coin. If they are to be considered meaningfully as

ethnographic evidence in a wider context of “public silence”, peer
educator songs must form part of a wider spectrum in which alternative
interpretations (musical or otherwise) of HIV/AIDS ought to furnish the
researcher with a more balanced account. The failure, to date, of a
“medical ethnomusicology” to place accounts of HIV/AIDS in this wider
performative context has thus been problematic, and this should provoke
us into widening the ethnographic lens in search of evidence that lies even
further beneath the tip of Fabian’s metaphorical iceberg.

Zwilombe guitar songs

In order to achieve this in the present analysis, we must take a closer
look at the male dominated genre of Zwilombe. The word Zwilombe
(singular: Tshilombe) is the collective noun in TshiVenda for a specific
group of male musicians. The actual term Tshilombe is derived from the
same root as Malombo spirit possession cults, u lomba (to fetch a spirit
from far away), which are closely related to what is widely known in the
region as Ngoma cults of affliction (cf. Janzen 1992). Defined by
Blacking (1965, 28) as “wandering minstrel[s]”, they perform with
guitars, dende (gourd bows), tshidzholo (an elongated zither), and mbila
(lamellaphones/hand pianos). They occupy a distinctly peripheral social
and spiritual space between ancestral possession and “madness” (Kruger
1999/2000)—a position from which they construct and project
contemporary social and political critique without fear of retribution.
Zwilombe perform at beer halls, parties, and other public occasions (and
thus move in similar social environments to the peer educators), generally
in exchange for beer, tobacco, or food. Their songs are not exclusively
about AIDS, but are based on current affairs of the day; from witchcraft
accusations to suspected poisonings, internecine disputes to car crashes.
Kruger demonstrates that Zwilombe are spiritually sanctioned and
“consequently present themselves as the voice of ancestral spirits or God”
(1999/2000, 22). From this position of relative power, Zwilombe
musicians construct models for and of social reality, or

a comprehensive, long-tem strategy which attempts to influence the

attitude and behaviour of people in the promotion of an ordered, supportive
social environment … they are prophet-musicians who act on spiritual
command … their instruments become “spirit”, symbolic extensions of
religious authority. (Kruger 1993, 510)
The song below, Zwidzumbe (Secrets), was performed for the first time
in 2000 by the famous and prolific Tshilombe Solomon Mathase. At that
“We Sing About What We Cannot Talk About” 49

time, he called it Ri Tshimbilanayo (we walk with it) and complained

through the song that promiscuous women were “carrying AIDS around
with them”, infecting men and “the sons of the nation”. Through the
verses, he called for them to adhere to the laws of “God’s country”, in
which they walk without giving due thanks or deference, causing the
nation to perish (lushaka lu khou fhela) in the process. This general theme
is remarkably similar to a song by Tshilombe Mmbangiseni Madzivhandila,
recorded by Kruger (1993, 348-356), in which promiscuous girls, with an
insatiable sexual appetite, are blamed for harbouring and spreading HIV.
The point here is that Zwilombe musicians transcend the prevalent culture
of self-censorship, and can breach the public silence on AIDS (and other
causes of death) through their spiritual sanction and regular confessions of
mental torment. Indeed, Zwidzumbe is a powerful reflection of the extent
to which (initiated) Venda men and women understand HIV/AIDS as an
element in a complex web of health and sickness, bound to notions of
secrecy and suspicion that prevent the general population from including it
in open conversation. In articulating this through song, Mathase provides
us with evidence that—when backed up with conventional ethnographic
material—demonstrates how AIDS has been incorporated into the
conservative, patriarchal folk model of sexual illness.
In the song below, then, Solomon Mathase shifts between the voices
and opinions of many different actors, simulating the “backstage” gossip
and rumour that AIDS conversation is normally restricted to. Through this
process, he poses questions and provides answers to construct a complex
representation of the ways in which HIV/AIDS is understood
conventionally by initiated Venda men and women.

Zwidzumbe (Secrets)

(1) Ula mudulu, nda ndi sa mufuni That grandchild, I really loved
zwone, him,
na ula nwanawanga, and that other child of mine,
nda ndi sa mufuni zwone vhone I really loved him too,
(2) mara u pfi o vhulaya nga yenei Aidse but they say he was killed by this
(3) vhone, vho pfa uri zwone ndi zwifhi? You, what did you hear?
(4) U la munwe a ri hu pfi hai! No! He slept with (ate) another
O tou wela who was unclean after an abortion
(5) Munwe a ri hai! Another said No!
Thiri o vha a khou pfana isn’t it that he slept with
na ula musadzi wa Vho-Mukene? the wife of Mr Mukene?
50 Fraser McNeill

(6) nwana wa vho-Mugede vhone vha you know, the child of Mr

mutendela… Mugede…
(7) ula? That one? (Do you really trust
ari thiri u tou nga o vha o bvisa thumbu? Didn’t she have an abortion?
(8) Munwe a ri hai! Another said no!
Thiri o vha ana zwilonda nga nnyoni? Doesn’t she have sores on her
(9) Munwe a ri hai! Another said no!
O vha o tou vhifha muvhilini. she was pregnant (lit. ugly in the

[Now speaking as a practicing herbalist/spiritual healer (inyanga)]

(10) Ngoho mara zwone Really, but honestly,

zwa vhukuma ndi zwifhio? Which one of these is the truth
(11) Tshihulwane, o vhuya a dida they even visited me, the great
na kha nne a ri ndi mufhe mushonga so that I could give them real
(12) A tshiyo pfana na nwana wa Vho when she fell in love with
Mukedene Mukedene’s son
(13) thumbu a yongo fara but she could not conceive a
(14) Mulandu ndi zwilonda the problem was these vaginal
zwe zwi la zwa nwe ndi li a “gokhonya” they led to her getting “gokhonya”
(15) A vho ngo vhona na ula Didn’t you see that one,
nwana wawe wa u thoma, o mbo di lovha her first child died
(16) O vha o funa o ya hangeni she should have consulted
ha “vho Nyatshavhungwa”. Mrs. Nyatshavhungwa8
Thiri na vhala vha a zwikonda? Isn’t it that she can deal with
these difficult things?

In this extract (the full version of the song lasts for over fifteen
minutes), Solomon raises closely related key issues from which we can
sketch the parameters of a Venda AIDS aetiology. They are abortion and
contraception (o tou wela or u bvisa tumbu—phrases 4 and 7), notions of
socially unacceptable promiscuity, and the role of the inyanga in treating
conditions such as vaginal sores (zwilonda, leading to gokhonya—phrases
8 and 14). These are connected, and although constraints of space prohibit
a comprehensive explanation, a partial deconstruction of this is necessary.
Towards the end of the extract, in phrase 13, Solomon makes an
explicit connection between AIDS and the woman who “could not
“We Sing About What We Cannot Talk About” 51

conceive a child”—a reference to the use of the contraceptive injection or

pill. In his own words (translated from TshiVenda):

[W]hen a woman is sleeping with a man, the bloods are not the same. This
has not really been a problem unless you were sleeping with a prostitute
who had too many mixtures in her blood and could always give you illness.
The problem now is that women are using the pills and injections from the
doctors … they get them at the clinics … I have seen them trying to hide it.
When women use the pills to prevent a pregnancy, she will get pimples and
wounds and smelly discharge from her vagina. You know women every
month they have their period—now after taking the pill the periods
disappear—so where does the dirty blood go?! It gets into their veins, they
cannot conceive and they get this AIDS—then the man gets inside her
without knowing and catches it.

As I suggested above in the discussion on female initiation, “dirty

blood” taken out of the female body through the menstrual cycle or
trapped in it through female contraception or abortion is connected to
notions of pollution, sexual taboo, and physical illness throughout
southern Africa. For Karanga speakers in Zimbabwe, it is known as svina
(lit. dirt) and avoidance of it is fundamental to the maintenance of fertility
and sexual health. In Malawi, some traditional practitioners have
conceptualised the symptoms of AIDS as a direct result of female
transgression of blood related sexual taboos (Lwanda 2003). Similar
observations have also been made in Botswana (De Bruyn 1992; Heald
2006) and South Africa (Leclerc-Madlala 2002; Liddle et al. 2005).
In Zwidzumbe, Solomon is thus articulating a widely held, but little
spoken of, “folk model” of sexual health. He is reminding his audience—
at beer drinks, taxi ranks, market places, and soccer games—that the
natural cycles responsible for the removal of pollution from the female
body are central to healthy individual, and thus social, reproduction. The
introduction of methods of contraception such as the pill, the injection, or
abortion put this system in immanent danger by regulating the flow in
“unnatural ways” that result in the trapping of pollution.9 Women who
use—or are suspected of using—these forms of contraception thus harbour
dams of dangerous pollution that will eventually make them, and any man
who comes into their sexual contact, sick. As we see in Solomon’s
statement above, so-called “dirty blood” has “not really been a problem
unless you were sleeping with a prostitute … the problem now is that
women are using the pills and injections from the doctors … they get them
at the clinics … I have seen them trying to hide it.”
In addition to full- or part-time sex workers who have “always” been
bearers of illness, then, there is the more recent category of women using
52 Fraser McNeill

biomedical techniques to regulate their fertility. With their prolific

knowledge of biomedical concepts, declared publicly on a weekly basis
through songs, peer educators fall directly into this recently-constructed
category. They are guilty through association; caught in the web that
connects their expert knowledge of biomedicine with direct experience of
its use. The diseases they are believed to harbour take a variety of forms,
such as zwilondo (open genital sores that resemble the third phase of
syphilis), gokhonya (also known as lukuse—through which a woman’s
children die in infancy), tshidzwonyonyo (a red, burning rash on the
genitals similar to thrush), or tshimbambamba (through which a woman
develops yellow pimples in the genital area that burst during sex and cause
a white, smelly discharge similar to gonorrhoea). This complex of sexually
transmitted infections is known in TshiVenda as malwadze dza
vhafumakhadzi—the illnesses of women, and as a blood borne infection
transmitted through sexual intercourse, HIV has been incorporated into
this framework with ease.
In this way, even a cursory look at an alternative musical interpretation
of HIV/AIDS gives us an indication as to the ways in which peer
educators—and the biomedical gospel with which they proselytise—can
be constructed as carriers of the virus, as opposed to cornerstones in the
promotion of its avoidance. As young women fluent in biomedical
discourse from which the contraceptive pill and injection emerged, they
are viewed suspiciously by men who may otherwise consider them
potential sexual partners. Moreover, they have broken the public silence
on HIV/AIDS in a way that renders them vulnerable to implication in the
spread of the disease. Unlike Zwilombe musicians, whose spiritually
sanctioned social critique is legitimated by claims to madness, I have
sought to demonstrate that peer educators have no such haven of
liminality, and their conspicuous knowledge of unnatural death places
them in a very ambiguous position. As it stands, current biomedical
interventions are easily incorporated into the pre-existing folk model, with
the result that most men continue not to worry about how they sleep with a
woman (i.e. with or without a condom), but rather make choices about
sexual encounters based on who they sleep with (in terms of a woman’s
perceived potential for harbouring pollution).

It appears, then, that in a socio-cultural environment characterised by
the avoidance of open conversation about highly stigmatised topics, music
provides a potentially fruitful avenue for ethnographic analysis. Whilst
“We Sing About What We Cannot Talk About” 53

conventional ethnographic methods may lead to conclusions based on

notions of “public silence” (albeit with private acceptance of AIDS as a
cause of death) or “collective denial”, performance-based evidence
provides a distinct line of enquiry that allows the ethnographer to build an
analysis not on what people say (or cannot say), but on what (and why)
they choose to sing. In the present case, this has taken us from the complex
and multi-dimensional uses of music incorporated into peer education
projects to the realisation that such an enterprise may, in fact, be
understood in contradictory terms by many of those in the audiences of
peer educators’ public performances.
But if this chapter has demonstrated that a culturally nuanced content
analysis of songs can reveal important aspects that may well constitute
ethnographic evidence where it appears to be thin on the ground, then it
also points towards the need to exert caution in doing so. Both the peer
educators and Zwilombe musicians represent divergent worldviews,
expressed through different artworlds. Taken as separate bodies of
evidence, or as two accounts of the same perceived crisis in social and
sexual reproduction, they could potentially lead the ethnographer to
contradictory conclusions about the role of music in constructing and
articulating interpretations of HIV/AIDS. It is only through a combined
analysis, in the absence of open conversation, that we may venture into
making tentative conclusions, and begin the process of making sense of it
all through supplementary forms of evidence such as gossip, interviews, or
observations “on the ground”.

I wish to acknowledge financial assistance from the Economic and
Social Research Council (UK) for funding the research upon which this
article is based, and to thank my research assistant Mushaisano Colbert
Tshivhase, Mashudu Madazhe for permitting research with the Centre for
Positive Care, and Solomon Mathase for permission to write about his
songs. I would also like to thank Casey High, Liana Chua and Timm Lau
for organising the "Questions of Evidence" conference, inviting me to
participate, and for providing me with insightful comments on earlier
drafts of this chapter.

Venda was a former “independent” homeland under the Apartheid regime. Since
1994, it has been incorporated into the Northern (later Limpopo) Province of South
54 Fraser McNeill

Africa—bordering with Zimbabwe to the north and the Kruger National Park to the
east. The current argument is based on several periods of fieldwork in this region,
stretching over the last ten years.
Fabian and Erlmann here use “consciousness” to refer to the active process in
which human actors deploy historically salient cultural categories to construct their
self-awareness (Comaroff and Comaroff 1987).
It is worth emphasising here that the social environment in question is quite
different to the more liberal urban cityscapes of Johannesburg, Durban, or Cape
Town, where so-called civil society movements, notably the Treatment Action
Campaign (TAC), have made significant in-roads encouraging people living with
HIV/AIDS to disclose their status. In South Africa’s cities, the TAC has instigated,
with at least partial success, a long process designed to combat the social stigma
attached to HIV infection. However, the TAC has no significant presence in the
Venda region, and it remains extremely rare for an individual in this region to live
openly with HIV or AIDS. Indeed, in over ten years, I have known of only two
people who disclosed their positive status in public.
Due to space constraints I provide only a small number of examples here.
However, I have digitally recorded and transcribed over two hundred such songs.
The role that music played in the struggle was recently commemorated in the
excellent SABC documentary “Amandla!—A revolution in four-part harmony”
(BMG South Africa 79102-21510-2).
The version was recorded by Canadian journalists at the ANC headquarters in
Lusaka, Zambia, in 1985 and later appeared as track 7 on the commemorative
Radio Freedom CD (Rounder Records USA 11661-4019-2). Radio Freedom was
the “voice of the ANC”, broadcast from countries throughout southern Africa.
The crocodile is a symbol of chieftainship, and of incontestable strength.
This refers to a specific group of traditional healers called maine. They are
comparable to family doctors and specify in treating specific ailments. See below.
The case of abortion is more complex than that of contraception. The folk model
holds that abortion can be conducted safely through the use of specific type of
traditional healer who conducts rites of purification afterwards. The objection held
by Solomon in the song is that in bio-scientific clinics this purification is avoided.

Ashforth, Adam. 2004. An epidemic of witchcraft? The implications of
AIDS for the post-Apartheid state. African Studies 61 (1): 121-145.
—. 2005. Witchcraft, violence and democracy in South Africa. London:
University of Chicago Press.
Blacking, John. 1965. The role of music in the culture of the Venda of the
Northern Transvaal. In Studies in ethnomusicology, Vol. 2, ed.
Mieczyslaw Kolinski. New York: Oak Publications.
“We Sing About What We Cannot Talk About” 55

—. 1969. The songs, mimes, dances and symbolism of Venda girls’

initiation schools. Part 1: Vhusha. African Studies 28 (1): 3-35.
—. 1973. How musical is man? London: University of Washington Press.
—. 1985. Movement, dance, music and the Venda girls’ initiation cycle. In
Society and the dance: The social anthropology of process and
performance, ed. Paul Spencer. Cambridge: Cambridge University
Barz, Greg. 2006. Singing for life: HIV/AIDS and music in Uganda. New
York: Routledge.
Campbell, Catherine. 2000. Selling sex in the time of AIDS: The psycho-
social context of condom use by Southern African sex workers. Social
Science and Medicine 50:479-494.
—. 2003. Letting them die: How HIV/AIDS prevention programmes often
fail. Johannesburg: James Currey.
Comaroff, John, and Comaroff, Jean. L. 1987. The madman and the
migrant: Work and labour in the historical consciousness of a South
African people. American Ethnologist 14 (2): 191-209.
—. 2004. Notes on Afromodernity and the neo world order: An afterword.
In Producing African futures: Ritual and reproduction in a neoliberal
age, ed. Brad Weiss. Leiden: Brill.
De Bruyn, Maria. 1992. Women and AIDS in developing countries. Social
Science and Medicine 34 (3): 249-262.
De Gruchy, John. W., and Steve de Gruchy. 2004. The church struggle in
South Africa. London: Augsburg Fortress.
Erlmann, Veit. 1996. Nightsong. Chicago: University of Chicago Press.
Fabian, Johannes. 1990. Power and performance: Ethnographic
explorations through proverbial wisdom and theatre in Shaba, Zaire.
Wisconsin: University of Wisconsin Press.
Farmer, Paul. 1990. Sending sickness: Sorcery, politics and changing
concepts of AIDS in rural Haiti. Medical Anthropology Quarterly 4
(1): 6-27.
Gell, Alfred. 1999. The art of anthropology: Essays and diagrams. LSE
Monographs on Social Anthropology. London: Continuum International
Publishing Group.
Gilbert, Shirli. 2007. Singing against Apartheid: ANC cultural groups and
the international anti-Apartheid struggle. Journal of Southern African
Studies 33 (2): 423-441.
Good, Byron J. 1994. Medicine, rationality and experience: An
anthropological perspective. Cambridge: Cambridge University Press.
Heald, Suzette. 2006. Abstain or die: The development of HIV/AIDS
policy in Botswana. Journal of Biosocial Science 38:29-41.
56 Fraser McNeill

James, Deborah. 1999. Songs of the women migrants: Performance and

identity in South Africa. Edinburgh: Edinburgh University Press.
—. 2002. To take the information down to the people: Life skills and
HIV/AIDS peer educators in the Durban area. African Studies 61 (1):
—. 2006. Black background: Life history and migrant women’s music in
South Africa. In The musical human: Rethinking John Blacking’s
ethnomusicology in the twenty-first century, ed. Suzel A. Reily.
Aldershot: Ashgate.
Jansen, John. M. 1992. Ngoma: Discourses of healing in Central and
Southern Africa. Oxford: University of California Press.
Kruger, Jaco. 1993. A cultural analysis of Venda guitar songs. PhD thesis,
Rhodes University, South Africa.
—. 1999. Singing psalms with owls: A Venda twentieth century musical
history. Part 1: Tshigombela. African Music 7 (4): 122-146.
—. 1999/2000. Of wizards and madmen: Venda Zwilombe. Part 1. South
African Journal of Musicology 19/20:15-29.
La Fontaine, Jean S. 1985. Initiation. London: Penguin.
Lambek, Michael. 1993. Knowledge and practice in Mayotte: Local
discourses of Islam, sorcery and spirit possession. Toronto: University
of Toronto Press.
Leclerc-Madlala, Susan. 2002. On the virgin cleansing myth: Gendered
bodies, AIDS and ethnomedicine. African Journal of AIDS Research
Liddle, Christine, Louise Barrett, and Moya Bydawell. 2005. Indigenous
representations of illness and AIDS in Sub-Saharan Africa. Social
Science and Medicine 60:691-700.
Lwanda, John. 2003. The (in)visibility of HIV/AIDS in the Malawi public
sphere. African Journal of AIDS Research 2 (2): 113-126.
Mann, J. 1987. Statement at an informal briefing on AIDS to the 42nd
session of the United Nations AIDS General Assembly. New York, 20
McGregor, Liz. 2005. Khabzela: The life and times of a South African.
Johannesburg: Jakana.
McNeill, Fraser G. 2007. An ethnographic analysis of HIV/AIDS in the
Venda region of South Africa: Politics, peer education and music. PhD
thesis, University of London.
Nowlin, Bill. 1996. Liner notes to Radio Freedom: Voice of the African
National Congress and the People’s Army Umkhonto We Sizwe.
Rounder Records ISBN 11661-4019-2.
Reid, Graham, and Liz Walker. 2003. Secrecy, stigma and HIV/AIDS: An
“We Sing About What We Cannot Talk About” 57

introduction. African Journal for AIDS Research 2 (2): 85-88.

Stadler, Jonathan J. 2003. The young, the rich, and the beautiful: Secrecy,
suspicion and discourses of AIDS in the South African Lowveld.
African Journal for AIDS research 2 (2): 127-139.
Stayt, Hue. 1931. The Bavenda. London: Oxford University Press for the
International Institute of African Languages and Cultures.
Turner, G., and J. Shepherd. 1999. A method in search of a theory: peer
education and health promotion. Health Education Research 14 (2):
Turner, Victor. 1982. From ritual to theatre: The human seriousness of
play. New York: PAJ Publications.
UNAIDS. 2007. South Africa country profile.
Vail, Leroy, and Landeg White. 1983. Forms of resistance: Songs and
perceptions of power in colonial Mozambique. American Historical
Review 88 (4): 883-919.
—. 1990. Power and the praise poem. Charlottesville: University of
Virginia Press.
Van Warmelo, Nicholas J. 1989. Venda Dictionary: Tshivenda-English.
Pretoria: J.L van Schaik.
Wilson, David. 2000. Peer group education guidelines: Volumes 1-6.
Unpublished training manual. Department of Psychology, University
of Zimbabwe, Harare.


This chapter discusses the multiple forms that audiovisual media—
specifically video—take in the production of anthropological knowledge.
It draws from my own experience of using video as part of my doctoral
research to demonstrate how the dualities suggested by participant-
observation operate simultaneously in the experience of producing an
ethnographic video. I discuss specifically how an ethnographic video at
once generates material evidence of social relations and constitutes a body
of knowledge. Crucially, in this situation, the process of producing the
audiovisual itself becomes a form of anthropological knowledge.
Making an ethnographic video as part of an academic research
programme challenged me to consider two different conceptual and
technical ways of approaching my fieldsite. One was text-based,
argumentative, and supported by academic references; the other image-
based, suggestive, and composed solely of materials I collected in the
field. In retrospect, the difficulty I encountered in bringing these two
approaches together into one project was mainly due to my assumption
that they were incompatible and needed to be addressed independently.
Throughout my research I found that both items, the video and the text,
were born out of the same process of fieldwork, and responded to the same
constellation of circumstances and subjectivities that characterised my
research experience. In this chapter, I argue that the main challenge
presented in a social anthropology project that uses audiovisual media
does not only concern the productive use of the visual versus the textual,
or the subordination of one to the other. The challenge, in my experience,
is to embrace and navigate the tension generated between the different
types of knowledge with which one is confronted; the final product
reflecting how such a tension was negotiated or balanced out.
Audio-Visual Evidence and Anthropological Knowledge 59

My suggestion is that audiovisual media, by their capacity to produce

meaning at various levels (through their content and form, for example),
are capable of presenting and representing different types of knowledge
simultaneously. In this chapter I will discuss how audiovisuals carry out a
double role in anthropology. I will argue that this double function occurs
through ethnographic video’s capacity to record events and processes that
can be scrutinised at a later time, while at the same time offering an open-
ended, suggestive, and reflexive account of fieldwork. In both cases,
knowledge is generated in a peculiar state of awareness that is mediated
through the camera and sound recording equipment.
I will begin by discussing an application of audiovisual media that
prioritises their function as material evidence of social events and
processes. In this approach, films are treated as exemplary recorders of
ethnographic data, capable of capturing social events and processes as they
occur in real time. To make films in the spirit of collecting evidence
suggests that audiovisuals contain some kind of consistency or truth-value,
perhaps by virtue of being experienced through the visual and aural senses,
from which anthropological knowledge can be extrapolated and validated.
However, in this approach, the consistency of audiovisuals is merely
technical rather than analytical, requiring an external component such as
text or narration to communicate ethnographic knowledge. The production
of audiovisuals for evidentiary purposes can be a valuable addition to an
ethnographic project because their examination can provide insights that
were not appreciated with direct observation, but were realised after
careful and repeated examination of the footage.
I will then address an alternative framework by which anthropologists
use audiovisuals as the principal means of carrying out their ethnographic
research. In this approach, the very experience of making and watching a
film constitutes anthropological knowledge. Audiovisuals are valued for
their capacity to communicate a sense of social experience through their
cinematic and auditory complexity. They thus invite the viewer's
imagination and intelligence to explore the sensuality and sensibility of the
sounds and images, perhaps better than any other medium, and are capable
of presenting and representing the totality of movement, vision, and sound
simultaneously “as the originary structures of embodied existence and the
mediating structures of discourse” (Taylor 1996, 80). This approach
suggests that images and sounds precede knowledge; they constitute what
we know before explanation (MacDougall 2006). They allow us to
examine processes and events in their state of being, untapped by political
and historical contexts yet mediated by the filmmaker’s subjective stance
towards the film subjects and fieldsite. Knowledge, from this perspective,
60 Carlo A. Cubero

is not exclusively related to the historical, discursive, or political. Rather,

the production of knowledge is contingent on the experiential, the
propositional, and the affective, and is generated in an inter-subjective
relationship between the filmmaker, film subjects, and the viewer
performing an exploration of lived experience in all its ambiguity (Taylor
1996, 80).
Finally, I draw on my own experience of making an ethnographic
documentary as part of my PhD project in order to discuss the ways in
which the process of making a video produced knowledge at the same time
as it created audiovisual objects from which analytical concepts were
extrapolated. My dissertation includes a fifty-minute ethnographic film,
called Mangrove Music, which explores the ways in which mobile and
insular practices and discourses operate simultaneously in the constitution
of an island musical identity on the Spanish-speaking island of Culebra in
the northeastern Caribbean. I focus on two specific ways in which making
a video as part of my fieldwork sensitised me to approach my fieldsite in a
way that enriched the research process and led to the inclusion of
analytical categories that would not otherwise have been part of my
dissertation. I first show how my continuous attention to the landscape of
Culebra, an exercise that was born out of collecting footage that was not
intended to form part of the final video, provoked me to approach my
writing from a perspective that connected islanders’ aesthetic
qualifications of the landscape to contemporary political controversies. I
then discuss how the experience of filming and editing a seemingly
ordinary and routine musical performance led me to consider interpretive
ways in which I could explain aspects of the creative and performative
processes of the musicians I was filming. In each case, the filming process
worked through different channels and to different ends. However, I
suggest that both cases reflect audiovisual media’s capacity to produce
evidence and knowledge simultaneously.

Film as evidence
The idea that audiovisuals function as a device for collecting evidence
in the field dates back to the origins of modern anthropology itself. At the
turn of the twentieth century, anthropologists applied the recently invented
film camera to the study of surface markers of cross-cultural difference,
such as skin colour, body shape, and technology, using language and
taxonomy associated with biology (Hart 1998). The camera was viewed as
a research tool that, by virtue of being a mechanical object detached from
the body, could record social interactions and processes objectively. The
Audio-Visual Evidence and Anthropological Knowledge 61

assumption was that mechanically-created images and sounds could be

scrutinised and exhibited alongside other materials collected during
fieldwork, such as objects made by informants, in order to extract from
them a narrative that would give insight into the social relations and
subjectivities characteristic of the fieldsite.
An oft-cited precedent of this approach is the 1898 Cambridge
Anthropological Expedition to the Torres Strait led by Alfred C. Haddon.
The intention of the expedition was to produce a taxonomical record of
cultures and societies under threat by imperialist expansion (ibid.). The
expedition included film equipment, and produced a series of short sets of
rushes depicting specific events and practices understood to be in the
process of disappearing. Cinematically, these sequences were characterised
by a static camera on a tripod, which filmed whole bodies in their
immediate space. Although the rushes suggest that Haddon’s team was
trying to record authentic moments of social life in real time, the irony is
that we now know that the team had in fact persuaded their informants to
reproduce performances that had not been practised for a generation. This
contradiction did not dissuade anthropologists from purveying the notion
that film would be an appropriate scientific data-gathering device, akin to
the astronomer’s telescope or the biologist’s microscope. Haddon and his
team designed their audiovisuals to function as reliable data to sustain
arguments and corroborate observations, rather than to narrativise or to
turn audiovisuals into a channel of anthropological discourse.
Almost forty years later, Margaret Mead and Gregory Bateson used
audiovisuals to construct and represent a narrative that would
communicate anthropological knowledge. They intended to use film in
order to explore internal processes that led to the formation of cultural
attitudes. Mead and Bateson aimed to document, with film and
photographs, external expressions of socially learned behaviour—
specifically child rearing and non-verbal communication—and evaluate
how these physical expressions reflected cultural values and relationships.
Their footage and thousands of photographs were intended to accompany a
text and construct an argument about the culture of Bali and describe an
“ethos” of the people.
The social ethos that interested Mead and Bateson was more
behavioural than performative. Their attempts to address historical cultural
patterns and broader social contexts, as they relate to internal processes,
proved elusive to the cinematic and narrative techniques that they
employed. Their cinematic methodology and narrative choice suggests that
they—particularly Mead—saw photography and film as an enhanced note-
taking device, as a record from which direct observations made in the field
62 Carlo A. Cubero

could be corroborated and examined in more detail after the filming had
taken place. It was assumed that the footage would be a theoretically
neutral piece of evidence to which the anthropologist could refer when
constructing the ethnography (Mead 1962, 138, cited in El Guindi 2000,
The seven edited films produced in this collaboration rely heavily on a
didactic voice-over that explains and interprets for the audience the
dynamics presented in the film, as if Mead herself were guiding us through
her movie, leaving little room for alternative interpretations. The narration
is used to corroborate what the images are already showing, or to offer an
interpretation that lies outside of the film itself, such as an explanation of
the broader social and political context in which the particular scene takes
place. This narrative choice suggests that the images and sounds are
subordinate to the voice and the text, but it can also indicate a possible
predisposition of the research (MacDougall 2006, 224). Film and
photography were not the means through which Mead and Bateson did
their research, but rather constituted a method through which they used the
camera to enhance their direct observations. The Mead-Bateson project
valued audiovisuals for their illustrative and didactic potential. However,
they did not see the camera as mediating the relationships fostered in
fieldwork, and the film remained a detached observation of events and
processes. The pedagogical value of these films lies in their capacity to
illustrate abstract anthropological concepts by showing how they appear in
the field, much like an illustrated lecture, offering the student a richer
learning experience.
The intentions and outcomes of the Mead-Bateson project demonstrated
the difficulties of communicating anthropological knowledge without
accompanying narration. In her later writings, Mead would argue that
anthropology’s historical tradition as a text-based discipline hindered the
development of audiovisual techniques and sensibilities within
anthropology (Mead 1975). At the core of Mead's argument is the
suggestion that audiovisuals and texts operate in fundamentally different
ways, and that the project of using audiovisuals in a “discipline of words”
(ibid.) may require a re-examination of what are valid anthropological
questions, as well as the qualities of anthropological knowledge (Pink

Film as knowledge
Following an alternative framework to the traditional practice of
collecting film as unmediated evidence, a series of filmmakers have been
Audio-Visual Evidence and Anthropological Knowledge 63

striving to use film to produce anthropological knowledge and move

beyond the use of audiovisual for illustrative or ancillary purposes. They
draw their inspiration from film theorists such as Béla Balázs and Sergei
Eisenstein, pioneers in writing about film’s capacity to communicate
abstractions, and from film practitioners such as Dziga Vertov and Robert
Flaherty, whose works are renowned for their reflexivity, collaboration,
and narratives that place viewers at the centre of the action (Loizos 1993;
Winston 1995).
In this approach, anthropological knowledge is produced through a
reflexive engagement with the film’s subjects and fieldsite, which is
socially mediated by cameras and sound recording equipment. The
knowledge these anthropologists wish to produce is contained within the
films themselves. This approach is primarily interested in communicating
the experiential and affective dimensions of social life, rather than being
motivated by the precision of scientific methods (MacDougall 2006; El
Guindi 2000). Their narratives are designed to communicate a sense of
place and show the complexity of experience that characterises people’s
daily life, as opposed to the demonstrative documentary (Winston 1995).
Filmmakers associated with this approach strive to make films that evoke
an intimate and empathetic relationship with their film subjects.
Two filmmakers associated with this approach are John Marshall,
known for his films amongst the !Kung of the Kalahari Desert, and Jean
Rouch, best known for his work in francophone Africa. While each
filmmaker is associated with employing different cinematic techniques and
working methods, this body of work shares an agenda of attempting to
understand, through filming, a social world from within the physical
spaces of the film subject, which necessarily entails a collaborative effort
between the filmmaker and the film subjects (MacDougall 2006).
An important characteristic of these filmmakers’ approach is an
interest in the workings of film language and a desire to grasp the
imaginative and intellectual attention of the audience through the
audiovisual medium. To achieve this engagement requires an honest
assessment of the distinct potential and limitations of text and visual
media, and an attempt to exploit the visual accordingly. Hence, the
ethnographic questions posed in these films and the ways of answering
them are necessarily of a different nature to the pursuits of anthropologists
like Mead and Bateson. For example, instead of looking for underlying
structures of culture, which relate more to discursive formations and
argumentation, their work suggests that the demonstrative and expository
language of film is well-suited to address the embodied experience of
individuals, the relation of people to places and material objects, the
64 Carlo A. Cubero

performative aspects of social life, emotions, time, the body, and the
textures and nuances of a place (MacDougall 2006, 220).
The inter-subjective accounts of these filmmakers recognise that
anthropological knowledge is the product of relationships of power
between the anthropologist and informants; the edited film is the result of
such a negotiation. The idea that audiovisuals can convey a sense of social
experience embraces the communicative ambiguities of non-textual media
as a stimulating and creative tension from which anthropologists can raise
questions about the properties of culture.

Fieldwork with video

While the two approaches I have outlined above may appear
antagonistic (see Grimshaw 2001; Hastrup 1992; Loizos 1992; Taylor
1996), I did not experience a similarly clear-cut contradiction in my own
field research. I found that the process of making a video as part of an
academic project placed me in a position where I relied on audiovisuals’
capacity to function as evidence of social relations—as an object that I
collected during fieldwork and referred to on occasion in the writing of my
dissertation—and as a form of anthropological knowledge, whereby the
process of filming, editing, and showing the film itself constitutes
knowledge. The following sections examine how these two paradigms
continuously folded into each other during my fieldwork, rather than
representing mutually exclusive discourses and practices.
My experience points to the multivalency of audiovisual media and its
capacity to generate multiple affectations and reactions in different
contexts, without altering its actual content. In the present case, I was
attempting to construct a cinematic narrative by using materials I collected
in fieldwork, not all of which I had originally accrued for that purpose.
When I returned from the field and examined the footage, I realised that
some of the material I shot that was intended for the video did not fit with
the themes and arguments that I wanted my project to address. In the same
way, some footage that I shot for more personal reasons ended up
becoming important scenes in the final edit. This was possible because of
video’s peculiar capacity to manifest multiple meanings and purposes in
the same images and sounds. It is through this ambiguity that video
expresses its double condition as a medium that can engage with the
viewer imaginatively and experientially, while at the same time containing
evidence of fieldwork and social processes that can enrich the descriptive
and analytical dimensions of written ethnography.
Audio-Visual Evidence and Anthropological Knowledge 65

My research project was concerned with the social processes by which

Caribbean island spaces are constituted and lived. In my dissertation I
examined different ways in which Caribbean migrations, creolisations,
transnationalisms, and other discourses and practices of mobility and
travel are produced and reproduced in relation to colonial and post-
colonial discourses that separate and fragment the region into discrete
island and regional spaces. The video included in my dissertation explores
the ways in which mobile and insular discourses and practices are related
to the construction of musical identities on the Spanish-speaking island of
Culebra, an offshore municipality of Puerto Rico. The main goal of the
video was to explore how the production and reproduction of a musical
insular identity is formed and motivated by the mobile dimensions of life
on Culebra.
My intention was to make this ethnographic video in an observational
style. My interest in observational cinema stemmed from my training in
Visual Anthropology at the Granada Centre for Visual Anthropology in
the University of Manchester. An important pedagogical goal of the
Granada Centre, at the time, was to place the student in a position to
consider the research process as sensorial and experience-based learning.
Ethnographic filmmaking was not about adding filmmaking to
anthropology but to visualise anthropology, which involves

a fundamental reorientation of perspective such that the world is not

primarily approached discursively through language, explanation,
generalisation; but through a re-embodiment of the self as the foundation
for renewed engagement with everyday life. (Grimshaw and Ravetz 2005,

At the Granada Centre we were encouraged to explore this renewed

engagement with the quotidian through observational cinema, understood
broadly as a methodology and cinematic aesthetic that is associated with
“ethnographic cinema”, “cinéma vérité”, and “direct cinema”. Working
observationally challenges the student to think visually and to look at a
fieldsite through the quality of its space and through the daily situations
and processes that take place within it; to appreciate the place’s sense of
time, pace, textures, aural, and visual qualities alongside the relationships
that these quotidian processes and events suggest. In these exercises, the
students were provoked to consider the production of knowledge by
looking at processes and events in their state of “being” (MacDougall
2006, 1-9). Looking at something “as it is” prompts questions that lead
towards a more experiential knowledge, stimulating the observer to be in
tune with the corporeal and emotive qualities of perception.
66 Carlo A. Cubero

Amongst the conventions of observational cinema that students were

encouraged to reproduce was to work alone, perhaps mimicking the
practice of classical anthropological fieldwork. The student was in charge
of the sound, image, and editing of the video. This practice suggests that
the researcher’s relationships in the field are mediated by the camera and
the equipment, that one’s identity as a researcher is enacted as a
cameraperson, sound person, video editor, and writer. Making a video
therefore constitutes doing ethnography, and vice-versa.
During my fieldwork and writing, I found myself continuously gauging
the narrative possibilities for my text and video. This meant that I was in a
position to take note of the visual, expository, and demonstrative aspects
of my fieldsite at the same time as I took note of the political, historical,
and social contexts at the heart of my research. This challenge became
more difficult in the post-fieldwork stage, when I felt that I had to
reconcile two disparate ways of thinking about my fieldwork. In
retrospect, I have come to realise that all along, and perhaps unbeknownst
to me, the visual and the textual were folding into each other throughout
my PhD research, and that both products emerged out of the same
experience, continuously informing each other during my fieldwork and

Understanding the political through the audiovisual

Throughout my year of fieldwork, the main village on the island of
Culebra1 was undergoing a process of urban regeneration. The island’s
municipal government enacted the development programme with the
financial support of the Puerto Rican state. The construction projects were
delayed because the engineers had overlooked the fact that, historically,
the village was constructed over reclaimed swamplands, such that it
flooded when the ground was excavated. Heavy machinery and special
cement pipes were brought in to deal with the flow of water and mud
seeping from under the main street. I collected a lot of footage of
bulldozers pushing mud, their mufflers belching smoke while the workers
applied and spread cement as fast as they could on the bits of the street
that were dry. It was a noisy and smelly affair that suggested to me a kind
of violent modernity that countered my previous experiences of visiting
the island years before, when I had admired its seemingly pristine
I continued focusing my attention on the landscape, appreciating the
changing colours of the sea as the days passed, the different tones of the
grass according to the rainy or dry season, and the sound of the wind
Audio-Visual Evidence and Anthropological Knowledge 67

ruffling the trees in my backyard. I contrasted these sensations of the

landscape with the mechanical sounds, mud, and smoke of the village.
This exercise of filming the construction sites and my appreciation of the
island’s landscape developed as a tangent to the core themes of my
research project.
As my fieldwork progressed, I gained a more sophisticated
understanding of the political dimensions of the island’s development
programme and the variety of social contests that were arising from it. The
goal of the infrastructure development, as stated publicly by the mayor, his
advisers, and the Puerto Rican senators who visited the island to inspect
the construction, was to alleviate poverty in Culebra. The infrastructure
development was justified as part of a broader goal to pair the socio-
economic resources of Culebra with Puerto Rico, and further democratise
Puerto Rico’s resources. The notion that Culebra lagged behind Puerto
Rico’s national development agenda was, in many ways, reproduced by
Culebra’s physical separation from Puerto Rico (they were connected by
an irregular ferry service). The antiquated architecture of Culebra’s main
village and the island’s dry landscape dominated by thorny trees were
often described in contrast to Puerto Rico’s lush rainforest interior and
dense urban spaces.
The main critics of the development programme, including those from
within the mayor’s own political party, pointed to the fact that the projects
were targeted more to expand tourism facilities than to address the genuine
needs of the islanders. These critics argued that the type of tourism
development proposed by the municipality responded more to
development plans originating in Puerto Rico, which in turn followed
planning discourses of tourism resorts in the United States. Tourism
development, the critics argued, should follow the example of other West
Indian islands that had successfully implemented eco-tourism policies or
maintained a high degree of control over their own tourism development
plans. There was a concern that the type of tourism development that the
mayor was implementing on the island would foster processes of land
privatisation and disenfranchise islanders from the island’s resources,
especially in the coastal areas.
As I wrote about this controversy in my dissertation, I became
increasingly frustrated in attempting to communicate how the political
camps that emerged from such a social contest did not follow the
traditional colonial politics of Puerto Rico and Culebra. The traditional
political parties and interest groups of Culebra were either fragmented or
held inconsistent positions. I also found that basic social and economic
data about the island, such as population statistics, unemployment rates,
68 Carlo A. Cubero

and income per household, were inconsistent with arguments and

expressions printed in the media and voiced in public meetings. To
approach this conflict through government statistics and the island’s
political parties, which were predicated on regional colonial politics,
would paint a contradictory picture of the ways in which political
processes are connected to Culebra islanders’ sense of island identity.
Instead, I focused on the different types of imagery that were evoked in
the rhetoric of development debates on the island. I found that when
people discussed their positions on the development, whether in public
meetings, published editorials, or private conversations, they referred
approvingly or disapprovingly to the mayor’s understanding of the
Culebra landscape. Critics of the development programme did not see the
island as harsh, inhospitable, and in need of domesticating, but as having a
beautiful natural landscape that had served as a means of subsistence in the
past and is the main tourist attraction of the island today. Critics also made
historical connections between the current struggle for their control of the
lands and the struggles carried out thirty years ago, when the islanders
mobilised themselves to successfully oust a US navy base that was using
the island as a practice bombing range. Corroded war tanks and shrapnel
still dot the island’s beaches and hills, and I often heard people draw
parallels between military tanks and bulldozers, and manned checkpoints
that blocked access to militarised zones during the 1970s and
contemporary “Private Property” signs that restrict public access to paths
and the coast.
I found that the arguments and rhetoric deployed during the
development debates were mediated through the visual; they placed the
island’s natural landscape at stake by drawing on the relationship between
the island’s appearance and a sense of history. For the development critics,
the notion of being a person from Culebra, or culebrense, was rooted in a
positive valuation of the landscape, which related to their memory of
struggles for access to it. In this perspective, being culebrense also evokes
a connection to the English-speaking Caribbean through a history of
movements throughout the archipelago and the aesthetic similarities
between the smaller islands of the northeastern Caribbean. For the
development supporters, to be a culebrense was to be Puerto Rican and an
active contributor to the Puerto Rican national project. They suggested that
the people of Culebra should demand that the Puerto Rican state correct
infrastructure disparities between Puerto Rico and Culebra following
Puerto Rico’s model. A truly democratised Puerto Rico, the development
argument suggested, would entail Culebra achieving an urban landscape
that was aesthetically compatible with Puerto Rican national standards.
Audio-Visual Evidence and Anthropological Knowledge 69

I suggest that my sensitivity to the visual and aural properties of my

fieldsite emerged from a heightened state of awareness that came from
recording video footage for an extended period of time.2 In order to be
successful in my filming, I could not record dispassionately. I had to film
knowing that my footage would be edited later into a narrative. In order to
do this, I had to be actively engaged in the events that I was recording, be
aware of my shots and their sequence, attentive to detail, as well as patient,
sensitive, and respectful of my surroundings. I had to maintain a sense of
confidence and determination in what I was doing, both technically and
conceptually, while at the same time fostering a relationship of trust and
empathy with my informants.
This state of awareness and excitement would not subside after a
recording session was finished. My eyes would continue to sting for hours
after recording and, when in certain social interactions, I would often find
myself “practising” camera movements and shot positions without my
camera, as if I were still filming. Recording video in the way that I did
caused me a certain visual excitement that continued throughout my
fieldwork. I suggest that this visual sensitivity had a significant effect on
the way in which I approached the social processes that I was trying to
This excited visual state resonates with the experience of other
filmmakers, such as Dziga Vertov, whose writings suggest that his
engagement with the world through filming offered him a special insight
into the world, which he labelled “film-truth” (Vertov 1963, cited in
Rouch 1975, 91). Also, Jean Rouch described his filmmaking
experience—both the filming and editing—as a kind of trance or “cine-
trance”, in which the camera and the body of the filmmaker were meshed
together in the process of observing and participating in social events. For
Rouch, the camera acquires an agency during the filming process. He
called the film camera a “capricious but faithful little machine, which has
an infallible visual memory” (ibid., 82) that participates alongside the
filmmaker and the film subjects in an improvised “ballet in which the
camera itself becomes just as alive as the people it is filming” (ibid., 89).
In my case, the video equipment became a powerful mediator and
instigator in the process of gaining an understanding of my fieldsite at the
same time as it functioned as a valuable collector of ethnographic evidence
that I referred to while writing my dissertation.
I did not actively seek to visualise the Culebra development debates
and write about them as such. I believe that my decision to write about and
understand this process from this perspective emerged from the experience
and mindset of making a video. That is, making a video led me to
70 Carlo A. Cubero

appreciate the descriptive and evocative dimension of my fieldwork and

the processes that characterised it. It offered a way to navigate the
complexities and contradictions of this particular political process.

Video recording as knowledge

Another instance where recording and editing video research materials
unexpectedly became an experience that I incorporated into my writing
occurred early in my fieldwork, while filming an otherwise ordinary
musical performance. The most active music group in Culebra at that time
was a conga percussion trio called The Wiki Sound. I had chosen to
include them in my video because the type of music they played and the
instruments they used suggested the histories of movement and insularity
that were at the core of my research project. More importantly, I had
established a good relationship with the three musicians, which, combined
with their charisma and sense of humour, made filming with them very
During the Christmas season, three months into my fieldwork, The
Wiki Sound invited me to a private party in San Juan, the capital of Puerto
Rico, where they had been hired to play. I had recorded their musical
performances before and had a general idea of how I wanted their
performance to be shot. Still, I was particularly excited and nervous about
the trip because I thought that a journey—with its clear beginning, middle,
and end—would provide a natural narrative for a sequence in the video,
should I choose to use it in the edit. We left on the morning ferry to
mainland Puerto Rico and took public transport to the capital. I shot, in a
fairly straightforward manner, sequences to show the nature of the trip. I
filmed the trio on the ferry, with shots of the sea, the musicians sleeping in
the car, arriving to the city, going for lunch, seeing the sights of the
capital, and going to a music shop to buy new skins for their congas.
We arrived at the party early, and while the musicians were tightening
their conga skins, I filmed people mingling, the party food, the bar, and the
view of the city from the apartment. When the band was ready to start I
positioned myself so as to be able to capture the entire band, and to pan
easily to the audience. Since I had filmed the band before, I had an idea of
what to expect from the musicians and from the audience. As the
musicians took their places, I put the viewfinder up to my eyes and held
the camera with my two hands, pressing my elbows against my torso to
keep the camera as steady as I could. As a result, the viewfinder lay firmly
against my eye, ensuring that I would only see through the viewfinder.
Audio-Visual Evidence and Anthropological Knowledge 71

The group started with their usual salsa number. Wiki, the band’s lead
member, counted off, and Rubén, his brother, set the speed on the cowbell.
Jorge played the base of the number on three congas, a 1-2-3-3-2
syncopation that emulated a bass guitar playing the standard salsa line.
Wiki followed softly on the highest pitched conga, improvising a gentle
flutter of beats above the bass line. They started their salsa number very
slowly, with Wiki and Jorge swaying side to side. I had seen this before,
but never before had I been in a position to watch it so intently. I
straightened my shoulders and bent my knees, putting my weight on my
thighs so I would be able to pan more smoothly with the camera.
The 1-2-3-3-2 bass line continued slowly in the dark, smoky room. The
repetitive beat, while tiring at first, slowly enticed me and drew me into
what they were doing, and I slowly began to feel as if I were caught under
some kind of spell. I panned down to the conga nearest to me and stayed
on it for a few seconds, perhaps a minute. The disembodied conga was
also swaying with the beat; its cream coloured skin was stained with beer,
sweat and blood—from an accident where Jorge had cut his hand a few
weeks before. I panned up again to the slow rhythm of the congas,
focussing on the other two musicians. I then slid as smoothly as I could
towards them, staying on them for another lapse of time. As I moved back,
returning to a wider shot of the three players, the cowbell accelerated and
Wiki closed his eyes. By the time I slid back to my initial position, Jorge’s
bass line had accelerated and shifted to another mood, and Wiki’s fingers
were now fluttering faster over his conga, occasionally beating the entire
face of the drum with the palm of his hand.
The audience was swaying to the faster beat but still not dancing—a
normal occurrence since it usually takes a few songs before the musicians
and the audience get warmed up. As I panned back along the faces of the
onlookers, the host of the party pulled a woman from the crowd and
started dancing with her, encouraging other people to do the same.
Moments later, a second couple emerged from the crowd and joined them.
When I panned back to the drummers, they had transformed their rhythm
once again. I could vaguely make out Jorge's 1-2-3-3-2 beat as he moved
his arms along his three congas at a frantic speed. Wiki now opened his
eyes and his hands were a blur to my camera’s shutter speed, which had to
be kept open because of the light conditions. Without knowing it, I had
moved from my original position and found myself behind one of the
congas. I was filming the backs of the musicians as the audience in the
background became more and more enthusiastic about the drumming. I
recorded one continuous shot for the length of the song, around ten
72 Carlo A. Cubero

minutes. During the course of the evening I filmed two more songs with
similar intent.
I knew something special had happened during the filming of the gig,
but I could not explain it in my fieldwork diary. I felt that I had learned
something about the band’s way of playing and the relationship they had
with each other, but I could not quite articulate it. I attempted to address
this revelation when it came to writing about the theme of music on the
island of Culebra in my dissertation. Like the other chapters, my writing
on music describes the process by which a sense of island uniqueness is
constituted in a context of high mobility and travel. I wrote about the
mobile histories behind the music groups of Culebra, focusing on their
choice of instruments and musical genres. However, my examination of
the Wiki Sound video material revealed more than just the historical and
geopolitical dimensions of music-making in Culebra. My close
observations of the musicians at work, repeated through the process of
editing the video, revealed something about the way in which the
musicians relate to each other when they play, their relationship with their
audience, and their composition process, which are informed by their
After multiple drafts and efforts to find the correct words to express
what I had seen as the musicians’ creative process, I reached a
compromise and wrote about “structure” and “improvisation” as categories
that not only suggest the creative sensibilities of The Wiki Sound, but also
reflect the social practices of musicians in Culebra more generally. I
expanded on this idea by showing how social relations amongst musicians
on the island can be described as simultaneously improvised and
structured. Improvisation refers not only to a compositional style, but also
a social category that addresses the dynamics of a highly mobile society.
Structure refers to the regionalisation of Caribbean music and the
concomitant insularisation, as well as the compositional style of the
island’s music.
Like my analysis of local ideas about the island’s landscape, my
consideration of island music was provoked by a state of mind brought
about through the process of making a video. But unlike the previous
example, my ideas emerged from the actual experience of filming and
editing the video. I suggest that the experience of recording and editing a
specific event revealed more to me than what appeared on the footage
itself. My attempts to describe this process of filming and editing
generated categories that I included in my dissertation, adding a new
descriptive and analytical layer to its argument.
Audio-Visual Evidence and Anthropological Knowledge 73

My intention in this chapter has been to address two disparate
approaches towards the use of audiovisual media in anthropological
research, and to offer an example of how they operate simultaneously in
practice. Audiovisuals, I argue, provide this simultaneity because of their
capacity to offer multiple meanings and narratives without altering their
form. I wish to suggest that the generation of audiovisual materials in
anthropological research constitutes a gathering of evidence and a
production of knowledge simultaneously. In this way, the distinction
between the process of generating knowledge and knowledge itself is not
necessarily clear-cut.
Ethnographic films, like cinema more generally, generate meaning
through both their content and form. Meanings are also created through
the viewer’s reflexive stance towards the characters and events depicted.
It requires from the audience “a form of engagement closer to the
experience of an onlooker at the event than to a reader of an ethnographic
monograph” (Taylor 1996, 77). The viewers are thus empowered to
create meaning in the film by appreciating, for example, the emotive
features of social life, the quality of the relationships within the film, and
the pace of the film. Ethnographic films and videos do not necessarily
state their message explicitly; they instead allow the themes and issues to
emerge through the edited narrative and suggestive scenes in such a way
that viewers learn through acquaintance rather than through lecture.
In my own practice, I attempted to follow conventions of observational
cinema because I was attracted to its potential for sharing with my
audiences the process of discovery. I especially strived to adhere to the
convention of keeping as little distance between the video equipment and
myself, as if to suggest that the footage would represent my own
subjective position in relation to my film subjects and fieldsite. The rushes
represent my shifting interests, my quest for narrative, and my emerging
relationships with the people I filmed. They mark my progressions during
fieldwork, which can be characterised as a tension between appreciating
the aesthetic, the textual, the aural, and the “being” of social events,
against gaining an understanding of the meaning behind the utterances,
social processes, and relationships with which I was engaging.
The two filming contexts I have described illustrate how making a
video enriched my ethnographic research. In the first example, knowledge
was produced through my sustained attention to the island’s landscape, an
attention that I may not have maintained had I not been engaged in making
a narrativised video of my fieldwork experience. In the second example,
74 Carlo A. Cubero

the actual filming and editing was an experience that I could not fully
explain in words. My attempt to do so resulted in the inclusion of
interpretive categories that would not otherwise have emerged in my
ethnographic writing.
The tension between the “being” (appreciating events and processes as
they occur) and “meaning” (understanding the social and political
significance) of social events manifests itself more acutely in audiovisuals
than in text, which points to a general tension in anthropology
(MacDougall 2006). Embracing this tension might pose a problem for an
anthropology that seeks consistency and certainty associated with
academic writing. However, it can serve very well an anthropology whose
goal is to represent social worlds as processual, discursive, and metaphoric
(see Geertz 2000). I believe that the multivalency of audiovisual media
challenges the researcher to confront these tensions in ways that amplify
the reflexive aspects of research. More importantly, audiovisuals promise
to contribute to broadening the perspectives of anthropological discourse
through a renewed examination of anthropological concepts by non-verbal
means, and by provoking new ways of assessing traditional
anthropological ideas.

This paper benefited from remarks and suggestions from Prof. Peter
Wade of the University of Manchester and Mrs. Anne Schiltz, an
independent ethnographic filmmaker based in Luxembourg. I would also
like to thank the editors of this volume for their painstaking work in
proofreading and helping me to clarify my ideas. I am deeply in debt with
the people in Culebra who have offered me their trust and friendship. A
friendship that has gone beyond the research process.

The island of Culebra is one of two offshore municipalities of Puerto Rico. Since
1898 Puerto Rico has been an Overseas Territory of the United States. In 1952
Puerto Rico and the US Congress approved a Constitution for Puerto Rico, which
designated internal island matters to be administered by a locally-elected
government, but also stipulated that the US Congress retain control over external
matters and sovereignty over Puerto Rico.
I recorded a total of sixty hours of footage during my twelve months of fieldwork.
Audio-Visual Evidence and Anthropological Knowledge 75

El Guindi, Fadwa. 2000. From pictorializing to visual anthropology. In
Handbook of methods in cultural anthropology, ed. H. Russell
Bernard. Walnut Cree: Altamira Press.
Geertz, Clifford. 2000. Local knowledge: Further essays in interpretive
anthropology, 2nd ed. New York: Basic Books.
Grimshaw, Anna. 2001. The ethnographer’s eye: Ways of seeing in
anthropology. Cambridge: Cambridge University Press.
— and Ravetz, Amanda. 2005. Introduction to Visualizing anthropology:
Experimenting with image based ethnography, ed. Anna Grimshaw
and Amanda Ravetz. Bristol: Intellect Ltd.
Hart, Keith. 1998. Anthropology and psychology: The legacy of the Torres
Straits expedition, 1898-1998. Lecture given at the opening session of
the Anthropology and psychology: The legacy of the Torres Strait
expedition, 1898-1998 conference, 10-12 August, in Cambridge, UK.
Hastrup, Kirsten. 1992. Anthropological visions: Some notes on visual and
textual authority. In Film as ethnography, ed. Peter Crawford and
David Turton. Manchester: Manchester University Press.
Loizos, Peter. 1992. Admissible evidence? Film in anthropology. In Film
as ethnography, ed. Peter Crawford and David Turton. Manchester:
Manchester University Press.
—. 1993. Innovations in ethnographic film: From innocence to self-
consciousness, 1955-1985. Manchester: Manchester University Press.
MacDougall, David. 2006. The corporeal image: Film, ethnography, and
the senses. Princeton: Princeton University Press.
Mead, Margaret. 1962. Retrospects and prospects. Anthropology and
human behavior. Washington DC: The Anthropological Society of
—. 1975. Visual anthropology in a discipline of words. In Principles of
visual anthropology, ed. Paul Hockings. Berlin: Mouton de Gruyter.
Pink, Sarah. 2006. The future of visual anthropology: Engaging the
senses. New York: Routledge.
Rouch, Jean. 1975. The Camera and the man. In Principles of visual
anthropology, ed. Paul Hockings. Berlin: Mouton de Gruyter.
Taylor, Lucien. 1996. Iconophobia. Transitions 69:64–88.
Vertov, Dziga. 1963. Kinoki Perevorot. Cahiers du Cinéma 144: 32-34.
Winston, Brian. 1995. Claiming the real: The Griersonian documentary
revisited. London: BFI Publishing.


The chief difference between European and American savages lies in the
fact that many tribes of the latter have been eaten by their enemies, while
the former know how to make better use of their conquered enemies than
to dine off them; they know better how to use them to increase the number
of their subjects and thus the quantity of instruments for even more
extensive wars.
—Immanuel Kant, Perpetual Peace.

In January 2006, a feature film about indigenous Huaorani people in
Amazonian Ecuador opened in more than a thousand cinemas across the
United States. End of the Spear was based on the history of spear-killing
raids for which Huaorani people became famous in the second half of the
twentieth century. It gives particular attention to an event in 1956 in which
five North American Protestant missionaries were killed while attempting
to make what was assumed to be “first contact” with the Huaorani. The
release of the film coincided with the fiftieth anniversary of this event,
which became one of the most widely publicised cases of martyrdom in
Christian missionary lore in the United States and Europe. The film largely
follows an evangelical narrative of how an uncontrollably violent cycle of
native revenge-killing was put to an end by the efforts of American
Christians who sacrificed their lives so that the “word of God” would
reach Ecuador’s remaining “savage” Indians. Various aspects of the
narrative I refer to here have been reproduced in dozens of books written
since 1960, some of which were authored by relatives of the dead
missionaries.1 While the film in some ways follows the evangelical
narrative reproduced in missionary texts about the Huaorani, I suggest in
this chapter that it goes further than previous historical sources in
End of the Spear: Re-imagining Amazonian Anthropology and History 77

highlighting the relationship between missionary discourses of Christian

martyrdom and Huaorani notions of victimhood and predation.
When the film premiered, I was in the final stages of writing my
doctoral dissertation about the meanings and uses of past violence in the
Huaorani communities where I carried out my ethnographic fieldwork. My
initial response to the film was one of irritation upon seeing actors playing
the roles of some of my closest Huaorani friends and informants according
to a Christian-inspired Hollywood script. My second reaction was one of
self-conscious apprehension regarding my own anthropological project.
What alarmed me most was the correlation between the film’s title and
that of my dissertation: From enemies to affines: Conflict and community
among the Huaorani of Amazonian Ecuador (High 2006). I wondered if I
was not simply reproducing an American Christian story in the guise of
anthropological research. However, upon further reflection I began to look
at how the film’s evangelical narrative might be related to the oral
histories told to me by Huaorani people, particularly those concerning
their experiences of missionary settlement in the 1960s and 1970s. I have
since come to see the narrative promoted in the film and in scores of
missionary books written about the Huaorani as a form of evidence which
can shed new light on ethnographic and historical research. I suggest that
the intersections between missionary and Huaorani ideas about violence,
victimhood, and history illustrate how anthropologists can fruitfully use as
evidence not only documentary historical sources, but also transcultural
and non-textual sources, such as popular cinema.
My aim in this chapter is not to compare the missionary narrative
developed in the film to the Huaorani stories I was told in order to
establish a more accurate or definitive “history” of Huaorani violence. My
intention is rather to explore the film narrative through the lens of
Huaorani ethnography with the aim of illuminating the shared and
contrasting uses to which missionaries, filmmakers, and Huaorani people
put depictions of past violence. I suggest that although the filmmakers and
my Huaorani hosts use narratives of violence for different purposes, both
strongly engage the notion of victimhood that has been described as lying
at the core of Huaorani cosmology and ethnic identity (Rival 2000, 2002).
I examine how indigenous ideas about predation and victimhood described
in previous Huaorani ethnography are related to the trope of martyrdom
that has become prominent in Christian representations of the Huaorani
since the 1950s. I suggest that this comparison provides new insights into
how Huaorani people have historically reacted to missionaries as well as
possible reasons for why my informants talked about past violence in the
ways they did.
78 Casey High

By examining the place of Christian imagery in Huaorani war

narratives, I suggest that a combination of ethnographic research and
closer attention to missionary discourses allows us to rethink one of the
dominant approaches in Amazonian anthropology. Contemporary Huaorani
ideas about violence and victimhood are not merely versions of a general
“Amazonian ontology”, but equally a historical product of relations
between indigenous people and missionaries. More broadly, I use this case
as an example of how anthropological knowledge can be produced through
the combination of fieldwork and consideration of less conventional
sources of evidence, such as missionary narratives and popular cinema—
even when the latter appear to be temporally and spatially distant from the
words and practices of our informants. I suggest, moreover, that expanding
the range of phenomena that we might treat as anthropological evidence
allows us to rethink the analytical models that have taken centre-stage in
regional studies. In relating Huaorani narratives of violence to both
missionary discourses and broader sociocosmological models developed in
Amazonian ethnography, this chapter illuminates the ways in which the
same objects of ethnographic enquiry can be employed as evidence for
multiple analytical purposes.

Amazonian anthropology: in search of El Dorado

Since the 1970s, anthropologists have repeatedly attempted to establish
a pan-Amazonian theory uniting the practices and cosmologies of
hundreds of indigenous groups living in several different countries, spread
across millions of square kilometres. Focusing primarily on small-scale
indigenous groups in remote areas, early ethnographies were particularly
successful in demonstrating the uniqueness of Amazonian people through
rich descriptions of egalitarianism, autonomy, and often atomised
household groups.2 Amazonian ethnographies thus presented a contrast to
the formal hierarchies and lineages described in Africa and elsewhere
(Riviere 1993). Alongside this emphasis on Amazonia as a unique
ethnographic area, some of the best-known anthropologists working in the
region continue to posit a general social structure, aesthetic, or ontology to
be found throughout Amazonia (Descola 1992; Overing 1983; Overing
and Passes 2000; Viveiros de Castro 2001).
In some ways this is a continuation of Lévi-Strauss’s famous claim that
all Amerindian people in North and South America share variations of the
same mythology (1970, 1973, 1978, 1981). However, the explosion of
Amazonian ethnography, with a few notable exceptions,3 has strangely
coincided with its relative isolation within mainstream anthropology. For
End of the Spear: Re-imagining Amazonian Anthropology and History 79

example, despite the important historical trading, ritual, and marriage links
between Amazonia and the Andes,4 Amazonia is often represented as a
world at odds with the scale and complexity of Andean society
(Heckenberger 2005). Even as most social and cultural anthropologists
have today abandoned outmoded notions of environmental determinism, it
is as if we still assume that a cultural area like Amazonia can somehow be
adequately circumscribed based on the presence of tropical rainforest.
While ethnographers have been successful in establishing indigenous
Amazonian people as the ethnographic “other”, they tend to neglect issues
and concepts that would situate Amazonian ethnographies within wider
historical, political, and economic processes in the region. For example,
given the pace of economic development in the past thirty years that has
placed many indigenous Amazonian groups and their lands under
increasing pressures, it is striking that so many anthropologists continue to
study violence and alterity as expressions of a universal Amazonian
ontology rather than, say, responses to structural violence. This is possibly
because many such works are based on insights generated from fieldwork
in remote places with small and relatively autonomous indigenous groups.
Such insights have thus become the evidentiary basis for the generation of
broader encompassing models of Amazonian sociality and cosmology.
This search for structural continuity and the authentically “indigenous”
partly explains why so little attention has been given to the large and
diverse groups of non-indigenous people living in the Amazon (Nugent
and Harris 2004). Regional studies clearly privilege certain types of
evidence over others, as remote, small-scale societies have been implicitly
taken as a more representative or authentic picture of indigenous
This picture of Amazonia has in recent decades been challenged by
ethno-historical research, which shows how warfare and the relative
isolation of some indigenous groups can be directly traced to economic
changes associated with colonialism, state formation, and missionaries
(Ferguson 1992, 1995; Whitehead 1992, 1993). Historical and
archaeological research reveals a marked contrast between the small and
relatively isolated societies described in much twentieth century
Amazonian ethnography, and the large-scale settlements and regional
networks that existed in the region prior to colonialism (Heckenberger
2005; Lathrap 1970; Roosevelt 1994). While such approaches have forced
us to revise our assumptions about Amazonian history, in this chapter I
want to explore ethnographically how we might rethink the notion of
Amazonian ontology as part of a wider historical process that extends well
80 Casey High

beyond the communities where we conduct fieldwork—and indeed beyond

“Amazonia” itself.
Attempts to establish a pan-Amazonian logic or sociocosmology have
ranged from describing an Amazonian aesthetic in which conviviality and
shared consumption establish consanguinity and the basis of sociality
(Overing and Passes 2000) to a “symbolic economy of alterity” to be
found in relations of exchange affinity, and warfare throughout the region
(Viveiros de Castro 1996, 190). Central to the latter approach is the notion
that ontological predation or symbolic cannibalism is the key
“sociocosmological operator” throughout Amazonia. These two analytical
approaches are not mutually exclusive (Santos-Granero 2000), and indeed
are unified in representing Amazonia within a single encompassing
framework, even if one focuses on peaceful household relations and the
other on violence and alterity.
My intention in this chapter is not to dismiss these approaches or other
claims for pan-Amazonian ideas and practices—which have led to
interesting comparisons and debates within regional ethnography. Rather, I
suggest that a historical and transcultural perspective allows and indeed
forces us to rethink some of the dominant theoretical claims of Amazonian
anthropology. In my view, what is at stake is not the discovery of a truly
indigenous ontology, but an understanding of indigenous Amazonian
people as historical agents whose ideas and practices both inform and are
transformed by events (Gow 2001). I suggest that by insisting that what
we hear or observe in the field is quintessentially “Amazonian”, we fail to
see how so-called Amazonian ontologies are produced and transformed
within wider historical contexts that include non-indigenous people. In this
chapter I am concerned with how a decidedly non-indigenous source,
namely a Hollywood film, allows us to re-imagine how Huaorani ideas
about violence have been generated within historical processes extending
well beyond Amazonia. By examining the intersections and disjunctures
between Huaorani oral narrative and American missionary discourses, I
suggest that such a historical perspective helps to pull Amazonian research
out of the comfortable and self-isolating search for an authentic “El
Dorado” of Amazonianness. In turn, this transcultural approach to
Amazonian history and anthropology raises questions about what
constitutes evidence in anthropology and the evidentiary uses to which we
put less conventional sources, such as film and missionary narratives, in
the production of anthropological knowledge.
End of the Spear: Re-imagining Amazonian Anthropology and History 81

Huaorani violence: from Palm Beach to Hollywood

The spearing of the five missionaries upon which the film End of the
Spear is based, often referred to as the “Palm Beach” killings, made the
Huaorani a powerful symbol within the evangelical world. In the 1950s
there were no more than about five hundred Huaorani living in eastern
Ecuador,5 many of whom were involved in an intense cycle of inter-
household revenge-killing (Yost 1981). They were also well-known in the
region for their ongoing conflicts with other indigenous groups and oil
companies, whose work camps were often attacked or looted while
conducting oil exploration on Huaorani lands.6 As a result, until very
recently the Huaorani were commonly referred to by Ecuadorians as
aucas, a term meaning “wild” or “enemy” in Quichua, Ecuador’s dominant
indigenous language.7
This reputation for violence and self-isolation made them an important
target of North American missionary work in Ecuador, and by the 1950s
missionaries were dropping gifts from planes that they flew over remote
Huaorani longhouses. In January 1956, five missionaries landed a small
airplane on an exposed bank of the Curaray River in an attempt to make
contact with and eventually proselytize a nearby Huaorani group. Only
days later their bodies were found in the river after being speared; a story
that soon reached the American public via an article with photographs in
Life magazine (“Go Ye and Preach the Gospel”, 1956). As the killings
received increasing international attention, the campaign to establish the
“Auca mission” intensified. In the years following the Palm Beach attack,
a widow and a sister of two of the missionary “martyrs” entered the
Huaorani territory and eventually established the first mission in the area
in the 1960s. They did this with the help of a young Huaorani woman who
had surfaced on a plantation in the 1940s and taught one of the
missionaries the basics of Huaorani language.8 By the 1970s the mission
had attracted the vast majority of Huaorani to live in a single settlement,
where many ceased spear-killing and eventually converted to Christianity
(Robarchek and Robarchek 1996; Stoll 1982; Yost 1981). Despite the
decades of missionary presence that continued until the missionary
organisation9 was banned from Ecuador in 1981,10 however, few Huaorani
today identify themselves as Christians.
End of the Spear primarily concerns a series of spear-killings leading
up to and including the deaths of the missionaries at Palm Beach. The key
protagonist in the story is Mincayani,11 a Huaorani man who participated
in the Palm Beach killings before later renouncing spear-killing and
converting to Christianity. The others are Nathan Saint, the pilot who died
82 Casey High

with the four other missionaries in 1956, and his son, Steve Saint, who
returns to live with the Huaorani years later to discover that Mincayani
was his father’s killer.12 The film paints a picture, consistent with much
missionary literature, of an extremely violent society unable to escape a
pattern of revenge-killing until the Christian martyrs and their surviving
relatives teach them to forgive one another and follow the bible. This
transformation is achieved in two ways: firstly, we see the missionaries
choosing not to fire their guns when attacked, a sacrifice that is made clear
by Nathan Saint’s dying words in which he tells his killer (Mincayani)
that he is his friend. Secondly, in the film the Palm Beach killers are
clearly inspired upon seeing that instead of taking revenge, the relatives of
the dead missionaries seek to live among them. The missionary narrative
becomes less subtle as the film progresses, culminating with Mincayani
taking Steve Saint to the Palm Beach site in the 1990s to describe how he
killed his father decades before. As Mincayani narrates the story, the film
replays the fatal spearing and shows a bright light forming above Nathan
Saint as he lies on the beach with a spear in his body; a clear statement of
the martyr’s ascendance to heaven. The film concludes with Mincayani
insisting that his victim’s son spear him, with Steve Saint refusing and
thus forgiving his father’s killer.
The film, produced by Every Tribe Entertainment, was shot in Panama,
casting most of the Huaorani characters with local indigenous people.
Though it claims to be based on a true story, the film is clearly not meant
to be a documentary and is shot within the conventions of a Hollywood
production. Despite high hopes, the film’s box-office potential was
apparently diminished when Christian writers discovered that the lead
actor—cast as Nathan Saint, the heaven-ascending Christian martyr—was
an openly homosexual man who publicly advocated gay marriage on the
Larry King Live Show only days before the film’s release. The box-office
fate of End of the Spear reveals that, despite being a feature film viewed
by millions of Americans, it depended in the end on attracting a Christian
audience. This became evident in the weeks following its opening, as the
official End of the Spear website posted numerous personal testimonies
offered by American Christians telling of how the film’s story had
strengthened their faith or even inspired them to become missionaries.

History, violence, and sociality

While the missionary narrative represented in the film has been
influential in the United States and Europe for decades, the story it tells is
also an important part of Huaorani oral history. In particular, the killings
End of the Spear: Re-imagining Amazonian Anthropology and History 83

which Huaorani people describe as plaguing previous generations play a

key role in contemporary social life, as individual men and women place
themselves morally in relation to various audiences through their
narratives of past violence. I have argued elsewhere (High 2006) that by
narrating these stories from the victim’s perspective, the storytellers are
able to define their commitment to peaceful sociality within a wider
contemporary community that includes former enemies.13
It should be no surprise that much Huaorani storytelling today
concerns past violence. Anthropological research in the 1970s estimated
that inter-group violence was responsible for nearly seventy percent of
Huaorani deaths in the decades preceding missionary settlement (Yost
1981). Their internal revenge killings and conflicts with other indigenous
groups and oil companies during much of the twentieth century led
anthropologists to describe the Huaorani as a “culture of violence”
(Robarchek and Robarchek 1998). While it is easy to treat this as an
extreme case of the “tribal warfare or “blood feuds” described by Chagnon
(1968) and others, the violence that has attracted so much attention from
missionaries, anthropologists, and the Ecuadorian public should not be
essentialised as a purely “indigenous” phenomenon. For example, despite
the relative isolation of groups like the Huaorani prior to the 1960s, the
demographic collapse wrought by colonial penetration into the Northwest
Amazon is likely to have had a major impact on violent conflicts within
and between indigenous groups.14
In the wake of increasing attention to history in Amazonian
anthropology, Viveiros de Castro (1996) has cautioned against the
tendency to attribute seemingly negative aspects of the people we study
(such as violence) to colonial or Western influences. However, as in the
case of the Yanomami described by Ferguson (1995), Western economic
expansion into the Amazon likely had an important role in indigenous
warfare. For example, the infamous rubber boom at the end of the
nineteenth century, in which rubber barons searched for slaves in and
around what is today the Huaorani territory, almost certainly influenced
patterns of Huaorani revenge-killing and relative isolation (Cabodevilla
1999; Muratorio 1991). While this helps to explain Huaorani conflicts
with other Ecuadorians in the twentieth century, more recent changes in
the regional political economy also affected patterns of internal revenge
killings between Huaorani household groups. As plantations and oil
development encroached on Huaorani lands, for example, it appears that
internal conflicts intensified as competition for manufactured steel goods
increased (Cipolletti 2002).
84 Casey High

Past killings are an important part of Huaorani oral history, as parents

and elders often tell impassioned stories of how their relatives were killed
by enemies in the past. These stories, which are often told in the evening
while families sit together around the cooking fire, often evoke even the
most remarkable details of past spear-killing raids. Imagery of violence
and victimhood in these stories appears to constitute a view of the past that
is central to Huaorani conceptualisations of a wider ethnic identity. Laura
Rival (2002) has suggested that the “victim’s point of view” so often
expressed by Huaorani people is based on a shared identity as “prey” to
aggressive outsiders. A clear example of this identity as prey is the widely
accepted assertion that until relatively recently the Huaorani assumed all
non-Huaorani people were cannibals intent on capturing them (Rival 2002;
Robarchek and Robarchek 1998; Yost 1981). Whether or not this
historical picture of the Huaorani is accurate, they are an interesting case
within regional ethnography, particularly in relation to Viveiros de
Castro’s suggestion that notions of predation are part of a universal
Amazonian cosmology (2001). Whereas in many Amazonian cases society
and persons are produced through the necessary incorporation of outside
“prey”, the Huaorani invert this model by placing themselves
unambiguously as the victims of outsider aggression (Rival 2002). Rather
than envisioning killing as a form of productive reciprocity between
enemy groups, or as a form of incorporating the enemy, they position
themselves firmly as victims, thus strengthening the moral opposition
between self and other. Still, this identity as “prey” remains an example of
the emphasis placed on symbolic cannibalism in Amazonian anthropology.
In addition to the pervasive image of victimhood in local oral histories,
my Huaorani hosts present themselves as victims in a number of other
contexts, ranging from unfair treatment by thieving neighbours and greedy
oil companies to more serious claims about suffering witchcraft attacks by
shamans from other villages. Although I know of no single word in the
Huaorani language that can be easily translated as “victimhood”, the
notion of unjust suffering and being wronged by various sources external
to the household is an important and very noticeable feature of Huaorani
personhood and moral identity. I suggest that in narrating past killings
critically from the victim’s perspective, individuals are able to achieve a
degree of moral legitimacy within Huaorani sociality. Narratives of
violence thus become a practice by which people express their
commitment to the idealized conditions for peaceful conviviality in
contemporary villages.
End of the Spear: Re-imagining Amazonian Anthropology and History 85

Victims and martyrs: contrasting and converging histories

What interests me in this chapter is the relationship between the central
idiom of victimhood in Huaorani oral histories and the Christian narrative
of martyrdom advanced in End of the Spear and in missionary texts. To
some extent, Huaorani people tell a story similar to that of the film,
emphasising the intensity and seemingly chaotic nature of past spear-
killings. Since the villages in which I carried out much of my fieldwork
are near the site of the 1956 killings, I encountered many of the killers’
relatives who were able to recount stories about the attack. In most
respects, accounts of the Palm Beach killings I recorded during fieldwork
are consistent with that of the film. There is, however, one key difference
which I find particularly revealing, given the centrality of victimhood in
Huaorani cosmology. While the missionary narrative emphasises that the
American martyrs chose to die themselves rather than fire their guns on
their attackers, many of my informants—including those who had lived
with missionaries for decades—suggested that a Huaorani man was fatally
wounded in the head by a gunshot during the struggle. Although they
criticised the actions of the killers, it is striking that even this story is
ultimately cast in terms of Huaorani victimhood. Another telling aspect of
this narrative is that, even for the few remaining Huaorani evangelical
Christians living in the former mission settlements, Palm Beach is a far
less common story than the frequent accounts of how their kin were killed
by other Huaorani in the past. It appears that these stories of internal raids
have greater salience for Huaorani people insofar as they present an
unambiguous image of the self as victim.
Interestingly, this insistence on victimhood also surfaced in an official
denouncement of the film published on the internet by the Huaorani
political organisation (ONHAE)15 soon after the opening of the film:

as the legal representative body of the Huaorani Nation we give to you our
formal denouncement against those who have taken advantage of our
innocence and produced a film by the name of ‘END OF SPEARS’ [sic],
the same one which speaks of our Huaorani history and that was filmed in
Panama by actors who were looking to imitate us.16

Without assuming this statement to be representative of wider Huaorani

opinion, it certainly points to the ethics and politics of representation
relevant to both filmmaking and anthropology. It is also an example of the
prevalence of “innocent” victimhood that emerges in Huaorani violence
86 Casey High

Despite the organisation’s rejection of the history depicted in the film,

the story of Christian martyrdom introduced by the missionaries through
biblical teachings and translations in the 1960s resonates strongly in
Huaorani communities today, surfacing in many of their own victimhood
narratives. While in the film relatives of the martyred missionaries teach
Huaorani killers to forgive, subsequent missionaries were accepted at least
partly as a result of their association as close kin of the Palm Beach
victims (see Rival 2002). That is, seeing that the outsiders were not
cannibals intent on attacking them, many Huaorani identified the
missionaries as members of a victim group much like themselves, thus
creating potential social relations between themselves and the
missionaries. I suggest that it is this relationship based on local
conceptualisations of common victimhood, rather than the Palm Beach
attack itself, that has made particular missionaries an important part of
Huaorani history. This is because, in the eyes of many Huaorani people in
the former mission settlements, the missionaries shared the same history of
losing close kin to violence.
A number of anthropologists working in diverse settings have noted
the processes by which shared experiences of violence transform social
identities in its aftermath (Bowman 2001; Malkki 1995; Swedenburg
1995). For the Huaorani families with whom I lived, victimhood continues
to be a key marker of identity and potential sociality. For example, in the
wake of a major incidence of violence that occurred within the Huaorani
territory in 2003,17 many of my informants subsequently responded by
identifying themselves with the victim group despite the lack of certainty
about their origins and native language. In this case they posited potential
ties of kinship with the victim group and expressed a desire to establish
relations with them upon hearing of the attack. Whether this emphasis on
victimhood in Huaorani moral identities is rooted in an indigenous model
of Amazonian ontological predation or in notions of Christian martyrdom
introduced by missionaries, or both, it is clear that victimhood has been a
key marker of sociality since at least the 1960s.
Despite the relatively small number of Huaorani Christians today and
the general absence of explicit biblical references in everyday
conversation, my informants often evoked a clear image of Jesus-like
martyrdom in their stories about how their own kin became victims of
violence. In describing to me how his uncle was killed long ago, the
following quote from one of my closest friends and informants provides a
brief example of how missionary discourses still emerge within Huaorani
historical narrative today:
End of the Spear: Re-imagining Amazonian Anthropology and History 87

My mother heard the men planning to kill Toña, and she went to warn him.
Early in the morning, they took Toña to a hill to kill him. They told Toña
to chop down a tree to make spears. While he was chopping, the other men
speared him in the back. With spears in his body, Toña said that he had no
problems with the men who had speared him, forgiving them, saying that
he was a real Huaorani person. He said “don’t kill” and “I die so that you
should no longer kill”. Then they speared him more and Toña cried.18

It is clear from this brief quote that missionary theology, and

particularly ideas about compassionate self-sacrifice emphasised in the
recent film, continue to have a significant place within Huaorani ideas
about violence and victimhood. On the one hand we can speculate that the
common emphasis on victimhood between the missionaries and the
Huaorani after the 1956 attack facilitated a relationship at the mission that
continues to the present. On the other, quotes like the one above raise
questions about the extent to which Huaorani notions of victimhood and
predation have been transformed or inspired by missionary ideology. What
I am suggesting here is that missionary and Huaorani perspectives on
violence may not be as independent of each other as one might assume.
Given that missionaries lived in Huaorani communities and taught them
the bible for decades, this should not be completely surprising, even
though most Huaorani today have abandoned Christianity.
To consider whether to explain Huaorani oral histories in terms of
missionary history or a generic indigenous model of Amazonian cosmology
points to the very nature of what “evidence” is within ethnographic
research. As Hastrup suggests, to invoke the concept of evidence posits a
relationship rather than an independent substance “external to the context
of the situation” (2004, 455). Evidence and knowledge clearly do not exist
without a purpose or a question. To offer missionary images and
discourses as evidence of historical processes posits an actual relationship
between missionary and indigenous ideas. To offer Huaorani narratives as
an example of Amazonian “ontological predation”, on the other hand, is to
posit an analogy between a particular (Huaorani) case and a wider
(Amazonian) encompassing model. In this sense, Huaorani notions of
predation can serve multiple evidentiary purposes in anthropological
research, whether as evidence of an actual relationship or as evidence for
comparison and analogy.
However, the presence of Christian missionary symbolism in war
narratives does not eliminate the potential for making analogies between
Huaorani notions of victimhood and broader socio-cosmological approaches
to Amazonian violence. In fact, analogies provide a useful way of relating
diverse phenomena, as long as we stop short of presupposing a common
88 Casey High

field against which to make ethnographic comparisons (Strathern, this

volume). That is, drawing analogies between Huaorani notions of
victimhood and broader anthropological models of Amazonian ontology is
a valuable tool for ethnographic comparison, as long as Huaorani
perspectives are not encompassed within these models as examples or
variations of a single, unified cosmology. To suggest that indigenous ideas
or practices are expressions of a common Amazonian cosmology risks
disregarding the diverse historical processes that underlie what we hear
and observe in fieldwork. More specifically, such approaches may
overlook the ways in which local “cosmologies” are connected to broader
social and political processes that extend beyond “indigenous society”.
It would be equally suspect, however, to suggest that Huaorani notions
of victimhood and sociality are an example of Amazonian people simply
adopting Christian ideology in the wake of mission settlement. While
more than two decades of missionary presence have transformed many
aspects of Huaorani social life, notions of violence and victimhood are as
much a part of indigenous conceptualisations of personhood and cultural
identity as they are the property of Christian missionaries. To examine
Huaorani and missionary narratives in terms of “acculturation”,
“modernity”, or other similarly encompassing frameworks risks ignoring
the agency and creative imagination which indigenous Amazonian people
themselves bring to historical changes. Rather than treating all indigenous
Amazonian expressions of violence and predation as illustrative cases of a
common ontology, comparing Christian notions martyrdom to Huaorani
narratives of past violence offers an opportunity to relate Huaorani ideas to
both missionary theology and other indigenous Amazonian ethnography.
What is at issue here is not whether Huaorani notions of victimhood
are “authentic” expressions of a generic indigenous Amazonian cosmology.
As with all practices and ideas described by anthropologists, “indigenous”
conceptualisations of violence are produced within diverse and changing
sociocultural environments that include multiple ethno-linguistic,
religious, and other groupings extending well beyond the temporal and
spatial boundaries of indigenous communities. What is at stake is where
anthropologists locate ethnographic knowledge and use it for different
evidentiary purposes. In the present case I suggest that popular cinema and
missionary narratives can be usefully employed as part of anthropology’s
arsenal of evidentiary conventions, as evidence for rethinking both a
particular historical relationship and, more generally, large-scale
anthropological models of Amazonian sociality. Such a move pushes us to
critically examine the encompassing models with which we work and
End of the Spear: Re-imagining Amazonian Anthropology and History 89

allows us to place our research within new and expanding analytical

In this chapter I have contrasted Huaorani oral history and the
missionary narrative exemplified by the film in terms of a few details and
the uses to which they are put before radically different audiences. Such a
comparison, however, also reveals a historical convergence of missionary
and Huaorani discourses. I am not suggesting that the Huaorani
identification as paradigmatic victims was simply produced through
biblical teachings on the mission in the 1960s. However, this case does
reveal how indigenous Amazonian ideas are often part of a history that
extends far beyond relations between indigenous people. It is the
ideological project of Christian martyrdom promoted by missionaries,
their various books, and the recent film, that I suggest provide evidence of
this historical intersection when placed alongside Huaorani oral histories.
Regardless of the claims we make about what ideas and practices are or
are not “indigenous” to Amazonian people, it is clear that the missionary
narrative expressed in the film has become an important part of the lived
experiences of Huaorani people.

Viveiros de Castro’s (1992) suggestion that predation has a central
place in Amazonian cosmologies has been particularly influential within
the regional literature. His comparison of symbolic cannibalism in
contemporary cosmology with historical cases of sixteenth century
cannibalism is highly relevant to many case studies in Amazonia,
including, I suggest, the Huaorani. However, what strikes me about the
continuities between the history and ethnography he presents is a seeming
lack of “history” in the sense of how and why notions of predation might
have changed over time in the context of colonial relations.19 Perhaps a
project of this sort would be overly optimistic given the nearly five
centuries between his two objects of study. However, in the case I have
described, missionary texts and popular cinema provide a significant form
of evidence to begin situating Amazonian ideas about violence in a
historical framework by examining the ideological projects of missionaries
who lived in Huaorani villages between the 1960s and 1980s.
On the one hand, we should recognise the ways in which Huaorani
violence and relative self-isolation during much of the twentieth century
emerged within the political and economic history of the Northwest
Amazon. On the other, we can consider, as I have in this chapter, how
indigenous ideas about violence, victimhood, and alterity more generally
90 Casey High

are created and transformed historically. This case suggests that we should
look to wider translocal relations and processes in examining the stories
told by the people with whom we work, even when such stories seem to
conform to typically “native” ideas or cosmologies within regional studies.
I am not suggesting that anthropologists should abandon their attempts to
define ideas and practices shared across particular ethnographic areas.
However, instead of insisting that diverse ideas and practices are somehow
versions of the same basic phenomenon, we should recognise that the
indigenous cosmologies and practices we describe today are inevitably
part of a process of constant innovation and historical transformation.
To ignore this wider historical context risks exoticising Amazonian
people as “people without history” (Wolf 1982), rather than recognising
them as historical agents in their own right. By insisting on the
“Amazonianness” of what we describe, we ultimately risk repeating the
mistakes of explorers, tourists, and much of South America’s non-
indigenous population: rendering Amazonian people as relics of
prehistory, and in so doing, reaffirming the opposition between “us” and
“them” at the very point where a historical understanding of the
relationship between “our” history and “theirs” might tell us a great deal
about contemporary social life.
This chapter also reveals that attention to missionary discourses may
lead to a better understanding of how “indigenous” ideas are adapted to
historical changes, even if these discourses appear in a Hollywood feature
film steeped in American evangelical symbolism. Students of film,
literature, and anthropology have long recognised the allegorical value of
fiction writing and film as responses to and evidence of the social contexts
in which they are produced. Some anthropologists have suggested that, in
order to come to grips with studying the politics and moral dilemmas of
the world today, anthropologists will benefit from looking to art and
literature, which may reveal increasingly translocal and intercultural
processes in a more productive way than the traditional ethnographic
monograph (Hart 2005). My aim in this chapter has not been to suggest
that the ethnographic and historical evidence gleaned from fieldwork and
the archives is obsolete and should be replaced by reviews of films and
novels. Rather, I suggest that less obviously “local” sources, such as End
of the Spear, can become valuable forms of evidence that re-contextualise
and deepen ethnographic knowledge within a wider intercultural space and
End of the Spear: Re-imagining Amazonian Anthropology and History 91

This chapter is based on a paper originally presented for the
“Ethnographies of Violence” panel at the 2006 meeting of the American
Anthropological Association. I would like to thank Neil Whitehead for
inviting me to write a paper about the film for the panel. Versions of the
present chapter were also presented at the “Questions of Evidence”
conference at Cambridge University and at a research seminar in the
anthropology department at Goldsmiths College in 2007. In addition to the
useful comments provided by the conference discussants, I would also like
to thank Stephen Nugent, Jason Sumich, and Mette High for commenting
on previous drafts of the chapter. My co-editors, Timm Lau and Liana
Chua, provided valuable critiques and suggestions which the present
chapter has only begun to answer. Above all, I am grateful to the men and
women of various Huaorani communities whose kind hospitality and
remarkable craft in oral histories have made this project possible. The field
research upon which this paper is based was carried out between 2002 and
2004 and was funded by the Wenner-Gren Foundation, a Fulbright
Scholarship, and a Central Research Fund travel grant from the University
of London. I would like to thank these institutions, as well as the
anthropology department at the London School of Economics, for their
crucial support at various stages of the project.

The following are just a few examples of the Christian literature on the Huaorani:
Elliot 1957, 1961; Kingsland 1980; Liefield 1990; Saint 2005; and Wallis 1960,
See for example the review articles by Jackson 1975; Overing 1981; and Viveiros
de Castro 1996.
Most notably, Viveiros de Castro’s (1998) work on perspectivism follows Lévi-
Strauss in positing structural continuities between Amerindian societies throughout
the Americas. This work has recently been adopted by anthropologists working in
other parts of the world (Pedersen 2007; Willerslev 2004).
See Oberem 1974; Salomon 1981, 1986; Taylor 1988; and Uzendoski 2004.
Today there are around two thousand Huaorani living in more than thirty
See Cabodevilla 1999; Robarchek and Robarchek 1998; and Wallis 1960 for
descriptions of the early conflicts between Huaorani groups and oil companies.
The Quichua word auca has also been incorporated into popular Ecuadorian
Spanish usage with similar connotations of “savage” or “wild”.
92 Casey High

The missionaries were accompanied by Quichua-speaking indigenous people
from neighbouring areas.
Summer Institute of Linguistics (SIL) and Wycliffe Bible Translators.
One of the original missionaries, the sister of the mission pilot killed in 1956,
remained in the Huaorani territory until her death in 1994 despite the official
closing of the mission.
The film character “Mincayani” appears to combine the actions associated with a
number of Huaorani individuals, including one of the well-known elders who took
part in the Palm Beach killings and still lives along the Curaray River.
Steve Saint continues to be involved in Huaorani communities today. He was a
consultant for the film and appears in an interview at the end of the film. He has
also authored a book by the same title, End of the Spear (Saint 2005).
Such practices are particularly important as larger villages bring together groups
and individuals with a history of mutual avoidance and conflict (High 2006, 2007;
Rival 1992, 1996).
Balee 1995 and Denevan 1976, among others, have examined how the
indigenous population of Amazonia was drastically reduced during the colonial
Organisation of Huaorani Nationalities of Amazonian Ecuador.
This statement, which appeared on the “People of the Path” website, is signed by
a well-known Huaorani ONHAE representative and political activist. (See
In May 2003, a group of Huaorani men attacked a distant and relatively isolated
longhouse group, resulting in approximately twenty-five deaths. The incident
received considerable media attention in Ecuador given the number of deaths,
disagreements regarding the identity and origins of the victim group, and
allegations that the killers were paid and supplied weapons to carry out the attack
by illegal loggers operating in the area. See Cabodevilla, Smith, and Rivas 2004,
and High 2006.
This quote is taken from a much longer narrative I recorded in 2003.
I do not suggest here that Amazonian “history” should necessarily be defined in
relation to colonialism or European history (see Gow 2001). Rather, I refer to
“history” in the sense of how ideas and practices change over time.

Balée, William. 1995. Historical ecology of Amazonia. In Indigenous
peoples and the future of Amazonia: An ecological anthropology of an
endangered world, ed. Leslie Sponsel. Tuscon: University of Arizona
Bowman, Glenn. 2001. The Violence in identity. In Anthropology of
violence and conflict, ed. Bettina Schmidt and Ingo Schroder. London:
End of the Spear: Re-imagining Amazonian Anthropology and History 93

Cabodevilla, Miguel A. 1999. Los Huaorani en la historia del Oriente.

Quito: CICAME.
Cabodevilla, Miguel A., Randy Smith, and Alex Rivas. 2004. Tiempos de
guerra: Waorani contra Taromenane. Quito: Abya-Yala.
Chagnon, Napoleon. 1968. Yanomamö: The fierce people. New York:
Holt, Rinehart, and Winston.
Cipolletti, Maria S. 2002. El testimonio de Joaquina Grefa, una Cautiva
Quichua entre los Huaorani (Ecuador, 1945). Journal de la Societé des
Américanistes 88:111-135.
Denevan, William. 1976. The aboriginal population of Amazonia. In The
native population of the Americas in 1492, ed. William M. Denevan.
Madison: University of Wisconsin Press.
Descola, Philippe. 1992. Societies of nature and the nature of society. In
Conceptualizing society, ed. Adam Kuper. London: Routledge.
Elliot, Elizabeth. 1957. Through the gates of splendor. New York: Harper
and Brothers.
—. 1961. The savage my kinsman. New York: Harper and Brothers.
Ferguson, R. Brian. 1995. Yanomami warfare: A political history. Santa
Fe: School of American Research, University of Washington Press.
—. 1992. A savage encounter: Western contact and the Yanomami war
complex. In War in the tribal zone: Expanding states and indigenous
warfare, ed. R. Brian Ferguson and Neil Whitehead. Santa Fe: School
of American Research Press.
— and Neil L. Whitehead, eds. 1992. War in the tribal zone: Expanding
states and indigenous Warfare. Santa Fe: School of American
Research Press.
“Go Ye and Preach the Gospel”: Five do and die. 1956. Life. January 30.
Gow, Peter. 2001. An Amazonian myth and its history. Oxford: Oxford
University Press.
Hart, Keith. 2005. The hit man’s dilemma: On business, personal and
impersonal. Prickly Paradigm #18. Chicago: Prickly Paradigm Press.
Hastrup, Kirsten. 2004. Getting it right: Knowledge and evidence in
anthropology. Anthropological Theory 4 (4): 455-472.
Heckenberger, Michael. 2005. The ecology of power: Culture, place, and
personhood in the Southern Amazon, A.D. 1000-2000. New York:
High, Casey. 2006. From enemies to affines: Conflict and community
among the Huaorani of Amazonian Ecuador. PhD thesis, Department
of Anthropology. London School of Economics, University of London.
94 Casey High

—. 2007. Oil development, indigenous organisations, and the politics of

egalitarianism. Cambridge Anthropology 26 (2): 34-46.
Kingsland, Ruth. 1980. A saint among savages. London: Collins.
Lathrap, Donald. 1970. The Upper Amazon: Ancient peoples and places.
London: Thames and Hudson.
Lévi-Strauss, Claude. 1970. The raw and the cooked. London: Jonathan
—. 1973. From honey to ashes. London: Jonathan Cape.
—. 1978. The origin of table manners. London: Jonathan Cape.
—. 1981. The naked man. London: Jonathan Cape.
Liefeld, Olive. 1990. Unfolding destinies: The untold story of Peter
Fleming and the Auca mission. Grand Rapids, MI: Harper Collins.
Malkki, Lisa. 1995. Purity and exile: Violence, memory, and national
cosmology among Hutu refugees in Tanzania. Chicago: University of
Chicago Press.
Muratorio, Blanca. 1991. The life and times of grandfather Alonzo:
Culture and history in the Upper Amazon. New Brunswick: Rutgers
University Press.
Nugent, Steve and Mark Harris, eds. 2004. Some other Amazonians:
Perspectives on modern Amazonia. London: Institute for the Study of
the Americas.
Oberem, Udo. 1974. Trade and trade goods in the Ecuadorian Montaña. In
Native South Americans, ed. Patricia Lyon. Boston: Little and Brown.
Overing, Joanna. 1981. Amazonian anthropology. Journal of Latin
American Studies 13 (1): 151-164.
—. 1983. Elementary structures of reciprocity: A comparative note on
Guianese, Central Brazilian and North-West Amazon socio-political
thought. Antropologica 59-62:331-348.
— and Alan Passes, eds. 2000. Conviviality and the opening up of
Amazonian anthropology. In The anthropology of love and anger: The
aesthetics of conviviality in native Amazonia, ed. Joanna Overing and
Alan Passes. London: Routledge.
Pedersen, Morten A. 2007. Talismans of thought: Shamanist ontologies
and extended cognition in Northern Mongolia. In Thinking through
things: Theorising artefacts ethnographically, ed. Amiria Henare,
Martin Holbraad, and Sari Wastell. London and New York: Routledge.
Rival, Laura. 1992. Social transformations and the impact of formal
schooling on the Huaorani of Amazonian Ecuador. PhD thesis,
Department of Anthropology. London School of Economics,
University of London.
End of the Spear: Re-imagining Amazonian Anthropology and History 95

—. 1996. Hijos del sol, padres del jaguar: Los Huaorani de ayer y hoy.
Quito: Colleción Biblioteca Abya-Yala.
—. 2000. Marginality with a difference, or how the Huaorani preserve
their sharing relations and naturalize outside powers. In Hunters and
gatherers in the modern world: Conflict, resistance, and self-
determination, ed. Peter Schweitzer, Megan Biesele, and Robert
Hitchcock. New York: Berghan Books.
—. 2002. Trekking through history: The Huaorani of Amazonian Ecuador.
New York: Columbia University Press.
Rivière, Peter. 1993. The Amerindianisation of descent and affinity.
L’Homme 126-128:507-516.
Robarchek, Clayton and Carole Robarchek. 1998. Waorani: The contexts
of violence and war. Fort Worth: Harcourt Brace.
—. 1996. The Aucas, the cannibals, and the missionaries: From warfare to
peacefulness among the Waorani. In A natural history of peace, ed.
Alvin Wolfe and Honggang Yang. Nashville: Vanderbilt University
Roosevelt, Anna. 1994. Amazonian anthropology: Strategy for a new
synthesis. In Amazonian Indians: From prehistory to present. Tucson:
University of Arizona Press.
Saint, Steve. 2005. End of the spear: A true story. Carol Stream, Illinois:
Tyndale House Publishers.
Salomon, Frank. 1981. Killing the Yumbo: A ritual drama in northern
Quito. In Cultural transformations and ethnicity in modern Ecuador,
ed. Norman Whitten. Urbana, Illinois: University of Illinois Press.
—. 1986. Native lords of Quito in the age of the Incas: The political
economy of North-Andean chiefdoms. Cambridge: Cambridge
University Press.
Santos-Granero, Fernando. 2000. The Sisyphus syndrome, or the struggle
for conviviality in native Amazonia. In The anthropology of love and
anger: The aesthetics of conviviality in native Amazonia, ed. Joanna
Overing and Alan Passes. London: Routledge.
Stoll, David. 1982. Fishers of men or founders of empire? The Wycliffe
Bible translators in Latin America. London: Zed Books.
Swedenburg, Ted. 1995. Memories of revolt: The 1936-1939 rebellion and
the Palestinian national past. Minneapolis: University of Minnesota
Taylor, Anne-Christine. 1988. Al Este de los Andes: Relaciones entre las
sociedades amazónicas y andinas entre los siglos XV y XVII, Vol. 2.
Quito: Abya-Yala.
96 Casey High

Uzendoski, Michael A. 2004. The horizontal archipelago: The

Quijos/Upper Napo regional system. Ethnohistory 51 (2): 317-357.
Viveiros de Castro, Eduardo. 1992. From the enemy’s point of view:
Humanity and divinity in an Amazonian society. Chicago and London:
University of Chicago Press.
—. 1996. Images of nature and society in Amazonian ethnology. Annual
Reviews of Anthropology 25:179-200.
—. 1998. Cosmological deixis and Amerindian perspectivism. Journal of
the Royal Anthropological Institute, n.s., 4:469-488.
—. 2001. GUT feelings about Amazonia: Potential affinity and the
construction of sociality. In Beyond the visible and the material: The
Amerindianization of society in the work of Peter Rivière, ed. Laura
Rival and Neil Whitehead. Oxford: Oxford University Press.
Wallis, Ethel. 1960. The Dayuma story: Life under Auca spears. New
York: Harper and Brothers.
—. 1973. Aucas downriver. New York: Harper and Row.
Whitehead, Neil L. 1992. Tribes make states and states make tribes:
Warfare and the creation of colonial tribes and states in Northeast
South America. In War in the tribal zone: Expanding states and
indigenous warfare, ed. Barbara Ferguson and Neil L. Whitehead.
Santa Fe: School of American Research Press.
—. 1993. Historical discontinuity and ethnic transformation in Native
Amazonia and Guayana. L’Homme. 28:289-309.
Willerslev, Rane. 2004. Not animal not not-Animal: Hunting, imitation
and empathetic Knowledge among the Siberian Yukaghirs. Journal of
the Royal Anthropological Institute, n.s., 10:629-652.
Wolf, Eric. 1982. Europe and the people without history. Berkeley:
University of California Press.
Yost, James. 1981. Twenty years of contact: The mechanisms of change in
Wao (Auca) culture. In Cultural transformations and ethnicity in
modern Ecuador, ed. Norman Whitten. Urbana: University of Illinois


Introduction: Empirical value

Biomedicine tends to split afflictions of the mind from those of the body.
However, various ethnographic data can show this assumption to be
rudimentary and incorrect ... In failing to surmount mind/body dualism, the
concept of somatisation is no longer useful.
—Second-year medical anthropology essay, 2005.1

Supervising undergraduates offers critical purchase on the topic of

anthropological knowledge production. Though none of the Medical
Anthropology lectures attended by my students directly addressed the
evidentiary character of ethnography, during our discussions the value of
anthropological insight emerged as a topic of some concern. Students
wanted to know what makes one analytical position better than another;
and further, what good comes from debunking dominant discourses. Of
course, this epistemological unease is endemic to the social sciences, and
indeed, to any intellectual exercise that draws upon theoretical models to
illuminate empirical material. But for those studying medical anthropology
the problem is amplified. The tentative authority of ethnographic
representation—balanced between scientific method and literary craft—is
compounded by the uncertain utility of its claims. Whether or not an
anthropologist’s observations of medical practices are “correct” sits
awkwardly with the question of whether or not they are “relevant” to the
well-being of those involved. Here, generating evidence—in the sense of
information that gives a reason for believing something to be true—
implicitly presupposes an audience (doctors, patients, researchers, and
policy makers) for whom those claims will make a difference.
Take, for example, an average undergraduate medical anthropology
paper. Because students are asked to include Western science within the
boundaries of their investigations, they learn to regard positivism as a
historically contingent and culturally specific style of reasoning (Good
98 Ann Kelly

1994; Martin 1991). Medical diagnoses are thus exposed as dualistic,

reductionist, and objectifying. Proceed down the page and these same
measures are celebrated as vehicles of patient empowerment; read further,
and they are dismissed as politically naïve. For instance, one student
demonstrated how, in addition to influencing scientific practice and policy,
patient-activist groups also provide a resource for the pharmaceutical
industry. She concluded that clinically-defined identities are not only
intellectually suspect, but also socially unjust, because they are shaped by
commercial logics. Our discussion of her work revealed that she was not
altogether happy with this critique. For what was the impetus behind her
interpretative decisions? Should her analyses be guided by the critical
criteria of anthropological theory or the interests of patient welfare? When
faced with situations of suffering and illness, she was at a loss as to what
anthropological investigations of biomedicine are meant to accomplish.
Conventionally, medical anthropologists reconcile the pragmatic and
theoretical dimensions of their work by focusing their ethnographic
attention on the practices, experiences and understandings that medical
knowledge excludes. Illustrative of biomedicine in context, ethnographic
data reveals the contingent and site-specific relations invisible to the
universalising and purifying gaze of medical science. The impact of
anthropological evidence on personal and collective well-being has,
therefore, been understood in terms of the social and cultural dimensions it
adds to clinical research and practice (Hemmings 2005; Pool and Geissler
2005). The corroborative potential of this perspective has been particularly
evident in the treatment of mental illness, for which patient narratives have
a direct impact on clinical diagnosis (e.g. Kleinman 1995; Young 1995); in
the formation of public health policy insofar as local perceptions about
health care are believed to inform health-seeking behaviours (e.g. Hahn
1999; Helman 1994); in the implementation of new medical technologies,
where user views are regarded as integral to innovation and development
(e.g. Lock et al. 2001; Rapp 1999); and finally, in the analysis of global
inequalities, where finely-grained ethnographic accounts of individuals’
struggles to gain therapeutic access are used to reveal and challenge
corporate agendas in the production and distribution of pharmaceuticals
(e.g. Biehl 2004; Petryna, Lakoff, and Kleineman 2006).
However, whether or not biomedicine is capable of accommodating
anthropological research findings is not what is at stake in this paper.2 Nor,
for that matter, is the degree to which ethnography can or should be
instrumentalised by disciplines other than anthropology. The subject of the
following discussion is, instead, the evidentiary conventions through
which anthropologists attempt to enhance the social consequence of their
Pragmatic Evidence and the Politics of Everyday Practice 99

investigations. I have chosen to explore those efforts in a medical context

because of the particular questions anthropologists face about the purpose
of their interventions. My sense of that epistemological unease has
ethnographic as well as pedagogic origins. Conducting fieldwork on the
production of medical evidence has made me particularly sensitive to the
instability of the ethnographic position vis-à-vis its object. On more than
one occasion, my ethnographic engagement with a medical study involved
interviewing and observing anthropologists working as a part of a clinical
research team. My initial response to these uncanny encounters was to feel
woefully inadequate. It seemed evident to me that from their position
within clinical research, these anthropologists’ day-to-day interactions
with health practitioners, medical researchers, and clinical trial volunteers
would generate rich empirical detail. Their findings would serve not only
to refine clinical accounts of therapeutic efficacy and thereby influence
health care policy, but also, from an anthropological perspective, to
generate profound insights about the creation of medical knowledge. In
contrast, my observations of an assortment of funding meetings, trial-
steering workshops, data and ethics monitoring discussions, patient-
activist meetings, and research methodology seminars seemed to lack the
ethnographic integrity associated with what George Marcus describes as
“being within a landscape” (1998, 99). The phenomena on which I was
reporting belonged to a different order than routine experimental research
procedure and its practical effects. Unsubstantiated by material reality, my
arguments about the construction of medical knowledge and public benefit
seemed speculative at best.
Though it took a great deal of time to think through my experiences
productively, it was ultimately this sense of falling short that inspired my
contribution to this volume. The key analytical step was to refrain from
contextualising medical research, and instead, to regard clinical trials and
ethnography as analogous inductive strategies. Recent ethnographic
investigations of legal and bureaucratic practices have distinguished
anthropological knowledge from other ways of knowing in terms of its
orientation toward means/ends reasoning (Riles 2004). To some extent,
my discussion follows a similar analytical tack. Drawing from my own
empirical material, I will describe a clinical trial investigating the
effectiveness of a screening programme for osteoporosis in the UK.
Pragmatic in approach, this trial is designed to reflect the complexity of
“everyday clinical practice”—or the social and practical interactions
attendant to routine health care delivery. Thus, as opposed to a true
measure of the biological effect of an intervention, the generalisability of
pragmatic evidence is intended as a realistic gauge of public health
100 Ann Kelly

impact. Put another way, the inferences drawn from the pragmatic clinical
trial are illustrative rather than explanatory. But accommodating
contextual features of medical care offers empirical advantages beyond
that of representation: my argument is that models of “everyday practice”
embed experimental protocols in the clinic. Thus, the methodological
value of a pragmatic research strategy is in the simultaneity with which it
implements the evidence it produces.
As an analogous take on “everyday practice”, I describe Annemarie
Mol’s ethnography of atherosclerosis, The Body Multiple (2002). Shifting
focus from epistemology—the experiences and understandings of
sickness—to ontology—the artefacts and technicalities that constitute
disease—Mol’s ethnography speaks to current anthropological
preoccupations with the material and practical nature of social meaning
(e.g. Hayden 2003; Henare, Holbraad, and Wastell 2007; Miller 2005;
Rabinow 1996; Riles 2006; Strathern 1999). Whether or not The Body
Multiple is representative of theoretical trends in anthropology is not
relevant to this discussion. What interests me is how a methodological
move from concepts to things challenges anthropological evidentiary
conventions. By re-describing biomedical reality as that which happens in
“everyday practice”, Mol opens up scientific knowledge for ethnographic
inspection and political consideration. As opposed to those concerned with
representation, this is an inductive strategy that claims to participate in the
world it describes. It is the purportedly productive character of
anthropological inferences drawn from material practices that engages my
By juxtaposing these two models of “everyday practice”, I seek to
come to terms with the scientific mileage and social value this empirical
unit affords to both clinical and anthropological research. At this point, it
should be clear that my comparative analysis will not stop at the
identification of difference or argue the potential for closer resemblance.
Ethnography and clinical trials exemplify distinct methodological
traditions3 and address divergent empirical concerns; my discussion is
unlikely to inspire anthropologists to randomise their observations.
However, appreciating how the pragmatic clinical trial constructs a
representative sample—the experimental links between subjects of study
and parameters of improvement—might encourage a more nuanced
consideration of the inductive steps we take from context-specific detail to
consequential statements about the world in which we live. As Strathern
argues in this volume, the critical advantage of analogic reasoning is that it
builds connections between different accounts of the world without
compromising their integrity. Thus, my hope is that drawing an analogy
Pragmatic Evidence and the Politics of Everyday Practice 101

between the pragmatic clinical trial and Mol’s ethnography will enrich our
understanding of the generic nature of anthropological claims. In
particular, I believe that the efforts made by the designers of clinical
experiments to transform evidence of “everyday practice” into evidence
for “everyday practice” may shed light on my anxieties, and those of my
students, about the relationship between anthropological evidence and its
impact in the “real world.”

The pragmatic clinical trial

Making evidence from scientific studies available to clinical practice has
been expected to directly improve quality of care, but this expectation has
not been realized … research is needed that takes into account the specific
characteristics of its population and the presentation and prevalence of
illness and disease. (Maeseneer et al. 2003, 1314)

Though only recently a focus of sociological and anthropological

inquiry, the relative benefits and potential biases of clinical research
evidence have polarised medical debate for the past three decades. At the
centre of this discussion is evidence-based medicine (EBM) (Guyatt et al.
1992), a methodological approach that endeavours to systematise clinical
decision-making through the application of critically appraised evidence of
clinical effectiveness. The key evaluative technique for an EBM model of
medical authority is the double-blinded randomised controlled clinical trial
(RCT). To randomise in a comparative study means that treatment is
allocated according to a chance mechanism, such as the tossing of a coin.
By introducing a fixed margin of error, experimenters can control for
unknown confounding interrelationships, and in so doing, accurately
interpret experimentally defined relationships between variables. In the
context of research on human subjects, random assignment also inhibits
the doctor’s inclination to give the sickest patients treatment rather than
placebos, and thereby eliminates selection bias (Marks 1997, 150ff.). The
advantages randomisation offers to evaluative practice are powerful: a
chance arrangement releases the experimenter from the burden of creating
comparable groups, and, hence, disassociates scientific outcomes from
expert judgment. The autonomy of research ensures not only the accuracy
of its outcome but also the liberal orientation of the information it
produces. As the founding father of EBM, British epidemiologist Archie
Cochrane succinctly put it, “[RCT] is the key to a rational health service”
(1972, 11).
As a strategy built upon the suspension of clinical judgement, EBM
has met with significant resistance from clinicians who regard
102 Ann Kelly

standardisation as an intrusion on their professional autonomy. Based

solely on protocols derived from population data, EBM, it is argued, is a
“cookbook” medicine incapable of responding to the demands of the
individual patient (Berg 1997; Timmermans and Berg 2003). In a
profession with an ever-expanding evidence base, the challenge facing
medicine is often described as closing the gap between medical knowledge
and daily clinical decisions (Bero et al. 1998; Schattner and Fletcher
2003). Sackett’s frequently-cited definition of EBM reflects these
concerns by encompassing both the science and the art of medicine:

Evidence-based medicine is the conscientious, explicit, and judicious use

of the current best evidence in making decisions about the care of
individual patients. (Sackett et al. 1996, 71)

But for some, the discrepancies between the evaluative aims of EBM
and the character of clinical work go beyond application to the ways in
which “current best evidence” is produced. While its core set of
methodological strategies—randomisation, blinding, clinical equipoise,
and informed consent—underpin its claims to universality, RCT processes
work to expunge the unpredictability that defines clinical practice.4 The
highly controlled setting of the RCT is unrepresentative of the realities of
health care, and thus fails to meet the practical needs of clinical decision
makers (Maesneer et al. 2003; Timmermans and Angell 2001). Critics of
EBM take issue with the “artificiality” of current approaches to the
production of evidence; they insist that reliable clinical data must take into
account the vicissitudes of physician care and patient health (Grol and
Grimsaw 2003; Wood, Ferlie, and Fitzgerald 1998). For professionals
concerned with the quality of health care delivery—and not simply the
efficacy of an intervention under ideal conditions—the reliability of
research-generated data depends more on the utility of its findings than on
the stability of its controls (Godwin et al. 2003; Tunis, Stryer, and Clancy
2003; Will 2004).5 In recent years, clinical researchers have begun to
develop hybrid research designs including the pragmatic clinical trial as an
alternative to the RCT (e.g. Kachur et al. 2001; Macpherson 2004).
Though still randomised, pragmatic clinical trials aim to generate practical
information to guide routine clinical practice rather than explanatory data
about the pharmacological or physiological effects of an intervention. This
approach strives to balance methodological rigour with clinical relevance,
generating evidence that will be more applicable to real-world settings.
My understanding of the methodological challenges attendant to
clinical research is drawn from fieldwork conducted with the Arthritis
Research Campaign (ARC), a UK-based charity that sponsors clinical
Pragmatic Evidence and the Politics of Everyday Practice 103

trials in rheumatology. During the January 2004 funding meeting of the

ARC Clinical Trials Collaboration (CTC), an application to conduct a
“pragmatic randomised controlled trial of the effectiveness and cost
effectiveness of screening for osteoporosis in older women in the
prevention of fractures” attracted the committee’s attention. The
application earned high marks from external reviewers for “the
significance of its question” and was selected by the committee as a
particularly strong candidate for ARC support. Though some concerns
were raised about the trial’s design, the ARC agreed to finance the trial as
an eighteen-month feasibility study involving eight hundred female
patients over the age of seventy at two Primary Care Trusts.6 Now
expanded into a seven year project (2007-2014), the SCOOP (SCreening
Of Older women for Prevention of fractures) Trial is set to recruit 11,600
women aged between seventy and eighty-five across seven sites in the
The trial will investigate a screening programme that combines clinical
examination with a DEXA scan, an x-ray technology that determines a
patient’s bone mineral density. Though there is little doubt that the DEXA
scan is the gold standard for diagnostic practice, what concerned SCOOP’s
principal investigator was that the instrument is neither cheap nor readily
available in the UK (Ferguson 2004, 180). The trial, therefore, was
directed toward the problem of distribution: for what group of people, at
what time in their lives, will the assessment of fracture risk by DEXA be
most beneficial and least costly to the National Health Service (NHS)?8 To
answer this question requires, first, an intervention that combines
technological and clinical elements. To that end, the research team plans to
incorporate the scan within a programme that effectively links patient self-
assessment with General Practitioner (GP) diagnosis. Addressing the
impact of such a programme on both patient and institutional outcomes
necessitates a proof that will demonstrate those complex and context-
specific therapeutic relations (Campbell et al. 2000; Will 2007).9 By
accommodating the human unpredictability attendant to the “real clinic”,
the pragmatic clinical trial locates its evidentiary context between the
rigour of an experimental protocol and the specificities of the health care
On the broadest level, the purpose of the trial will be to demonstrate
how an osteoporosis screening programme might work in everyday
practice. The challenge is primarily one of design: how can a trial
simultaneously model a clinical interaction and measure its effects on
patient and clinician behaviour? The research team has addressed this
problem by deploying an intriguing protocol. At the start of the trial, the
104 Ann Kelly

GP will send a letter to female patients aged seventy to eighty-five asking

them if they would like to participate in a study. The letter is accompanied
by a consent form, information leaflets on osteoporosis, and a baseline
information booklet designed to assess a patient’s overall risk of fracture.
Those patients who agree to participate will be randomised into a control
and a screening group. From clinical information provided on participant
self-assessment forms, the GPs will divide the screening group into three
subsets. The first group, considered to be at low risk, will not be contacted
further.10 The second group, found to be at high risk, will be given
appointments to discuss disease management. The third group, whose
level of risk could not be assessed from the self-reported forms, will have
their bone mineral densities measured by a DEXA scanner. If the scanning
determines that a patient is at high risk for a fracture, a course of treatment
will be recommended.
When it was initially presented, the trial raised a number of
methodological problems for the ARC funding committee. With regards to
the patient volunteers, would the use of self-assessment forms and the
distribution of informational leaflets generate an unusually proactive
sample population? Might the information provided about osteoporosis
alter the conduct of the control group? Now more aware of the risks
associated with the disease, would not all those enrolled in the trial be
more likely both to seek treatment and to adopt behaviours that would
reduce their risk of fracture? In short, this highly-informed sample could
hardly be considered representative of the general patient population.
Further, the trial grants GPs a degree of flexibility that would make its
outcomes exceedingly difficult to analyse. Concerned less with whether
the trial sufficiently resembled the “real clinic” and more with the
legibility of scientific proof, GPs might misdiagnose patients, or worse,
recommend a therapeutic regimen, such as Hormone Replacement
Therapy (HRT), that was no longer considered front-line treatment for
While acknowledging the legitimacy of these concerns, the trial’s
principal investigator (PI) maintained that the value of the evidence
generated about the screening programme was directly related to the
impact of the study’s findings on practice. In his response to the funding
committee’s review, the PI defended the trial’s design on two key grounds.
First, he argued that if participation in the trial inspired volunteers to seek
clinical advice about osteoporosis, this alone would demonstrate the
effectiveness of the screening programme. Further, a more proactive
research sample might be more likely to modify poor health habits,
comply with the GP’s treatment advice, and thus generate more conclusive
Pragmatic Evidence and the Politics of Everyday Practice 105

evidence about the value of a policy of treatment that combines clinical

examination with screening. Rather than an obstacle to experimental
rigour, the PI argued that the volunteer bias introduced by the self-
assessment forms and information packet would reinforce the evidentiary
context of the trial by narrowing the distance between trial and practice.
Thus, the trial participants would demonstrate the beneficial effects of the
programme by acting as functional elements of the programme itself.
Their relationship to the patient population would not be merely
epidemiological; the pragmatic clinical trial would secure the
representative nature of the research sample through a series of complex
experimental moves that cultivated and coordinated the technical and
human aspects of clinical practice.
Second, the PI stressed that the “sample” included both patients and
GPs. The GPs are integrated within the trial’s design, not as gatekeepers of
a patient population or as executers of a research project, but as part of the
intervention and its potential beneficiaries. The team framed variations in
GP clinical practice as a problem of information, rather than one of
clinical uncertainty. In the same way that pamphlets would educate
patients about the management of their disease, a sufficiently detailed
protocol would inform GPs about current standards of care, changing their
behaviour to improve their clinical practice.11 The changes initiated by the
experimental context to GPs’ practices and patient behaviours render moot
the accuracy of its depiction of the clinic. What makes the pragmatic trial
convincing as a demonstration of effectiveness is its potential to animate
clinical relations, and thereby generate better performance.
In short, the relationship the pragmatic clinical trial construes between
its model and its object is not simply mimetic (Armstrong 2002): here,
complexity is resembled so that it may be effectively managed (see also
Berg 1997 and Will 2007).12 The criticisms of the osteoporosis screening
trial focused on whether or not it could demonstrate the effectiveness of its
intervention. That the screening programme would work was never
questioned; its benefits to practice inhered in the consistency it promised
to generate. What convinced the committee to fund the trial were the steps
the design took to set this goal in motion. In both informing GPs about
current research in osteoporosis treatment and teaching patients how to
assess their own risk of fracture, the protocol reorganises clinical practice
before the beneficial effects of these processes are evaluated. One could
argue that as a rationalising strategy, the pragmatic clinical trial is
considerably more comprehensive than the RCT. Located within
“everyday practice”, the pragmatic evidentiary context reworks clinical
realities one technicality at a time. Though it is not one I will pursue here,
106 Ann Kelly

the argument could be made that as a form of regulation, which enhances

efficiency through the process of description, the trial is mirrored in the
audit. Equating control with social benefit, these techniques pre-empt
dissent not through “hard fact” but through their managerial grasp of the
“real world”.
Anthropologists might balk at the suggestion that the pragmatic
clinical trial could hold lessons for anthropology, and for good reason. It is
an obvious point that defining the outcomes of anthropological inquiry in
advance of its ethnographic investigations defeats the methodological
purpose of fieldwork. Writing fieldnotes entails anticipation rather than
prediction; anthropologists gather as much as possible with an eye to the
very things they would take for granted, knowing that at some point, they
might prove relevant. However, as a modelling exercise that involves
making decisions about where the boundaries between an inquiry and its
object should be drawn, the pragmatic clinical trial may help shed light on
the aggregative dimensions of anthropological evidence. In the next
section, I examine an ethnographic description of clinical reality. Here
again, empirical work centres on everyday practice, though in this case,
that methodological focus serves to preserve contingency rather than
reorder it. However, for both projects, models of everyday practice serve
to generate links between scientific description and social consequence. In
the case of anthropology, I believe this connection is a cause for concern.
Specifically, I am worried that attending to the vicissitudes of everyday
practice obscures to what and to whom our evidence speaks, and more
broadly, how we imagine the social.

Ethnography of disease
It is possible to refrain from understanding objects as the central points of
focus of different people’s perspectives. It is possible to understand them
instead as things manipulated in practices ... if instead of bracketing the
practices in which objects are handled we foreground them ... Reality
multiples. (Mol 2002, 5)

Though her observations are not evaluative in character, Mol also

seeks to reveal the effects of clinical interventions. The object at the
centre of her ethnography is atherosclerosis—a common arterial disease
associated with a lack of blood supply to the legs. Drawing critical
attention to the materiality of disease, Mol distinguishes her work from
the preoccupations of conventional medical anthropology. Rather than
investigate the different perspectives implied in treating and suffering
from atherosclerosis, Mol describes the surgical technologies, therapeutic
Pragmatic Evidence and the Politics of Everyday Practice 107

interventions, patient complaints, diagnostic techniques, vessels, and

fluids that shape atherosclerosis as a biological, pathological,
epidemiological, and social entity. Like the pragmatic clinical trial, The
Body Multiple aims to grasp medical realities by shifting its empirical
focus from representing disease to exploring how diseases are handled in
everyday practice.
Juxtaposing the notion of practice with the descriptive capacity of
scientific models is a familiar anthropological move. Taking a classic
example, Pierre Bourdieu finds systematic principles incapable of
grasping the productive dynamics of practical endeavour. In his two
seminal works on practice, Bourdieu attempts to concentrate
anthropological inquiry on the everyday constitutive behaviours of the
individual. Rather than attempting to explain or decode practice, he
privileges the forming and reforming of habitus—a term he defines as
“the durably installed generative principle of regulated improvisations”
(1977, 78). The regularity of habitus is of a different order than that of
obedience to rules. According to Bourdieu, the structure of social action
creates a format through which an actor can manipulate practical
situations according to his interests.
Mol’s empirical engagement with atherosclerosis retains Bourdieu’s
emphasis on the generative spontaneity of practical action. She illustrates
how atherosclerosis is constantly reformulated in response to specific
situations—a process which, she argues, is irreducible to a coherent
scientific rationale. Thus, like Bourdieu, Mol disassociates knowledge
from an objective “point of view”. However, Mol’s dogged attention to
physicality entails consequences more radical than the loss of an external
perspective. Here, practice does not describe the behaviour of doctors and
patients; rather, it materialises and animates a disease. Understood as an
interrelated series of events, diseases no longer serve as points of reference
around which to interpret social logics. As I suggested above, for Mol,
attending to practice implies a radical shift in the interpretive register from
epistemology to ontology.
If diseases such as atherosclerosis are not fixed entities but rather
emerge in practice, then it follows that the closer we are to everyday
events, the better we understand the real. Exploring this contingent work
of social and material mediation borrows from the observations made by
science studies, which has revealed the complex social and material
arrangements that underwrite reliable facts (e.g. Latour 1999; Hayden
2003). But mapping out atherosclerosis as a network of activities is only
the starting point for Mol’s analysis. It is not enough, she argues, to
recognise and represent the mess beneath seemingly objective practice:
108 Ann Kelly

What is required is that we find better ways of articulating the tensions and
the questions within treatment practices and pertinent to clinicians and
their patients. And that we design research activities in such a way that
these questions are dealt with more seriously. (Mol 2000, 533, emphasis

Despite the emphasis she seems to place on public engagement with

medical practice, Mol finds the patient-centred rhetoric of contemporary
healthcare policy poorly suited to advance that discussion. By the time
therapeutic action must be taken, several key decisions—which projects
get funding, how research will proceed, to what extent an intervention will
be subsidised by the state—have already taken place. Emphasising the
freedom of a patient’s choice over therapeutic intervention ignores the
material and technical contestations that shape those possibilities.
Jurisdiction over one’s health is fairly limited unless one can exercise
influence over what one has say.
Here, Mol’s argument takes an interesting turn. As a proof of the
contingency of medical knowledge, her ethnography offers evidence
against the scientific authority of the randomised controlled clinical trial
and, by implication, the precepts of evidence-based medicine. Describing
biomedicine—its instructions, interventions, diseases, and traditions—as
part of “common, day-to-day, socio-material practices” (2002, 6), Mol
reveals the myriad social and technical relations that underwrite the
effectiveness of an intervention. According to her model of clinical
practice, the number of steps in a patient’s home can be as constitutive of
atherosclerosis as the thickness of a vessel wall. Because the RCT controls
the interpersonal and contextual relations attendant to a therapeutic
intervention, guidelines based on its evidence weaken rather than reinforce
healthcare. In contrast, ethnographic attention to everyday practice
preserves the possibility of alternate configurations of atherosclerosis, and
thereby allows deliberation over the different ways in which the disease
can and should be handled. “The question that now needs asking,” she
writes, “is ‘what reality should we live with?’” Though her study leaves
this issue open, it is clear that for Mol, clinical practice can only be
advanced by understanding therapeutic value as a matter of detail and
practical relevance. By beginning to explore “the details that involve the
good life” (ibid., 165, 174, emphasis added), Mol believes ethnography
can become a resource for practical and social efficacy. “Like any other
representation this book is a part of practice, or a set of practices.
Attending to the multiplicity of the body and its diseases,” she insists, “is
an act” (ibid., 151-152).
Pragmatic Evidence and the Politics of Everyday Practice 109

I find this line of reasoning jarring. How does Mol move so seamlessly
from an indeterminate, varied reality to improved parameters of clinical
assessment? Whose values inform the “reality we should live with”? How,
indeed, do ethnographic descriptions of the material and practical
negotiations of everyday practice speak for that collective well-being?
More pointedly, what kind of “act” is ethnography? Are its representations
to be considered on par with a surgical intervention?
We might begin to address these questions by tracking how doing
away with perspective—partial or objective—tallies ethnographically.
Though Mol’s field includes research colloquiums, interviews with
specialists, conversations with doctors, and patient narratives, her interest
in medical rhetoric is limited. “This study is not,” she stresses, “about
styles of thought that dominate medical specialties or individual doctors”:

The different worlds may be inhabited by different people, but the people
do not make the difference. Even if all surgeons were to perceive
atherosclerosis as a gradual process of plaque formation tomorrow, they
wouldn’t be able to treat it in this way. (ibid., 112)

The people in her ethnography—patients, radiologists, vascular

surgeons, haematologists and pathologists—surface where they intersect
with x-ray plates, calcified arteries, blood velocity as part of the same
disease-event. In other words, Mol treats “talk” as she would any other
technical apparatus used for medical diagnosis, prognosis, and treatment.
She is specific, then, about the kind of talk that strikes her ethnographic
interest: “If one follows the argument through, it might be better when
interviewing people (whether they be doctors or patients) to ask them
about what they do and about the events that happen to them, rather than
about their thinking” (ibid., 12, emphasis added). What people say about
atherosclerosis is significant only insofar as it sets the stage for what gets
If, anthropologically speaking, what counts is what works, the question
arises: where, in Mol’s model of everyday practice, is the social located?
In The Body Multiple, the answer is everywhere: in instruments,
procedures, buildings, and events. The social, Mol argues, is no bigger
than the technical practices than her ethnography investigates. But how,
then, does Mol imagine society? The analogy she draws is telling:

The organism in hospital Z (and other places like it) has gaps and tension
inside it. It hangs together, but not quite as a whole. It is more than one and
less than many … what we end up with is an organism that clashes and
coheres—just like society. (ibid., 84)
110 Ann Kelly

If its analogue is the organism, then society is merely a composite of

practical connections. Because the collective is located in moments of
technical efficacy, producing evidence of “everyday practice” is a political
act. On the one hand, attending to the instruments, the procedures, and the
complex and mundane interactions of clinical practice opens scientific
truth to public debate. Which people make up that public is a question Mol
does not have to address: society exists in “daily life … where patients,
we, have to live with our doubts and our diseases” (ibid., 184). Thus the
act of ethnography also becomes a conjuring trick; the “we” in empirical
ontology is merely an extension of therapeutic preferences. To my mind,
this position short-circuits a more careful ethnographic consideration of
the ways in which well-being is and can be imagined. In so doing, it
forecloses the analytical potential of the evidence anthropology can

Conclusion: Pragmatic analogies

Analogy is a demonstrative or evidentiary practice—putting the visible
into relationship with the invisible and manifesting the effect of that
momentary unison … All of analogy’s simile-generating figures are thus
incarnational. (Stafford 1999, 105)

Like the pragmatic clinical trial, Mol’s ethnographic evidence is

simultaneously descriptive and productive. As a testimony to the
indeterminacy of medical reality, her accounts of clinical practice contain
their own uses. Investigating “everyday practice” gives her observations a
political gloss. By contrast, the inductive work that “everyday practice”
does for the pragmatic clinical trial suggests a greater purchase on the
politics of collective wellbeing. The social and scientific integrity of
clinical research rests on the representative character of the sample. If
clinical generalisation proceeds by inferring population benefits from the
effects of a therapeutic intervention on a representative group of patients,
the clinical trial must produce likeness between a sample and society. For
though the problems representation entails for politics are distinct from
those it provokes for epistemology, for clinical research, the tasks of
faithful and exact representation coincide (Epstein 2004). As an event (or
aggregation of events) that renders novel relations visible so that practice
might be transformed (Barry 2001, 177-178), the persuasiveness of a
clinical trial inheres in both its technical coherence and in its relevance to
patient populations. The pragmatic clinical trial which accomplishes this
work entails a coordination of indices: the number of fractures suffered,
the cost of a DEXA screening, the social and psychological characteristics
Pragmatic Evidence and the Politics of Everyday Practice 111

of the sample, the malleability of GP clinical expertise. The research team

cannot, in other words, take society for granted, but rather must articulate
it through a complex definition of patient well-being and population
benefit.13 “Everyday practice” provides the language to express that
pragmatic imaginary: it is the algorithm that directs the trial’s passage
from experiment to policy and from epistemological concept to
institutional arrangement.
As anthropologists, we may wish to distance ourselves from the notion
of representative groups. Still, drawing an analogy with the clinical trial
forces us to think critically about the ways in which we group our
observations and the everyday features (as opposed to ethnic or cultural
ones) around which generalisations are made. George Marcus has a point
when he argues that we have left the study of “peoples” behind;
“ethnography”, he suggests, “is predicated upon attention to the everyday,
an intimate knowledge of face-to-face communities and groups” (1998,
82). However, if we consider the conception of the social that inflects our
claims to anthropological knowledge, the problem of aggregation—how
we shift our analytic frames from everyday interactions to the constitution
of community—looms large. In the field of medicine, anthropologists
could start by scrutinising the categories through which they track medical
practice and imagine well-being. Alternatively, analytical attention could
be focused on the notions of society that guide our sense of the
“everyday”. This intellectual activity would have particular relevance to
discussions of the ethics of global health and human rights, where appeals
to civil society, public good, and social benefit are complicated by
transnational processes of governance.14 Beyond facilitating deliberations
on “which reality we should live with”, ethnographic attention to everyday
practice could begin to explore how, where, and when the collective is
evoked. Evidence of the ways in which individual projects are
scientifically or politically connected to collective well-being does more
than add context and contingency to dominant forms of knowledge.
Understanding the tools we use as analogous to other methods of
knowledge production illuminates the contemporary character of scientific
integrity and social value. Rather than asking “how can anthropology be
relevant to medicine?”, my students could begin to consider the ways in
which this relevance is understood.

This essay significantly benefited from the painstaking editorial work
of Liana Chua, Casey High and Timm Lau. The clarity of my argument is
112 Ann Kelly

a product of their conscientious attentions. Thanks are also due to my

father, William Kelly, a true alchemist of sentences. I also am grateful to
the Arthritis Research Campaign for their forbearance of an ethnographer
in their midst, and to Annemarie Mol, for providing the theoretical
impetus for this essay. Finally, I am in debt to Catherine Will, whose
critical insights continue to shape and inspire my thinking on clinical

From an essay entitled: “What does the notion of embodiment bring to an
understanding of mental illness in a cross-cultural context?” It is quoted here with
the student’s permission.
For a convincing argument about how evidence-based medicine accommodates
distinct evidentiary claims by encompassing them, see Lambert 2006.
Although perhaps not entirely: consider Haddon and Rivers’ investigations of the
perceptive sensitivity of natives to life in the Torres Straits. As Schaffer (1994)
illustrates, fieldwork techniques had their origins in the laboratory; the
anthropological interest in cultural difference evolved from efforts to generate
indexes of “normal behaviour” across human subjects. My point in juxtaposing
these two evidentiary practices is not to reveal or reject commonalities to clinical
approaches to investigations of the clinical world, but to inspire new ways of
thinking about anthropological evidentiary conventions.
Randomisation aims to control the confounding differences that an experimenter
cannot measure. In the context of research on human subjects, whereby radical
variability exists across an experimental sample, random assignment eliminates
any bias that might systematically put similar people in one group. In so doing,
randomisation creates a framework through which confounding interrelationships
are dissolved in order that the experimentally defined relationship between
variables can be interpreted.
The empirical trade-off is often phrased in terms of sacrificing “internal
coherence” for “external validity”; while large-scale RCTs with clear-cut
endpoints often produce elegant results, pragmatic evidence is believed to be more
clinically relevant.
Primary Care Trusts are statutory bodies responsible for delivery of health care
and the provision of community health services to a local area. Within the
parameters of the budgets and priorities set by the Strategic Health Care Authority,
they can fund GPs and commission hospital services from NHS trusts or the
private sector.
In January 2005, the ARC CTC awarded £200,000 to conduct a twelve-month
pilot study across two Primary Care Trusts in Norwich and Sheffield. In 2006, this
amount was increased to £380,000, and with the combined support of the
Osteoporosis Society the SCOOP trial has begun enrolment (for more information
see /scooptrial).
Pragmatic Evidence and the Politics of Everyday Practice 113

This information is drawn from a series of interviews and conversations
conducted by the author with the PI, Dr. Lee Shepstone (2004-2005), and from the
initial grant application for the trial (Shepstone, Harvey, and Fordham 2004).
To determine the effectiveness of this intervention, the team had to take into
account the number of osteoporotic patients identified, the subsequent
effectiveness of the treatment, the cost of the program measured in time, money,
and potential side-effects, and the ease with which GPs might integrate the
technique into practice. The expertise necessary for this research required the
formation of a multidisciplinary research team comprising orthopaedic specialists,
rheumatologists, public health specialists, economists, and statisticians.
They will receive what is described by the protocol as “usual care”, which, in
essence, denotes no proactive involvement whatsoever. Because a control patient’s
self-reported risk assessments will not be read, GPs will probably only see a
patient when she fractures a bone (Lee Shepstone, interviewed by the author, 21
May 2005).
For a discussion of how protocols become instruments of implementing trial
results see Berg 1997.
This strategy of contextualisation is quite different from the one Strathern
describes as characteristic of western epistemology. The demand for situatedness
resonates with the clinical trial, but the task of representation here is tied not so
much to generating a more holistic image, but rather, to activating links between
scientific research, medical practice, and scientific policy (1992, 73).
Here, I have drawn inspiration from Alberto Corsín-Jiménez’s analysis of
attempts made in economic theory to measure social well-being. What captures his
anthropological interest is the way in which economic theorists have resisted
reifying society, by asking question about how the social is distributed. In his
phrasing, I would argue that the clinical trial could be viewed as a tool through
which “the social apportions itself” (2005, 12).
James Ferguson has shown how for many low-resource countries, where
sovereignty no longer implies the provision of public services or even control over
national resources, the “national state” has become an increasingly irrelevant
category for conceptualizing political or social life (2006, 207). In its place, “civil
society” has emerged as a rubric better suited to the analysis of the real-world
processes of development and democratisation. However, Ferguson concludes that
emphasizing “civil society” obscures the financial institutions, development
agencies, and medical research collaborations that govern in place of the state.

Armstrong, David. 2002. Clinical autonomy, individual and collective:
The problem of changing doctors’ behaviour. Social Science &
Medicine 55 (10): 1771-1777.
114 Ann Kelly

Barry, Andrew. 2001. Political machines. London and New York: Athlone
Berg, Marc. 1997. Rationalizing medical work: Decision-support
techniques and medical practices. Cambridge, MA and London: The
MIT Press.
Bero, Lisa A., Roberto Grilli, Jeremy M. Grimshaw, Emma J. Harvey,
Andrew D. Oxman, and Mary A. Thomson. 1998. Closing the gap
between research and practice: An overview of systematic reviews of
interventions to promote the implementation of research findings.
British Medical Journal 317 (7156): 465-468.
Biehl, Joao. 2004. The activist state: Global pharmaceuticals, AIDS, and
citizenship in Brazil. Social Text 22 (3): 105-132.
Bourdieu, Pierre. 1977. Outline of a theory of practice, trans. Richard
Nice. Cambridge: Cambridge University Press.
Campbell, Michelle, Ray Fitzpatrick, Andrew Haines, Ann Louise
Kinmonth, Peter Sandercock, David Spiegelhalter, and Peter Tyrer.
2000. Framework for design and evaluation of complex interventions
to improve health. British Medical Journal 321 (7262): 694-696.
Cochrane, Archibald. 1972. Effectiveness and efficiency: Random
reflections on health services. London: Nuffield Provincials Hospitals
Corsín-Jiménez, Alberto. 2005. Well-being and anthropological theory:
an analytical provocation. Manuscript, University of Manchester.
Epstein, Steven. 1996. Impure science: AIDS, activism, and the politics of
knowledge. Berkeley: University of California Press.
—. 2004. Bodily differences and collective identities: The politics of
gender and race in biomedical research. Body and Science 10 (2-3):
Ferguson, James. 2006. Global shadows: Africa in the neo-liberal world
order. Durham and London: Duke University Press.
Ferguson, Niall. 2004. Osteoporosis in focus. London: Pharmaceutical
Godwin, Marshall, Lucia Ruhland, Ian Casson, Susan MacDonald, Dianne
Delva, Richard Birtwhistle, Miu Lam, and Rachelle Seguin. 2003.
Pragmatic controlled clinical trials in primary care: The struggle
between external and internal validity. BMC Medical Research
Methodology 3:28.
Good, Byron J. 1994. Medicine, rationality, and experience. An
anthropological perspective. Cambridge: Cambridge University Press.
Pragmatic Evidence and the Politics of Everyday Practice 115

Grol Richard and Jeremy Grimshaw. 2003. From best evidence to best
practice: Effective implementation of change in patients’ care. The
Lancet 362 (9691): 1225-1230.
Guyatt, Gordon, John Cairns, David Churchill, et al. (Evidence-Based
Medicine Working Group). 1992. Evidence-based medicine. A new
approach to teaching the practice of medicine. Journal of the American
Medical Association. 268 (17): 2420-2425.
Hahn, Robert. 1999. Anthropology in public health: Bridging differences
in culture and society. Oxford: Oxford University Press.
Hayden, Cori. 2003. When nature goes public: The making and unmaking
of bioprospecting in Mexico. Princeton, NJ: Princeton University
Helman, Cecil. G. 1994. Culture, health and illness: An introduction for
health professionals. 3rd ed. London: Butterworth Heinemann.
Hemmings, Colin. P. 2005. Rethinking medical anthropology: How
anthropology is failing medicine. Anthropology and Medicine 12 (2):
Henare, Amiria, Martin Holbraad, and Sari Wastell, eds. 2007. Thinking
through things: Theorising artefacts ethnographically. London and
New York: Routledge.
Kachur, S.Patrick, Salim Abdulla, Karen Barnes, Hassan Mshinda, David
Durrheim, Andrew Kitua, and Peter Bloland. 2001. Letters to the
Editors, Re.: Complex, and large, trials of pragmatic malaria
interventions. Tropical Medicine and International Health 6 (4): 324-
Kleinman, Arthur. 1995. Writing at the margin: Discourse between
anthropology and medicine. Berkeley: University of California Press.
Lambert, Helen. 2006. Accounting for EBM: Notions of evidence in
Medicine. Social Science & Medicine 62 (11): 2633-2645.
Latour, Bruno. 1999. Pandora’s hope. Essays on the reality of science
studies. Cambridge, MA: Harvard University Press.
Lock, Margaret, Allan Young, and Alberto Cambrosio, eds. 2001. Living
and working with new medical technologies: Intersections of inquiry.
Cambridge: Cambridge University Press.
Macpherson, Hugh. 2004. Pragmatic trials. Complementary Therapies in
Medicine 12:136-140.
Marcus, George. E. 1998. Ethnography through thick and thin. Princeton,
NJ: Princeton University Press.
116 Ann Kelly

Marks, Harry M. 1997. The progress of experiment: Science and

therapeutic reform in the United States,1900-1990. Cambridge:
Cambridge University Press.
Martin, Emily. 1991. The egg and the sperm: How science has constructed
a romance based on stereotypical male-female roles. Signs: Journal of
Women in Culture and Society 16 (3): 485-501.
Maeseneer, Jan M. de, Mieke L. van Driel, Larry A. Green, and Chris van
Weel. 2003. The need for research in primary care. The Lancet 362
(9392): 1314-1319.
Miller, Daniel, ed. 2005. Materiality (politics, history, and culture).
Durham and London: Duke University Press.
Mol, Annemarie. 2002. The body multiple: Ontology in medical practice.
Durham and London: Duke University Press.
Petryna, Adriana, Andrew Lakoff, and Arthur Kleineman, eds. 2006.
Global pharmaceuticals: Ethics, markets, practices. Durham and
London: Duke University Press.
Pool, Robert, and Paul Wenzel Geissler. 2005. Medical anthropology.
Maidenhead: Open University Press.
Rabinow, Paul. 1996. Essays on the anthropology of reason. Princeton, NJ:
Princeton University Press.
Rapp, Rayna. 1999 Testing women, testing the fetus: the social impact of
amniocentesis in America. New York: Routledge.
Riles, Annelise. 2004. Property as legal knowledge: Means and ends.
Journal of the Royal Anthropological Institute 10 (4): 775-795.
—. 2006. Documents: Artifacts of modern knowledge. Michigan:
University of Michigan Press.
Sackett, David L., William M.C. Rosenberg, J.A. Muir Gray, R. Brian
Haynes, and W. Scott Richardson. 1996. Evidence based medicine:
What it is and what it isn’t. British Medical Journal 312 (7023): 71-
Schaffer, Simon. 1994. From physics to anthropology—and back again.
Prickly Pear Pamphlet No.3. Cambridge: Prickly Pear Press.
Schattner, Amichai, and Robert H. Fletcher. 2003. Research evidence and
the individual patient. Quarterly Journal of Medicine 96 (1): 1-5.
Shepstone, Lee, Ian Harvey, and Ric Fordham. 2004. Pragmatic
randomised controlled trial of the effectiveness and cost effectiveness
of creening for osteoporosis in older women in the prevention of
fractures. Project grant application, Arthritis Research Council,
Clinical Trial Collaboration.
Pragmatic Evidence and the Politics of Everyday Practice 117

Stafford, Barbara M. 1999. Visual analogy: Consciousness as the art of

connecting. Cambridge, MA: MIT Press.
Strathern, Marilyn. 1999. Property, substance and effect: Anthropological
essays on persons and things. London: Athlone Press.
Timmermans, Stefan, and Alison Angell. 2001. Evidence-based medicine,
clinical uncertainty, and learning to doctor. Journal of Health and
Social Behavior 42 (4): 342-359.
Timmermans, Stefan, and Marc Berg. 2003. The gold standard: an
exploration of evidence-based medicine and standardization in health
care. Philadelphia: Temple University Press.
Tunis, Sean R., Daniel B. Stryer, and Carolyn M. Clancy. 2003. Practical
clinical trials: Increasing the value of clinical research. Journal of the
American Medical Association. 290 (12): 1624-1632.
Will, Catherine. 2007. The alchemy of clinical trials. Biosocieties 2 (1):
Wood, Martin, Ewan Ferlie, and Louise Fitzgerald. 1999. Achieving
clinical behaviour change: A case of becoming indeterminate. Social
Sciences and Medicine 47 (11): 1720-1738.
Young, Allan. 1995. The harmony of illusions: Inventing post-traumatic
stress disorder. Princeton, NJ: Princeton University Press.


In 2004-2005, I spent twelve months doing field research in a coastal
village of 370 people in the Republic of Benin, collecting ethnographic
data on the practice of vodou, the dominant religion in the country. My
original idea was to study the ways devotees of the vodou goddess Mami
Wata experience and learn to understand the world they live in, and
compare these to the ways religion, culture, and “modernities” have been
conceptualised in recent anthropological studies. Worship of Mami Wata,
whose devotees are mostly small-scale entrepreneurs, has increased
dramatically in the past twenty years. The collapse of Benin’s Marxist-
Leninist regime in 1989, the arrival of a capitalist-style market economy,
and the political freedom to practice vodou have all contributed to its
growing popularity. Since Mami Wata, a light-skinned mermaid goddess,
is fond of consumer items associated with a Western lifestyle and beauty,
it is not surprising that existing anthropological research on this
phenomenon has been dominated by arguments that contrast “Western”
modernity with “Beninese” tradition.
In this chapter I attempt to demonstrate that contrary to most of these
arguments, it is misleading to view Mami Wata worship as a convenient
case study of the effects of globalisation on a local level. Using a small-
scale ethnographic study as evidence, I argue that the most crucial
categories used by the worshippers of Mami Wata themselves are not
affiliated with common anthropological conceptions of modernity,
consumerism, and tradition. Instead, they are more concerned with
different types of exchange relations between humans and spirits, and
specific ways of acquiring the ambiguous privilege of engaging in these
relations. Drawing on this case study, this chapter argues that
ethnographically-specific issues and discourses can both nuance and
The Vodou Priest who Lost his Spirit 119

challenge the large-scale explanatory models which have dominated the

study of West African vodou. It concludes by assessing the wider
implications of rendering analytical (Toren 2002, 7) the categories that our
informants use as forms of anthropological evidence.

The world of the vodou spirits

The Beninese people refer jokingly to the eighteen-year period of
Marxism-Leninism (1972-1990) as Le Laxisme-Béninisme. Even though
vodou worship was theoretically forbidden during this time, people still
practised it quite openly without facing any serious consequences. When
Benin became a democracy in 1990, however, the position of vodou was
elevated within society. The Beninese state has since actively promoted
vodou as a part of the nation’s cultural and spiritual heritage. To this end,
in 1992 it partly financed the formation of le Bureau National de Vodou,
the main task of which is to regulate the practice of vodou throughout the
country. It authorises the formation of new temples, solves religious
disputes, and acts as an authority in determining whether a temple can
rightfully claim to have certain types of spirits residing in it. Even though
just over half of the Beninese population are officially vodou worshippers,
it would not be an overestimation to say that in the rural coastal region
where I did my fieldwork, everyone I encountered was engaged in some
form of vodou worship.
Beninese people divide vodou spirits into two different main
categories. First there are the ancestral spirits, Togbe Yéhou, which stay
among the living villagers. In each family compound there is a shrine built
for kuku, the dead ancestors. These spirits are consulted daily through the
shrine and their advice is asked on family problems and other domestic
matters. Old trees planted by ancestors, objects like pipes or clay pots that
belonged to kuku in life, and some animals of significance to the family’s
history are also regarded as Togbe Yéhou.
The second group of spirits is simply called Vodou or Yéhoue. This
group is divided into four spirit pantheons—air, water, fire, and earth—
each containing several hundred spirits that are related to each other. The
leader of the fire pantheon is a thunder god called Heviosso. The main god
of the earth pantheon is the god of smallpox, Sakpata. The main spirit of
the air pantheon is a rainbow serpent that lives in the sky, referred to as
Edan. The mermaid goddess Mami Wata leads the pantheon of water
Whereas Togbe Yéhoue spirits normally stay at family compounds and
are taken care of by ordinary people, the vodou spirits that belong to the
120 Anna-Maija Aguilera Calderòn

pantheons stay in temples specially constructed for them, where they are
cared for by a hounon: a vodou priest or priestess. Hounons have to
undergo an initiation ceremony before they are permitted to run a temple
anywhere in Benin. In order to become a hounon for a certain spirit group,
the aspiring priest will need to acquire a bo, which could be described as a
container or an edifice in which a specific type of spirit resides.
In Benin there are two ways to become an owner of a bo: through
inheritance or purchase. If a hounon has enough money he can also learn
the skill of manufacturing bos, in which case he is entitled to use the title
of botono—the maker of the bos. In the Mina language, the main language
of my fieldsite, an inherited bo is called bo pomési, which literally means
“a bo of the ancestral family”, and a purchased bo is referred to as bo
yovoni, which can be translated as “the bo of the strangers”. A bo yovoni
can become a bo pomési if it is passed on in the family through
inheritance. All the bos in my fieldwork village, apart from the ones for
the spirits of the ancestors, were originally purchased elsewhere.
By acquiring a bo one simultaneously acquires the privilege to engage
in direct exchange relations with vodou spirits, thereby becoming a
mediator between the spirits and the people who seek help from them.
Each of the vodou spirits from the main pantheons specialises in solving
certain types of problems. For example, Sakpata spirits cure people and
Mami Wata spirits help people gain employment or overcome financial
problems. In order to get help from a spirit one needs to give it a vossa
(sacrificial gift), and pay the hounon a fee for his services. All vodou
spirits need to eat and drink daily, and have their own distinct tastes. It is
the task of the hounon to feed the spirits; a vossa always consists of food,
drink, and cigarettes, and in some cases perfume and talcum powder.
The vodou spirits possess differing degrees of wealth. The spirits of the
Mami Wata pantheon are the richest; and a goddess called Mami Sika is
the most affluent among them. This type of Mami Wata spirit is not really
a mermaid or a Hindu-style goddess, like most spirits in this group. Mami
Sika, which literally means “Golden Mother”, could be described as a
seductive light-skinned woman who loves gold and diamond jewellery and
is always accompanied by a white mare. The economic status of each
spirit is also reflected in his or her taste. For example, Sakpata spirits—the
poorest group in the main Pantheon—mainly drink local palm liqueur
(sodabi) and consume cigarettes, palm oil, and chicken blood, whereas
Mami Wata spirits require gin, the blood of peacocks, and boiled sweets,
which are harder to acquire. Of course one can give luxurious drinks and
food to Sakpata if one wishes, but should one give the local palm liqueur
and other cheaper products to Mami Wata, she will not stay in her bo: the
The Vodou Priest who Lost his Spirit 121

spirits are free to leave, and will do so if they are not satisfied. A common
denominator among all the spirits in the vodou pantheon is their
unpredictability and short tempers. Spirits are possessive, ungrateful, and
impulsive: it is impossible for humans to predict their behaviour, and the
owner of a bo yovoni must learn to accept the stressful and financially
disastrous possibility of losing the spirit. Even though bos can be
purchased or inherited, they cannot really be owned; the spirits have a free
will and if they choose to abandon their residence, a bo will become
During my fieldwork, I was surprised at how little the ethnographic
data that was emerging from my research corresponded with the existing
anthropological literature on Mami Wata, which I shall explore further
later. Briefly, it has commonly been argued that the worship of Mami
Wata can primarily be seen as the devotees’ response to the changing
social and economic situation in post-Marxist-Leninist Benin. Indeed this
seems to be borne out by the fact that her shrines are adorned with
distinctly “foreign” and “modern” goods, including expensive beauty
products, Western movie posters, pictures of Hindu goddesses, and broken
mobile phones. At first glance such features appear to be “local” evidence
of the workings of “globalisation” or “modernity” on a small scale.
However, I shall argue in this chapter that this is only true when such
phenomena are identified and assessed through encompassing explanatory
models like “tradition” and “modernity”. Conversely, I suggest that rather
than looking for “local” manifestations of such models, we must also take
into consideration “native” categories and discourses, using them as the
analytical and evidential basis of our ethnography.
In the following section I shall approach these issues through a detailed
ethnographic account of a vodou priest who purchased a foreign Mami
Wata spirit, brought her to his village, and then lost her after failing to
properly care for her. I shall then conclude by examining the relationship
between my ethnographic data on vodou spirits and common
anthropological conceptualisations of “tradition” and “modernity”.

Losing the spirits

Sofo, a thirty-three year-old vodou priest was, or at least became, one
of the least popular characters in the Southern Beninese village where I
conducted my fieldwork. Sofo was the only son of a respected hounon, a
healer and a specialist priest of Sakpata, the earth god. People came to
Sofo’s father’s temple to be healed and to find help for their problems, and
he managed to run it quite successfully, making a relatively good income
122 Anna-Maija Aguilera Calderòn

from it. Sofo’s father had hoped that his son would follow in his footsteps
and take over the management of the temple, becoming at least a hounon if
not a botono. He had taught Sofo from a young age how to feed and serve
the three different Sakpata spirits residing in the shrine room that he had
inherited from his own father. However, when Sofo was only fifteen years
old he left the village against his father’s will. Rumours spread that, like
many young men from rural Benin, he had travelled to Gabon to work on
the fishing boats. Sofo spent sixteen years away from the village of his
birth. All his relatives thought that he had died, since he never contacted
anyone in Benin. After his father realised that his son may never come
back, he started to collect money to initiate his daughter, Sofo’s elder
sister, as the temple’s next hounon.
When I started my fieldwork, Sofo had already been back in the village
for two years, and taken over the management of the temple from his
sister. Before long, I noticed that the villagers gossiped a lot about him and
his time abroad, and that there were many wild stories circulating about
his past. Whatever the truth about Sofo’s wandering years, it was certain
that he had managed to make a considerable amount of money during his
time abroad, by the standards of the village. The villagers told me that he
had returned with his own car, and that once he had settled down in his
father’s temple compound, he had immediately built a new, rather
grandiose room for the shrines of the Sakpata spirits. They also revealed
that upon his return, Sofo had been in quite a bad physical condition,
suffering from severe headaches, palpitations, and insomnia. Sofo claimed
to have developed these problems a couple of years earlier, and was
convinced that they were caused by his father’s spirits disturbing him.
After his return Sofo had consulted several hounons in the area, who came
to the conclusion that he was constantly ill because he had escaped his
predestined duty of becoming a hounon. He had abandoned the inherited
Sakpata spirits and let them slowly decline in his family temple. Making
him ill was how they forced him to return home. Thus in order to regain
balance in his life, he should be initiated as a hounon in the same temple as
his father and grandfather.
Sofo followed the advice of the oracles and priests, and travelled to
Northern Benin to stay in the temple where all the people from his pomé
(ancestral family) were initiated. This temple specialised in initiating
priests who had inherited shrines for specific types of earth spirits, and
teaching some of them to manufacture bos. All the hounons initiated there
belonged to the Guen-Mina “tribe” which had a common Togbe: a
spiritual forefather who, according to the oral legend, had travelled to the
Ivory Coast 174 years ago, where he learned to manufacture shrines for
The Vodou Priest who Lost his Spirit 123

Sakpata spirits. After his initiation Sofo quickly became quite successful,
and aspired to make enough money to become a botono too. His Sakpata
spirits had a good reputation in the village, and since his father had taught
him how to manufacture medicine from leaves and plants, he also claimed
to be able to successfully treat infertile women. Sofo was now an
acknowledged hounon in Benin. He participated in the meetings of le
Bureau National de Vodou and also got involved in local politics.
Sofo did not suffer from a lack of confidence, and constantly told me
that he was going to become the greatest vodou priest in Benin—that his
reputation would be known all around the West African coast. In order to
gain such success, Sofo was convinced that he needed to travel to Ghana
and purchase at least three different types of Mami Wata bos to make his
ahwan, literally “a troop of spirits”, function better. He constantly hinted
that this was going to happen soon and that the villagers would not believe
their eyes when they saw the beautiful shrine, replete with Western
cosmetics and plastic toys, that he would construct for the new spirits.
Time passed, and I did not see Sofo much as he was working as a
treasurer for the le Bureau National de Vodou committee organising the
celebrations for National Vodou Day, which is held annually on 10
January. It was also the time of the presidential elections. As part of his
campaign, one of the candidates had donated the equivalent of £3,000 for
the organisation of the festivities. The presidential candidate lived in
France and did not even attend National Vodou Day. The celebrations
went well, despite many complaints about the quantity of the sacrificial
animals. Normally the whole vodou community would feast on a stew
prepared from three bulls slaughtered on the day. This year, however, the
animal sacrifice was very meagre, and only the highest-ranking priests
were offered meat to eat.
Soon after National Vodou Day, Sofo left the village. His wife, sister,
and several smotos (temple servants) took care of the temple while he was
away. They were very secretive about Sofo’s whereabouts, but since most
villagers knew about his desire to bring new Mami Wata bos to the village,
it was assumed that he had gone to Ghana to fetch them. Sofo returned to
the village in late February. One could see three Ghanaian cars on the road
and Sofo standing on the back of a truck, dressed in a white loincloth and
red beads that were worn only by Mami Wata hounons. In his hand he held
an object covered in a white cloth embroidered with the colours of the
rainbow. It was clear to everyone that Sofo had brought to the village a bo
for Mami Sika, which, as mentioned earlier, is the most affluent spirit of
the Mami Wata pantheon. This kind of spirit did not yet exist in the
village. Overjoyed by this, the people followed the convoy of cars to
124 Anna-Maija Aguilera Calderòn

Sofo’s temple. There Sofo gave his sister money to organise a feast, and
that evening the whole village celebrated the arrival of the new spirit. All
the Mami Wata priestesses and priests from nearby villages arrived at
Sofo’s temple and played drums until dawn. Many people fell into trances,
since they felt the presence of the new spirit so strongly. While people
celebrated, Sofo and the Ghanaian high priestess, or Tassino, set up the
shrine room where the bo would be kept. The Tassino had brought with
her a sacred table and plastic bags full of bath oils, cosmetics, mirrors,
movie posters, and broken mobile phones. After installing the shrine, she
instructed Sofo on how to feed the spirit that now resided there. After
about a week, the Ghanaians returned home and Sofo was left alone to
tend his spirit. Many people from the neighbouring villages visited him
daily and brought expensive gifts to the new spirit, asking services of her.
I continued to visit Sofo regularly, and after several months, noticed
that he had become anxious. The arrival of Mami Sika had altered Sofo’s
daily life significantly, and he often complained to me how hard it was to
obey all the restrictions that ownership of the bo and his new status as
Mamino (priest of Mami Wata) had imposed on him. Sofo could not drive
his car anymore. He could only wear shoes and a shirt on Wednesdays and
Mondays; the rest of time he had to wear a loincloth and twenty-five
necklaces made of heavy sacred beads. He could not have sex, drink
alcohol, or eat during the day. In the evenings his diet was limited to a
certain type of local fish called sika sika and steamed vegetables. Sofo also
faced growing pressure from other villagers and the vodou Bureau, as
there were rumours that he had acquired the money to buy the bo for Mami
Sika by stealing at least £1,000 from the presidential candidate’s donation
to the National Vodou Day festivities. The candidate claimed that he had
given £4,000 directly to Sofo, but the committee confirmed that they had
received only £3,000. Sofo was summoned several times to Cotonou, the
capital of Benin, to explain the disappearance of the money, but the
presidential hopeful left for France soon after losing the election, and
decided not to take any action against Sofo. However, Sofo’s reputation at
le Bureau National de Vodou was tarnished and he was excluded from
their meetings. Meanwhile, the villagers had also started to resent Sofo.
They had always been ambivalent about him, but the monetary scandal
had made them even more apprehensive. In general, however, Sofo was
still respected, since he still owned a very prestigious ahwan consisting of
both inherited bo pomési and purchased bo yovoni, to which the villagers
still came for help.
When I noticed Sofo’s weariness, I initially thought that it was due to
the pressures and problems he faced, but as I spent more time with him
The Vodou Priest who Lost his Spirit 125

learning about his ahwan I started to realise that his most serious problem
was fulfilling the needs of the spirits. Firstly, I noticed that Sofo spent
most of his time feeding the spirits or asking them what they desired. It
seemed that the spirits were never satisfied: the Sakpata spirits, for
example, had increased their food intake and required more and more
chicken blood. Each also had very specific demands for the type of
chicken it wanted. If one wanted a black rooster, the other would ask for a
brown chicken and then ask for something else. Sofo told me that the
Sakpata spirits were jealous and resented the presence of Mami Sika. I
asked him if he was afraid that the Sakpatas would leave their bos, but he
did not think that this was very likely, as the spirits were used to their bos
and loyal to his family. The problem was that if the spirits were
dissatisfied they would become less effective.
If Sofo was concerned about the Sakpatas, he was absolutely terrified
about feeding and serving Mami Sika in the right way. He told me that he
could not sleep because he was so concerned that Mami Sika would leave
him. He spent all the cash he could on expensive gin, soft drinks, sweets
and peacocks that his nephew bought in Cotonou. He told me that Mami
Sika found her shrine room too small, did not like being in a rural area,
and complained that Sofo could not serve her correctly. Most importantly,
the spirit claimed that the bo in which she resided was not well-made. I
found Sofo’s immense determination to keep Mami Sika in his bo a bit
disturbing, since he really seemed to be on the verge of a nervous
breakdown. He tried to hide this from his clientele as well as he could, but
soon rumours started circulating that the spirit was going to leave Sofo.
Several people even reported that Mami Sika was not cooperating with
Sofo and it was not worth giving gifts to her. I found this quite curious,
since on several occasions I had seen cases where people did not really
mind whether the spirit fulfilled their request. People went to the temples
to ask favours of the spirits and give them offerings, but by no means were
they certain that their wishes would be granted. However, it seemed that
Sofo himself and the villagers were slowly coming to a consensus about
the ineffectiveness of Mami Sika. A few weeks before I returned to
Europe, the secretary of le Bureau National de Vodou and his assistants
visited Sofo’s temple and examined the bo for Mami Sika. They reached
the conclusion that it was empty. Sofo had failed to keep the spirit with
him, and now he would have to purchase another bo from the same temple
in Ghana if he wanted to keep the status of Mamino.
Naturally, Sofo was devastated by the verdict. He told the whole
village that Mami Sika still resided in the bo and that the vodou Bureau
just wanted to tarnish his reputation in the wake of the monetary scandal. I
126 Anna-Maija Aguilera Calderòn

had noticed that the villagers did not have very much respect for le Bureau
National de Vodou and did not consider it an authority in determining the
effectiveness of different spirits. However, the departure of Mami Sika was
something that everyone including Sofo had been expecting, so this time
people agreed with the Bureau, and the village chief told Sofo that he was
no longer allowed to call himself a Mamino. When I was back in Europe a
friend of mine in the village wrote to me, saying that Sofo had left the
village again. His sister was now taking care of the temple and it was
believed that Sofo had gone to Burkina Faso to trade in old mobile
telephones and cars, though there were rumours that one day he might
return with a new bo for Mami Sika. My friend added that this time, if
Sofo was wise, he should also purchase a reme, a kind of spirit companion
from the same temple, so that Mami Sika would not feel so lonely in the
countryside with the jealous Sakpata spirits.

Can vodou spirits be “modern” or “traditional”?

In this section, I move from my ethnographic study of a particular
vodou priest to a consideration of its implications for our understanding of
large-scale abstractions such as “tradition”, “modernity”, “consumerism”,
and “globalisation. Sofo lost his precious vodou spirit because the
villagers, le Bureau National de Vodou, and even he himself believed that
the spirit had left her bo. In some ways, the villagers connected the loss of
the spirit to Sofo’s role in the monetary scandal; and in others, it was the
result of his ambiguous position within the village community. I have
presented an account of Sofo’s unsuccessful attempts to incorporate a
purchased deity to live alongside inherited deities, because I believe it
conveys a crucial point about what is missing in recent anthropological
studies of Mami Wata and West African vodou.
During the past three decades, a rich and extensive literature has
developed around the Mami Wata phenomenon (Bastian 1998; Drewal
1988, 1986; Fabian 1978; Frank 1995; Houlberg 1996; Jules-Rosetta
1994; Kramer 1993; Surgy 1994). The underlying assumption in many of
these works is that Mami Wata practices constitute an implicit critique of
the depersonalising and alienating effects that accompany Africa’s
integration into a capitalist system. Mami Wata practices are seen to
express both fascination and disenchantment with a world governed by
impersonal market forces and “anonymous foreign capital” (Fabian 1978,
327). For instance, Drewal has argued that Mami Wata devotees are
engaged in a dynamic process of redefining their selves and their position
in a changing society by interpreting Western consumer items through
The Vodou Priest who Lost his Spirit 127

indigenous concepts (1996, 308, 328-329). He suggests that Mami Wata

practices are ritualised reconstructions of European otherness, whereby the
devotees attempt to control and exploit external social, political, and
economic forces (1988, 102, 132). Mami Wata’s Caucasian features—
white skin-colour and long flowing hair—have similarly been interpreted
by Bastian (1998) and Houlberg (1996, 34), among others, as African
appropriations of European ideals of beauty that were introduced through
globalisation. According to Fabian (1978, 327) and Frank (1995, 342),
Mami Wata practices articulate local concerns about the displacement of
socially embedded subsistence economies within the impersonal logic of
capitalist market moralities. The “new economic morality” described by
Frank consists of a shift in conceptions of value, such that the European
tendency towards an “individualisation of economic goals” replaces social
relations within the traditional gift-giving economy (1995, 342).
The work of John and Jean Comaroff (1993) offers a useful starting
point for looking critically at the dichotomies and unilinear concepts of
modernity contained in much of the literature on Mami Wata. They point
out that such modernist ideologies are bound to an aesthetic of “nice
oppositions” (1993, xii). These ideologies reduce the complexity of reality
to binary contrasts (ibid., xii). Conversely, the Comaroffs argue that the
spread of ideas about modernity is based on a constant interaction between
the “global” and the “local”, and that this interaction simultaneously
changes the meaning and content of “modernity” and “tradition”. This
dialectical process leads to “multiple modernities”, whereby global
influences are translated into various languages of local meanings (ibid.,
Despite the usefulness of “multiple modernities” as a critique of
unilinear notions of modernity, however, I argue that this remains a
misleading framework by which to analyse the Mami Wata phenomenon.
According to Englund and Leach, anthropological attempts to highlight
and analyse the cultural heterogeneities of globalisation are organised by a
singular “meta-narrative of modernity” (2000, 225-228) that draws upon
familiar sociological abstractions such as time-space compression,
commodification, individualisation, disenchantment, and re-enchantment.
These underlying assumptions hamper the production of ethnography as a
form of reflexive knowledge and risk reducing it to an empiricist critique
of theories of modernity. The idea of “multiple modernities” thus
“bespeaks a view of modernity’s historical origins in the West” (ibid.,
237). Several anthropologists, including the Comaroffs, acknowledge the
danger this kind of Eurocentrism poses, and attempt to alleviate it by
claiming that Western modernity is no less magical than modernities
128 Anna-Maija Aguilera Calderòn

elsewhere (Comaroff and Comaroff 1988, 65). However, as Englund and

Leach suggest, the concept of “multiple modernities” creates an
epistemological vacuum in which the “wider context” inadvertently
represents the anthropologist’s own superior understanding of the world
(2000, 237).
It is true that at first glance Sofo seems to possess many characteristics
of the sort of Mami Wata devotee described in much recent
anthropological literature on the phenomenon. Sofo was interested in
European fashion and films, and was keen on cars, mobile phones, and
computer games. He dreamed about becoming rich and had several
business ideas that could be classified as “modern” within the literature in
question. Moreover, tensions between Sofo and other villagers came to a
head when he tried to incorporate a deity that he brought from a foreign
country into a temple where his inherited spirits—which recent studies on
Mami Wata might classify as “traditional”—resided. However, a closer
examination of the ethnographic and historical evidence pertaining to this
particular case reveals that it would be misleading to claim that Mami
Wata worship simply gave Sofo the means to cope with “modernity” and
“capitalism”. There are number of ways in which the small-scale
ethnographic material reproduced in this chapter challenges this claim.
Sofo was born into a village where spirits are an essential part of every
domain of life. From a very young age he learned from his father how to
feed the spirits and ask for their services. He was also well-connected in
the vodou community and very aware of spirit trading networks and
methods of buying prestigious bos. It would thus be misleading to argue
that an external force like a Western-style market economy was the
impetus for Sofo’s desire to purchase a bo for Mami Sika—although it did
require a new range of objects which he could offer to the spirit—or that
he somehow expressed his ambivalence about “modernity” and
“capitalism” by giving Western luxury items to her. Sofo was drawn to the
Mami Wata pantheon at least in part because possession of these spirits
leads to wealth and high status, but this was not because he saw Mami
Wata spirits as “modern” or “foreign”. Rather, his desire reflected the
widespread understanding that an ahwan of spirits becomes stronger if it
combines inherited and purchased spirits, since this widens the array of
problems they can solve. Thus Sofo’s desire to create a combined ahwan
can be seen as a part of a long historical process in which the regional
trade of spirits has well-established precedents. It is implausible to
imagine the coastal people of the Bight of Benin as “precapitalist”,
especially given their participation in several centuries of slave trading,
gold mining, oil palm plantations, and other forms of trade. Moreover,
The Vodou Priest who Lost his Spirit 129

“exotic” commodities from elsewhere have flowed into what is now

modern Benin for at least four hundred years, through shipments of
cowries, currency metals, and luxury goods during the era of the slave
trade (1553-1850).
Similar considerations arise in the study of the vodou spirits. Firstly,
since Mami Wata spirits are the richest spirits in the pantheon, it is logical
that they only want the most expensive food and items—which in this case
happen to be imported luxury goods. For the villagers I worked with, the
categories of “modernity” and “tradition” are insignificant in this context.
Secondly, monetary exchange is inherent to any vodou relationship: it
enables the owner of the bo to acquire and maintain the spirit, and
supplicants to engage with it. During my fieldwork I noticed that the
pedigree of every vodou spirit in the village apart from those of the
ancestors could be traced to a foreign country: Sofo’s inherited Sakpata
spirits, for example, were originally from the Ivory Coast. Some spirits
have thus stayed longer in the village than others; but the new arrivals
cannot be seen as the products of “globalisation” or arrival of “capitalism”,
since the purchase and trade of foreign vodou spirits has taken place for
several hundred years.
How, then, might an anthropologist begin to engage with a situation as
historically specific and complex as this? My answer is to look at the
categories the subjects of the anthropological analysis themselves use. For
the Beninese villagers with whom I worked, the two main categories of
spirit residences—purchased bo yovonis and inherited bo pomésis—do not
fit into the categories of “modernity”, “consumerism”, and “tradition” that
have commonly been used by anthropologists in the analysis of Mami
Wata worship. These categories do not fit well within theoretical models
that naturalise value and exchange-value or moral and market economies.
One must struggle to divide the economic formations of contemporary
Benin, and specifically this village, into separate beds of gift and
commodity exchange. Although the dichotomy between gift and
commodity has been criticised widely in anthropological literature over
the last thirty years (notably by Humphrey and Hugh-Jones [1990] and
Parry and Bloch [1989]), anthropological studies of Mami Wata worship
have rarely drawn on such insights. Rather they implicitly assume that
“capitalism” and the use of money, as Parry and Bloch put it, “acts as a
kind of acid which inexorably dissolves cultural discriminations, eats away
at qualitative difference and reduces personal relations to impersonality”
(1989, 6).
In recounting the story of Sofo, a somewhat marginal and socially
unpopular vodou priest who lost his spirit and fled the village, I have not
130 Anna-Maija Aguilera Calderòn

attempted to claim that his experiences as an individual are characteristic

of the general way of life and vodou worship in the village. Nor do I wish
to use his experiences as a way of criticising anthropological theories of
“globalisation”, “modernity”, and “tradition”. I have simply chosen to
present this account to convey information about the categories that every
single villager uses and finds important and self-evident. I believe that by
examining the categories of exchange that the villagers themselves use,
one can begin to understand how life in general is constituted in this
particular place. This does not mean that the arrival of the market
economy and the use of Western consumer items in vodou should be
dismissed as irrelevant, but merely that our analytical focus should be on
the categories that our informants use.
In this chapter I have used the narrative of Sofo’s triumphs and
misfortunes to render analytical the local categories of bo yovoni and bo
pomési, thereby treating them as the basis of my argument about exchange
relationships in vodou practice. In so doing, I have tried to avert a common
pitfall entailed in using anthropological notions of “modernity”,
“consumerism”, and “tradition” as analytical touchstones for local
ethnographic contexts: the danger of inadvertently bringing into the
analysis inaccurate assumptions about the existence and meanings of
spheres of exchange, which reproduce the conclusions already inherent in
such large-scale abstractions. In the process I have given these conceptual
contexts evidentiary and analytical significance. My ethnographic account
thus stands as a form of evidence in itself. According to Christina Toren,

This mode of analysis is capable of revealing the material validity of our

own categories and the limits of their application. By the same token it
allows us to access the preoccupations of the people with whom we work
and a means whereby their categories—perhaps very different from our
own—can be rendered analytical. (2002, 122)

By rendering analytical the categories surrounding networks of spirit trade

and relationships with the spirits used by the subjects of this study, one
can arrive at a rich and nuanced understanding of the dynamic and
complex social world that brings vodou worship in Benin into being.

I am grateful to the Laura Ashley Foundation for funding my doctoral
research in the form of a studentship. Fieldwork was in addition supported
by travel grants from the Nordic Africa Institute and the Finnish-African
Cultural Centre Villa Karo. This article began as a short paper at CRASSH
The Vodou Priest who Lost his Spirit 131

"Questions of Evidence" conference in Cambridge (2007). I am grateful to

Liana Chua, Casey High, Timm Lau, Andrew Beatty and James Frazer for
their remarks on earlier drafts.

Bastian, Misty L. 1998. Mami Wata, Mr White and the sirens off Bar
Beach: Spirits and dangerous consumption in the Nigerian popular
press. Afrika und das Andere: Alterität und innovation, ed. H.Schmidt
and A.Witz. Munster: LitVerlag.
Comaroff, John L., and Jean Comaroff. 1993. Introduction to Modernity
and its malcontents: Ritual and power in postcolonial Africa. Chicago:
University of Chicago Press.
—. 1988. Through the looking glass: colonial encounters of the first kind.
Journal Historical Sociology 1 (1): 6-32.
Drewal, Henry J. 1988. Mermaids, mirrors, and snake charmers: Igbo
Mami Wata shrines. African Arts 21 (2): 38-45.
—. 1996. Mami Wata shrines: Exotica and the construction of self. In
African Material Culture, ed. Mary J. Arnoldi, Chistraud M.Geary, and
Kris L.Hardin. Bloomington: Indiana University Press.
—. 1998. Interpretation, invention and re-presentation in the worship of
Mami Wata. Journal of Folklore Research 25 (1-2): 101-139.
Englund, Harri, and James Leach. 2000. Ethnography and the meta-
narratives of modernity. Current Anthropology 41 (2): 225-248.
Fabian, Johannes. 1978. Popular culture in Africa: Findings and
conjectures. Africa 48 (4): 315-334.
Frank, Barbara. 1995. Permitted and prohibited wealth: Commodity-
possessing spirits, economic morals and the goddess Mami Wata in
West Africa. Ethnology 34 (4): 331-346.
Hugh-Jones, Stephen, and Caroline Humphrey, eds. 1992. Barter,
exchange and value: An anthropological approach. Cambridge:
Cambridge University Press.
Houlberg, Marilyn. 1996. Sirens and snakes: Water spirits in the arts of
Haitian vodou. African Arts 29 (2): 30-36.
Jules-Rosette, Bennetta. 1994. Simulations of postmodernity: Images of
technology in African tourist and popular art. In Visualizing theory:
Selected essays from V.A.R 1990-1994, ed. Lucien Taylor. London:
Kramer, Fritz. 1993. The red Fez: Art and spirit possession in Africa.
London; Verso.
132 Anna-Maija Aguilera Calderòn

Parry, Jonathan, and Maurice Bloch, eds. 1989. Money and the morality of
exchange. Cambridge: Cambridge University Press.
Surgy, Albert. 1994. Togoville: Le Roi-pretre des Evhé du sud-Togo.
Lomé: Presses de l´UB.
Toren, Christina. 2002. Anthropology as the whole science of what it is to
be human. In Anthropology beyond culture, ed. Richard G. Fox and
Barbara J. King. London: Berg.


Over the last several decades, personal disclosure has become an
effective, albeit embattled, standard practice in ethnographic writing. Yet
anthropological reflexivity is often very selective, and can detrimentally
obscure the degree to which our own experiences—especially in unstable
fieldsites—can fundamentally shape ethnographic insight, or underlie
encompassing anthropological models or theories. In conflictive settings,
where “anthropologist and subject inhabit the same complex world”
(Jenkins 1994, 434), and where we are more likely to be emotionally
“penetrated … by another form of life” (Geertz 1988, 4), we can be
especially vulnerable. Indeed, when research occurs in perilous spaces rife
with political upheaval or interpersonal or religious turmoil, moral
uncertainties shape and embitter our on-the-ground interactions, while
private predicaments impact our social positionality, research choices, and
the ability to process or convey participants’ experiences. By
reconsidering knowledge acquisition in a way that stresses my positional
and largely unresolved fallibilities, my chapter answers Timothy Jenkins’
call for postmodern ethnographers to re-evaluate the wider spectrum of
their field-based moralities, and move beyond their own “official account”
(Bourdieu 1977, cited in Jenkins 1994, 436).
My Northern Pakistani doctoral fieldwork was an emotional and
dangerous time, when my academic professionalism and private beliefs
were unbalanced by everyday, interpersonal battles. It brought out some of
the very best, and the worst, in me. In response to the insights I gained
from my fieldwork struggles, this chapter attempts to augment and
challenge the widespread use of reflexive tropes that accord primacy to the
“naïve” ethnographer, whose early fieldwork mistakes, inter-cultural
blunders, and lapses in knowledge are then rectified over time. By
134 Emma Varley

reflecting on my own active and deliberate role in Shia-Sunni

sectarianism, interfamilial vengeances, and Northern women’s use of
magic and counter-magic, I discuss how my doctoral research was
anything but “disciplined stillness” (Lambek 1997, 48). The last half of
my fieldwork was mired in the chaos emanating from a violently contested
broken marriage with enmities infusing nearly every aspect of my field
research. Although these troubles resulted in enormous evidentiary and
theoretical benefits, my post-fieldwork efforts to find expressive space or
productive applications for my difficulties confirmed that ethnographic
reflexivity remains constrained by a quiet political correctness.
Notwithstanding repeated calls for ethnographers to address the hubris
inevitably attendant to fieldwork or its morally uncertain underpinnings
(see Jenkins 1994; Meneley and Young 2005; Strathern 1987), our use of
reflexivity demonstrates sequences of narrative distortion. Specifically,
even problematic admissions do not, by nature, tend to acknowledge
personal wrongdoing or mal-intent, even though the process of
acknowledging these can be an enormous site of cultural understanding.
To this end, Bourdieu comments on the charities inherent in ethnography’s
subtle “objectivication of the point of view” (Bourdieu 2001, 94). In our
writing, we are undeniably generous to ourselves, smoothing away many
of our own rough edges while foregrounding the foibles and missteps of
our participants. More importantly, by failing to reconcile our analysis
with this broader scope of our field-based “selves” and by delimiting the
realm of acceptable self-disclosure, we marginalise the evidentiary
potential of our shadowy subjectivities, moral lapses, or the social games
that go even further than the knowledge-seeking coercions of Rabinow’s
“symbolic violence” (see Geertz 1988, 97).

Reflexivity (un)done
To varying degrees and in different ways, we all face field-based trials
and tribulations. In retrospect, many of these difficulties reinforce the
veracity of our inquiry. But in the same way that ideological frameworks can
obscure relations of domination (D’Andrade 1995, 401), ethnographic
reflexivity tends to mask the wider experiential scapes underlying our
theoretical development. And in cutting away our predilections and
misbehaviours, we risk a kind of artificial professionalism or preclude
acknowledgment that our theorising might draw on deeply emotional or
problematic events. Because we “bring to bear on [our] work a unique
biographical background and configuration of personal interests and values”
(Honigman 1976, 244), consciously or not, ethnographic production is
Enmities and Introspection 135

inextricably enmeshed with the personhood of the anthropologist as social

actor, “self”, or author. Our fieldwork is tacitly guided by our personal
predilections, and our texts imbued with quiet autobiography. By logical
extension, our work inevitably reflects or integrates not only positive aspects
of our selfhood, but also the more problematic or morally uncertain. In
multiple ways, carefully editorialised reflexivity is imbued with an
unrealistic moral optimism. Borrowing from Holly Wardlow’s theorising of
Huli sex workers’ narratives (2006), in my opinion, such reflexivity often
represents the same kinds of “triumphalist language of self-determination”
or retrospective agency (Wardlow 2006, 158) that we work so hard to
trouble in our participant’s narratives of “self” and sociality.
It has been well demonstrated that reflexivity’s “personal approach”
encompasses “an observer’s sensitivity, depth of thought, speculative
ability, imagination” (Honigmann 1976, 244), yet current reflexive tropes
remain purposefully and unproductively incomplete. For instance, when
ethnographic research aspires to scientism and objectivity, reflexivity can
become nihilistic. By safely displacing our personhood from ethical
critiques, this type of reflexivity results in what John Van Maanen
humorously monikers “‘the doctrine of immaculate perception’” (Van
Maanen 1988, 74). Postmodernist reflexivity, on the other hand, has been
critiqued as “dead-end indulgence and narcissism” (Quayson 2003, 7,
cited in Gibb 2005, 225). Here the ethnographer’s emotionally charged
confessional excesses are usually under-attentive to local socio-cultural
landscapes or largely devoid of evidentiary value. The most frequently
employed tropes, however, narrate our fieldwork as rife with early
misunderstandings and benign missteps. Such reflexivity readily
acknowledges how our early transgressions provide preliminary insights
into social boundaries or cultural customs. In other less productive ways,
by prioritising fieldwork’s regular pitfalls and common inter-cultural
gaffes, ethnographers sideline more ominous personal failings, or bury the
behaviours that disrupt our self-styling as professionals or verge on ethical
indiscretions. These might include romantic trysts, exploitive work, or
research partnerships, or, as in my own case, active involvement in
protracted family battles and community enmities. Largely because of the
ethical challenges facing fuller disclosure or the absence of academic
support, ethnographers are left with few safe spaces in which to explore
their own problematic contributions to tenuous field encounters (see
Lambek 1997; Young 2005; Zulaika 2004). Moreover, ethnographers may
be reluctant to move into the shady spaces that unsettle their sense of
“self”. Paraphrasing Wittgenstein, the anthropologist Joseba Zulaika notes
that “genuine understanding is frequently impeded by pride, not lack of
136 Emma Varley

intelligence” (2003, 97). When we as ethnographers wrestle these realities

away from our written work, we can produce one-dimensional or
unrealistic reflexivity. Herein, our written personhood is incomplete and
strangely devoid of fallibilities. Seen in this way, our insertion of “self”
into texts can be, in some ways, a game of words or characterised by
perceptible subterfuge. Similarly, Geertz archly redescribed the
ethnographer’s contemplative stance as being, in fact, a “pretense” (1988,
98). To this point, Geertz commented on the predicaments of ethnographic
authorship, whereby “getting themselves into their text…may be as
difficult for ethnographers as getting into the culture” (ibid., 16).
Deeply mindful of the limitations of current reflexive tropes,1 various
anthropologists—including Bent Flyvbjerg (2001), Michael Lambek
(1997) and Nancy Scheper-Hughes (1995)—have re-oriented their self-
reflection to demonstrate something more “in-between”. As both concept
and practice, ends-directed, contextually contingent, hermeneutic, and
morally attentive reflexivity enables “meaningful engagement with others”
(Lambek 1997, 33), especially when conducted in a spirit of critical
ethicism, conscientious positionality, and recognition of the ethnographer’s
own historicity (ibid., 35). Such orientations enable intensely personal
disclosures, many of which carry evidentiary status. These include the
insights incurred by pregnancy termination (Rapp 1999), unauthorised
expeditions into phenomenally dangerous places (Bourgois 1990), erotic
and ethically ambiguous fieldsites (Kulick 1998), the relational anxieties
accompanying field relationships (Lambek 1997; Young 1996), or the
frustrated witnessing of infant deaths (Scheper-Hughes 1992). Alternately,
the anthropologist’s “native” identity confirms how ethnic or cultural
“insider” positioning affords additional insights (Abu-Lughod 1988;
Narayan 1993). Despite these attempts by practitioners of ethnography to
include the reflexive aspects of fieldwork in their accounts, what remains
fundamentally un-addressed is how even the most emotionally cumbersome
aspects of our personhood or ethically challenging events might also carry
evidentiary potential or invigorate our theorising and analysis.
Undoubtedly, while there are moving examples of on-the-edge and
insightful ethnographic confessions (see Rosaldo 1996; Zulaika 2003), my
chapter argues that the fundamental difference between those confessions
and mine lies in my not merely being a participant, but also an antagonist.
In using my fieldwork quandaries and ethical lapses as both an analytical
platform and ethnographic evidence, my chapter is not intended to
displace our focus on the “other” or unbalance reflexivity as a whole.
Instead, I seek to expand our practice of reflexivity by suggesting that the
unhappy dynamics affecting my fieldwork and stemming from myself—
Enmities and Introspection 137

which would normally be sidelined in ethnographic writing—were

ultimately a germane component of my analysis and theorising. Just as
importantly, this chapter will demonstrate that my contributions to, and
experience of, fieldwork difficulties were in and of themselves evidence.
Albeit unintentionally, my most troublesome and antagonistic involvements
with Gilgiti community life often produced more meaningful insights into
women’s life-worlds than the data collected using more standard
ethnographic methods. Through sharing my story, I will argue how, if only
occasionally, the ethnographer’s inclusion of their shady subjectivities and
moral gamesmanship can contribute to more productive, if embattled,
fieldwork accounts.

Strategising fieldwork enmities

My desire to counter partial reflexivity through more rigorous self-
accounting necessarily requires some disclosures. My own “long
apprenticeship” entailed acquiring a wide array of uncomfortable “habits
of action for coping with reality” (Bourdieu 2001, 5). After an absence of
nearly five years, my family and I returned to Gilgit in 2004 for my
doctoral fieldwork. Gilgit is the administrative and economic capital of
Pakistan’s semi-autonomous Northern Areas—a region of incredible
beauty. Surrounded by the Karakoram Mountains and cut through by
glacier-fed rivers, Gilgit Town is populated by fifty thousand Shia, Sunni,
and Ismaili Muslims in equal numbers, and happens to be the birthplace of
my Sunni husband, Wadood. For Gilgitis, shared social spaces have not
always been peaceful. Every few years since the mid-1970s, the town
erupts under the pressure of mounting sectarianism, fuelled by competing
Shia-Iranian and Sunni-Pakistani religio-political forces. The happy
commensality and inter-marriage that had once characterised Gilgiti Shia-
Sunni relations was irreparably fractured by clandestine, post-
Revolutionary Iranian politicisation, which was matched by Pakistani
General Zia Ul-Haq’s Sunni Islamisation schemes. Each had manoeuvred
Gilgit’s sectarian communities into opposing positions so as to uphold
their respective nationalist agendas. In 1988, a massacre of Shias by
marauding Sunni militants resulted in hundreds of deaths and property and
livestock destroyed. But where Shias and Sunnis have competitively
weaponised, Ismailis, led by the Aga Khan, remain staunch pacifists.
Between the late-1990s and our 2004 return, Gilgit had settled into an
uneasy peace. The bucolic summertime beauty of fruit-laden apricot trees,
blooming roses, and wind-rustled fields of barley concealed a disturbing
history. “There, a Shia had his throat cut,” or, “here, a Sunni was shot,”
138 Emma Varley

locals would recount, pointing to some desolate stretch of rocky hillside

above town, or an alleyway shaded by a mulberry tree.
Spurred on by my early married years in Gilgit when Ismaili and Shia
friends scoffed at my decision to marry into a “group of Sunni animals”
and then my ensuing adventures as the mother of four, I wanted to explore
how sectarianism invited discord into patient-physician interactions
between Shia health providers and Sunni maternity patients. I wanted to
explore how sectarian animosities led to Sunni women’s decreased uptake
of health services, and therefore exacerbated the region’s already dire
maternal mortality statistics. Six months into my fieldwork, the worst
sectarian fighting in Gilgit’s history powerfully confirmed my approach.
After the Sunni-orchestrated assassination of their Shia religious leader
Imam Syed Aga Zia-u’din Rizvi on 8 January 2005, Northern Areas’ Shias
pulled out all the stops to avenge his murder. Caught off-guard and
handicapped by persistent community infighting, Sunnis haphazardly
strategised their defence, and Gilgit was overwhelmed by protracted Shia-
Sunni fighting, targeted killings, army curfews, police surveillance, and
media blackouts.
Besides murdering more than twenty Sunnis in the hours following
Zia’s death, shootings, bombings, and hatchet attacks were staple features
on the eighth of every month as Shias killed Sunnis in memorial
retaliations. Victims included Gilgit’s Sunni Inspector General of Police
and his family, the Northern Areas Health Director, and Education and
Forestry officers. Subsequent to Zia’s death, roadblocks, threats of
violence, or the curfews often restricted maternity patients’ access to
health clinics. Shia strikes regularly coincided with the cloaking darkness
of timed electricity “load-shedding”. Because many power station
employees were Shia, Sunnis were convinced of extensive conspiracies.
Sunni neighbourhoods refused Shia police protection for fear of pre-strike
surveillance. With Sunnis murdered a few hundred metres from our front
door on 8 January, we had fled our riverside home in an all-Shia
neighbourhood. We relocated to one of Gilgit’s safest mountainside Sunni
enclaves where, surrounded on three sides by Shia boroughs, our only
feasible escape was through an adjacent, neutral Ismaili area.
Throughout 2005, Shias calculatingly interwove their regularly
occurring ritual observances with their anti-Sunni stance. Notwithstanding
their beauty, the regular burning of night-time mountainside þiragaan—
massive bonfires shaped out in the names of Shia Imams—were threatening
reminders that we, as Sunnis, were surrounded. By marriage I was
considered a Sunni, a religious assignment I was unable to disentangle
myself from during Gilgit’s “tension times”, as locals called the Shia-
Enmities and Introspection 139

Sunni hostilities. On festival nights, hundreds of young Shia men

synchronised their descents off the mountain and marched, shouting
threats and with flaming torches in hand, through oddly quietened Sunni
areas en route to attend rallies (jalsa) at local Shia Imambaraghs
(mosques). For the duration of each festivity, Sunni mohallas
(neighbourhoods) were closed off by army curfews to prevent street fights
between marchers and angry Sunni onlookers. One such night, as several
hundred men passed along the street near our house, I sat on our darkened
front steps and, along with my neighbours, yelled back anti-Shia slogans
for each anti-Sunni epithet. Rather than being distasteful, it felt
exhilarating to turn the tide of insults back at our “aggressors”. In the early
springtime of 2005, during the unsettling Ashura rituals where Shias mark
the death of the Caliph Ali’s son Hussain, police shut the town down as
thousands of Shia men took to the major roadways. My in-laws and I later
watched a video of that year’s Ashura procession, when men had whipped
at the flesh of their backs while marching through the center of town; the
tinkling sounds of the whirring chains a morbid cadence against the soft
shuffle of their feet. Many men had carried a set of chains with which to
strike themselves, called zahnzeer, with each chain ending in a carefully
sharpened blade. And for hours throughout the late afternoons and well
into the nights, particularly during the first half of the month of Muharram
(in which Ashura takes place on the tenth day), Sunnis living near to
Imambaraghs listened to the broadcast sound of Shias’ lamentations and
mourners rhythmically beating their chests (matam). Stuck inside their
homes by the curfews enforced on those days, Sunnis sat drinking chai
(tea) and munching, almost disrespectfully it seemed, on biscuits and pfitti
bread, while they decried Shias’ Ashura observances as bizarre recompense
for “failed” ancestral battles, and their invocation of the Prophet’s son-in-
law, the fourth Caliph Ali, as a kind of heresy (bid'at).
Because the Pakistani Army had stymied weapons shipments into
Gilgit, during times of increased tension Sunnis established night-time
“duty teams” to make the most of their dwindling resources. In two-hour
shifts throughout the night, squads of armed men guarded vulnerable
Sunni mosques, roadways, fields, and homes, including ours. When army
officers searched households one by one during the curfews, we plotted
how best to hide our unregistered weapons and ammunition; in the
absence of an effective police response, losing them could have been
disastrous. During one springtime weapons sweep, I was pleased with
myself for thinking to hide our Kalashnikov by tying it into the perfumed
branches of our cypress tree. During the safer daytime hours, our rustic,
verdantly green, and tranquil neighbourhood offered some respite from our
140 Emma Varley

night-time paranoias. We excitedly discussed rumours that Shias were

hiding in the darkened corn fields near our house, waiting to seize on
Gilgit’s disharmonies and strike at unsuspecting Sunni neighbours, many
of whom were also their distant relatives. In spite of the growing tension
of sectarian strife, my husband Wadood’s genuinely happy and ongoing
family relationships with our Shia neighbours, many of whom showed us
considerable hospitality, made describing the “tensions” using reductive
aggressor-victim dyads painfully awkward.
Yet, in the midst of my shames and fear, I was proud of my comportment
and my rapid acquisition of a conflict-related feel for the game. Our
neighbours, however, gossiped that I was a nervous wreck. Even after eight
years of marriage and being a known quantity—with most of my Western
exoticisms worn away—many women neighbours remained vengefully
preoccupied with me. I discovered that the reason for this preoccupation was
that Wadood had divorced his first wife Cheshmagula, a local Sunni woman
and the mother of his four daughters, before marrying me.
Shortly after we had relocated to our new home, it came to light that
even after five years in Canada with nary a phone call, letter, or message
between them, “Cheshma” wanted a reconciliation with Wadood. Upon
our return to Gilgit six months earlier, Wadood had realised with horror
that during his absence, his parents had angrily repudiated his original
verbal oath of divorce (talaq), and re-installed Cheshma in the family
home, proclaiming he would return to her after tiring of me. However, he
soon made it plain that this reconciliation would not take place. It then
took several months to shift Cheshma’s belongings back to her natal home.
Left behind, hidden in a dark corner of the family home, one of my sisters-
in-law discovered one of her suitcases, filled with amulets (tawiz) prepared
against me, each containing scraps of my clothing that had somehow been
stolen years before and were now emblazed with the arcane script
characteristic of “black magic” (kala jadu). Setting Cheshma's acrimony
for me aside, and still openly doubtful that Wadood was completely
finished with her, Wadood’s parents suggested he “re-divorce” her. To
their dismay, he quickly obliged. Cheshma and Wadood’s families were
deeply wounded by Wadood’s apparent disregard for the multiple
interlinkages their arranged marriage had spawned, many of which were
firmly reinforced by business and monetary dealings. However, and even
despite being visibly upset, my in-laws and Cheshma’s family seemed to
have accepted Wadood’s decision. I was convinced we were finally
finished with this protracted marital drama, but Wadood was not nearly as
optimistic and remained uncertain about what would happen in the coming
weeks and months.
Enmities and Introspection 141

Amid our fears of Shia aggressions, Cheshma’s family, who happened

to be our neighbours, secretly planned to scare me away from Gilgit and
from my marriage. When local tensions peaked and the staccato pops of
pistol fire and percussive blasts of LMGs reverberated off the rocky
hillsides, Cheshma’s brothers surreptitiously used slingshots to launch
volleys of rocks onto our corrugated tin roof with the intention that we
would assign blame to the Shias, which we did. Startled by the noise, our
“duty teams” climbed our garden walls with weapons carefully aimed to
spy out into the alleyways snaking around our house. Inside the house,
Wadood swore about “Shia maniacs”, and I raced to cut the lights and
settle our terrified children into corners safe from possible gunfire. Some
of the worst rock attacks came the fear-soaked night our next-door
neighbour, Wadood’s cousin, was killed by Shias in the minutes prior to
electricity being restored. I was certain that each barrage of stones
heralded an imminent assault by militants through our side gate, which
faced a row of Shia households; my terrors were indescribable. And while
trying to dodge the rocks as they flew into our garden intermittently from
the darkness outside, I had to wrestle Wadood back from joining our
enraged Sunni neighbours in the street; we were horrified to learn that
some of the men had torched local Shia stores later that night. Shortly after
this, thanks to the confession of one of Cheshma’s only-slightly guilty
accomplices, we discovered our nighttime aggressors’ true identities.
Shaking with rage and delayed fear, I realised that on any one of those
other nights, someone from her camp could have posed as a Shia militant,
shot Wadood, and successfully avoided blame. By necessity we worked to
continually, tactically distinguish between the threats coming from Shia
violences and the hazards posed by our neighbours, many of whom were
Cheshma’s relatives.
Cheshma and her family’s careful imbrication of their revenge with
Shia hostilities stunned and hardened me. Disgraced by his abandonment,
angered by my non-Gilgiti difference, and determined to get revenge
(badal) for the second divorce, Cheshma’s family were, we now realised,
far from accepting of the situation. In addition, Wadood confessed
intercepting several notes intended for me that were wrapped around rocks
and thrown into our garden, in which Cheshma proclaimed her love for
Wadood and his for her and warned me of my imminent divorce. She had
incorrectly assumed that in our years away, Wadood had become a
wealthy Canadian, and he now had no further use for me. While we were
most vulnerable, many of Wadood’s more amicably sociable family
abandoned us. Disappointed by my apparent unwillingness to provide
them with wealth in exchange for their earlier loyalties, they too hoped we
142 Emma Varley

would quickly divorce. Wadood admitted his mother had prophesised that
if I was enough disturbed, I would return to Canada and he might then re-
marry Cheshma. Alongside irrefutable proof of my relational vulnerabilities,
it hardly helped that at local social events or even in the clinics where I
was conducting my research, Sunni women, emboldened by Cheshma’s
story, confronted me for having “stolen” Wadood. One maternity nurse
dismissed my interview questions, arguing that I should be writing about
my own “big story”. Though the messy public-ness of this aspect of my
life felt chaotic and un-resolvable, by remaining in Gilgit, I tried to salvage
half-done research, confront my marital demons, and rescue the potential
innovations made possible by these conflict-addled interactions.

Learning (mis)steps
Set against the Sunni community’s defensive measures, Wadood and I
quietly strategised against fellow Sunnis, so as not to publicly feed Shia
characterisations of Sunnis as “deceitful” and “disloyal”. I quickly realised
that publicly fighting with Cheshma would only have partially ended my
anger in that I was also ashamed of having blamed “innocent” Shias for
many of the disturbances. Just as awkwardly, I was still resentful of the
Shias’ raucous intimidation tactics. In these moments, my unresolved fears
devolved into anger and prejudice, which for the next few months tainted
social encounters, conversations, and self-appraisal. And not unlike most
of women’s social embroilments, my tit-for-tat games with Cheshma were
conducted vis-à-vis gossip or through third parties, with only occasional
direct confrontations. Knowing the wide array of insults that Gilgiti
women hurled at each other during heated domestic fights, I framed
Cheshma as a mountain woman; unwashed, uneducated, bovine, shrewish,
and aggressively masculine. By doing so, I staggered my descriptions of
her against cherished ideals of Gilgiti womanhood which include qualities
such as modesty, subservience, kindness, and open-heartedness. And by
drawing on Sunnis’ frequent characterisations of Shias, I raged that our
sectarian antagonists were false victims, religious elitists, and duplicitous.
Distanced from both by perverse pride and the intractable boundaries
separating Sunnis from Shias, I actually knew very little about either
Cheshma or the Shias.
My resentments suffused interview sessions with neighbouring women
and, if pressed by indelicate queries, I mimicked the worst of Gilgiti
bigotry, endowing my vitriol with local, emotional weight. Most women
delighted in my moral degeneration, not only because it confirmed that I
was fallible, but because my struggles were comparable to their domestic
Enmities and Introspection 143

battles or loathing for Shia machinations. As I publicly navigated

Cheshma’s ongoing offensives, I garnered a small measure of sympathy
and my neighbours shared more of themselves. In turn, I pursued specific
narratives because they afforded comprehensibility of my predicament, a
fact that women were sometimes uncomfortably aware of because their
responses had the potential to personally challenge or disappoint me. As
delicately as possible, I strategically juxtaposed my experiences against
the lives of my participants, whereby my battles with Cheshma or
Wadood’s parents functioned as ethnographic equivalences and analogies.
Women seemed comforted not merely by my genuine interest or promises
of anonymity, but also by my ability to withstand the darker elements of
Gilgiti sociality.
There was an undeniably schizoid quality to those days, as I laboured
through protracted hospital and patient interviews during the day and
hardened myself to cosmologies of risk and violent threats at night.
Sectarianism’s predictable logistical hazards were eminently more
manageable than enmity’s more unwieldy, slippery emotional scapes. As it
was for most of my participants, everyday sociality was over-weighted
with uncertainty, watchfulness, and diagnosing intent behind every action.
The most generous gestures seemed the most suspicious. After graciously
accepting gifts of food or fresh fruit at our front gate, Wadood invariably
fed these succulent but potentially magically malicious offerings to our
dog. Every morning, Wadood searched the nooks and crannies of our
garden’s outer wall for hidden tawiz (amulets). At dusk, he circumnavigated
our house while reciting the Qu’ran’s protective Surah Yaseen, which the
Qari (mullah, cleric) described as the perfect “antidote” for Cheshma’s
jadu (magic). In lurid detail, the Qari recounted how women would pay
for tawiz to be made against their enemies and just before dawn, take them
to the local cemetery where they would be thrown into unkempt graves.
That night, I slept overburdened with images of tawiz bearing my name in
goat’s blood being thrown into broiling, glowing, red graves with puffs of
smoke signalling “magical” success.
When my health began to fail, the mullah suggested Wadood take me
to a spirit diviner who could, using the assistance of djinn, retrieve, “tawiz
nikalna” ('open') and disempower Cheshma’s remaining or newly created
amulets against me. Not long after, in the fetid gloom of the diviner’s
village house, she caught a small padlock from where it materialized in the
air in front of me. She ordered it smashed and the broken pieces thrown
into “clean water”. My bewildered silence was sharply interrupted by the
excited shoutings of the woman who had accompanied me: “Do you know
what this means? She has tried to lock your hearts against one another!”
144 Emma Varley

Though improvisation and mimesis are normative facets of our fieldwork,

my training had hardly equipped me to cosmologically mediate death
vengeances while remaining vigilant to everyday sectarian dangers. I had
known women’s games-playing existed, but this didn’t necessarily mean I
now knew how to join the fight or how to improvise my way out of
matrices of animosity and evil intent in ways that carried meaning for me,
Cheshma, and our family complexes. Unwilling to seek out local jadugar
(magicians) to neutralise the harms directed at me, I found other means of
attacking her. But what could I do and not regret? What carried the most
meaning and retaliatory weight in Gilgiti terms? How could I be seen as
“victim” and not also as “perpetrator”?
With relief, I found women’s tactics surprisingly predictable.
Cognisant that the emotional intensity of a reaction to an abrupt “hit” was
viewed as especially rewarding, I kept my miseries largely to myself.
Devastated that my responses to their antagonisms were not on view,
Cheshma had to keep “hitting me” or, as neighbours suggested, “picking at
our wounds” with phone calls, notes, and public confrontations. The more
graciously I refused to fight back, the more infuriated and harder she
pushed. The more morally I acted, the more she looked, by comparison,
devious or “jungulee” (uncultured). When she went with her mother to the
Qari (mullah, cleric), and swore on the Qu’ran that she was “giving”
Wadood’s four daughters to me, I argued that within Islam, it was not my
right (haq) to receive them. Neighbours approved of my quiet handling
and mocked Cheshma’s apparent disinterest in her children, though we all
knew better. By burdening me with the children, she had hoped I would
flee or that my refusal to accept them would demonstrate my supposed
callousness. When her cousins angrily confronted me at a wedding, asking
me why Wadood had divorced her, I “shyly” demurred with a smile,
suggesting that perhaps this was an affair her brothers should be bringing
to Wadood’s attention as it was not my Islamic responsibility. Islam’s
clear-cut strictures and emphasis on male responsibility became a refuge
of sorts, an inarguable “high road”.
The more I was stressed and sleep-deprived, the more susceptible I
became to breaks between my academic demeanour and my private feelings
and bad behaviour. Occasionally, I grew impatient with my performance and
seized on available opportunities to reverse the discomforts thrust at me by
Cheshma’s family. A relative of Cheshma’s married Wadood’s best friend
and, when showing me her wedding pictures, deliberately pointed at
Cheshma in each shot with a hennaed forefinger with the comment, “Pyaree
chachazad” (“Beloved cousin”). I held my composure but later that night
asked her mother-in-law, a dear ally of mine, to see the photograph again.
Enmities and Introspection 145

Conspicuously, I began taking digital pictures of the print. Watching and

unable to intervene, the bride’s embarrassment was soon mirrored by my
growing shame as it took snap after snap after snap to capture a clear
image. It became obvious that, in the same way Cheshma used portions of
my stolen dupatta (veil) to concoct spells against me, I wanted a piece of
her too. Knowing she would soon find out and being aware that
photographs are viewed as a particularly acute point of cosmological
access, I hoped she would feel as vulnerable as I did.
Unarmed with a more mature and resilient sense of “self”, I found
myself corroded by the daily onslaught. My most poignantly immature
jealousies arose when neighbours said Cheshma was unabashedly
discussing the sexual details of her married life with Wadood. Because her
behaviour was contrary to local norms, I felt safe saying she had no sense
of “shame” (besharam), and described her more lascivious accounts as
utterly inappropriate for a “rejected” woman.2 I then denied Wadood’s
daughters access to our house. Publicly, the reason I gave was that I didn’t
want them used to spy on us, but privately, I wanted to disconnect myself
from the intimacies of Wadood’s past. My greatest anger was sparked by
her relatives’ assertions that our second son’s stillbirth was somehow
divine retribution for Wadood’s sinful (gunah) marriage to me. Because
Cheshma reportedly took credit for putting a spell against the baby, I
called her a “churriyl” (witch), her behaviour idolatrous (shirk) and
invoked Prophetic admonitions to strike at “magic-makers”. As the mother
of two living boys, I then re-emphasised my ability to provide Wadood a
son. Entangled in a peculiarly intimate rivalry concerning Wadood, marital
rights, conjugality, and Cheshma’s Islamic rights to access and contact my
husband, my gossipy retorts began to destroy the qualities I fantasised
distinguished me from her: propriety, education, gentility. Just as
awkwardly, buttressed by the ethical values provided by years of university,
I retained a persistent academic positioning, uncomfortably aware of my
regrettable choices or deliberately choosing to do the wrong thing. Yet my
fears and angers formed an unquestionably productive undercurrent to my
research; “witnessing” and my local-level involvements were sometimes one
and the same, or the boundaries distinguishing these acts were blurred.

Entanglements as evidence?
George E. Marcus argued that standard ethnographic confessions are
“inadequate [because although] experiential, [they are] not properly
conceptualised” (Marcus 1994, 569 in Aunger 1995, 102, footnote). So,
could my more problematic subjectivities be rescued by their evidentiary
146 Emma Varley

potential? Paraphrasing Dorinne Kondo, the evidentiary weight of these

moments is upheld by her contention that “the specificity [of our]
experience … is not opposed to theory; it enacts and embodies theory”
(Kondo 1990, cited in Narayan 1993, 681, emphasis added). Kondo’s
position allows me to re-envision my own struggles as theoretically relevant,
where the knowledge I gained from navigating those tumultuous days
ensured far-reaching and genuinely productive analysis. In important ways,
the discoveries I made over the last six months of my research provided a
counter-balance to earlier fieldwork interviews, where men and women
plied me with unrealistically happy, “official accounts” of family integrity
and honour. As such, my story necessitated that I analytically trouble
people’s everyday stories of divorce, custody rights, women’s propriety and
sisterly affiliation, the supposed disconnections between ex-spouses, and the
deep fascination and ambivalence Gilgitis held for marital love.
Without my disputes with Cheshma, it would have been virtually
impossible to gain so many cascading waves of insight into how both
everyday quarrels and exceptional antagonisms intersected with women’s
reproductive health. My navigations of enmity were mirrored by women’s
stories of being purposely held back from hospitals or doctors’ offices, or
of being attacked by their in-laws, extended family, or jealous co-wives.
Given my initial focus on Sunni women’s access to health services, it
became fundamentally important for me to re-investigate women’s claims
that Sunni family and neighbourhood discord posed greater challenges to
their reproductive health than the dangers and logistical constraints
associated with the Shia-Sunni hostilities. Despite evidence that poverty,
nutritional deficiencies, medical mismanagement, or sectarian constraints
were primary factors underlying maternal illness or deaths, and with
Gilgit’s maternal morbidity and mortality ratios among the highest in
Pakistan,3 my participants argued that in-community enmities regularly
precipitated denied access to care, or that physical abuse, poisonings, or
malevolent “magical” intercession caused most of their health complaints.
Cosmological hazards entailed women’s uptake of home- or mosque-based
interventions or “alternative” therapeutic systems. Unlike biomedicine,
Islamic prayers (dua, dhum) and Prophetic medicine, mountain herbals
(desi bilehn), along with Islamic and indigenous sorcery practices
(As’Sihr, jadu), integrated Gilgiti notions of sociality and directly
redressed a wide array of physical and cosmological harms or risks. The
high costs of these therapies, which could far outweigh those of biomedical
services or drugs, poignantly demonstrated women’s prioritisation of
interpersonal discord and spiritual harms.
Enmities and Introspection 147

Now that I knew where to look, I found women’s stories of antagonism

and vulnerability were powerfully upheld by more “objective” accounts,
such as physicians’ accounts of their obstetric patient’s obvious use of
amulets (tawiz) , or women’s somatisation of the effects of social discord.
Backaches (derh shilahn), headaches (shish shilahn), uterine “weakness”
(bachitani kamzoree hen), and menstrual irregularities were common,
body-anchored points for women’s expression of unhappiness or
emotional turmoil. And during my research into women’s use of
abortifacients, I discovered that some pharmaceutical dispensers willingly
concocted poisons (zeher) for paying clients; these were used to lace their
unsuspecting victim’s food or drinks. Seen in this way, the local Sunni
community’s extraordinarily high maternal morbidity and mortality ratios
reflected conflictive sociality not only between sectarian communities, but
also within them.
Using my own reproductive and family difficulties as a segueway, I
was entitled to ask about women’s otherwise concealed deployment of
“black magic” to destroy an opponent’s fertility, happiness, sexual
compatibility with her husband, beauty, and peace-of-mind. On one level,
women’s relationships were truly harmonious and congenial, and I
benefited greatly from my neighbours’ empathies, many generosities, and
compassion. On another level, these relationships could be riddled with
insecurity and competitiveness that in women’s later years tended to
culminate in chronically antagonistic relationships. Domestic clashes were
the most protracted and multi-faceted, driven by the tensions accompanying
arranged marriages between unwilling partners, young brides working
under the supervision of a dictatorial mother-in-law, or the spatial frictions
of overcrowded households. I soon learned that extended family breakups
were often a product of the animosity of sisters-in-law. Of equal
importance, my ordeals highlighted the very different nature of men and
women’s confrontations. Wadood joked that men dispensed with their
angers with their fists or hard agricultural labour. Living socially
compressed lives and physically encircled by high-walled gardens and
family compounds, women nurtured revenge projects for years.4
Through their quietly conducted battles, women achieved forceful, far-
reaching domestic agency. Where men were symbolically, politically, and
materially privileged by sectarian fighting, they became sidelined and
disempowered by women’s everyday fights. With men’s happy
acquiescence, women subtly dominated the household, acting quickly and
covertly to reinforce their assets. But local antipathies toward their public
confrontation of rivals (in-laws, co-wives, neighbours) or enemies
(dushman) required that women resort to Gilgit’s many cosmologies—
148 Emma Varley

Islamic and indigenous—to indirectly redress emotional, economic, social,

and material imbalances. Women’s caustic jealousies were answered and
resolved by clandestine vengeances. Because Gilgiti women’s social and
familial importance is inextricably dependent on marriage and
childbearing, and with their bodies already over-determined by Islamic
honour (izzat) concepts, women’s reproductive ability provided a strategic
target for symbolic vengeances or direct interference by their family and
neighbours. What better way to cripple your enemy than through sexual
dysfunction, infertility, the birth of daughters, or persistent menstrual
difficulties—in other words, her “womanhood”? In this way, punitive
measures and retributive acts clearly demonstrated how Gilgiti women are
socially constituted through their bodies and reproduction.
Aside from Islamic and “magical” resort, Sunni women also used
biomedical drugs for enmity’s “symptoms”, including those associated
with “black magic” (kala jadu). Indeed, women prioritised the somatic or
physiological symptoms they felt resulted from “magical” interference
ahead of the dangers stemming from 2005’s hostilities. After ten months
of fieldwork interviews, it was clear that as much as I pushed women to
discuss how sectarian tensions impacted Sunni patient-Shia physician
encounters, or women’s access to health centres in Shia neighbourhoods,
women pushed back to prioritise how the “black magic” born of
interpersonal discord presented larger threats, more pervasive and
enduring health challenges. And in ways that impacted their overall
morbidity, women often read pre-eclampsia or anaemia as symptomatic of
“magic”, and sought care first from Islamic clergy, herbalists, or spirit
diviners. Self-medicating against the anxieties and unhappiness
accompanying relational turbulence was also quite normal. Besides
pharmaceutical anti-anxiety and pain medications, such as Valium and
Buscopan, women also used Gilgiti herbals in their efforts to relieve their
emotional tensions (aram; “relax”). Sensing my disquiets after yet another
neighbourhood spat with one of Cheshma’s many sisters, one auntie began
plying me with a local beverage. Pleased by its calming effects, I asked to
see the raw ingredients, and was not entirely surprised to realise it was a
weak opiate tea brewed from mountain poppies.
When I was especially weary, I found group interview sessions carried
a therapeutic quality. Women’s gup shup (conversations) provided helpful
opportunities to flush out angers and process the meanings and
motivations behind rivalries or hatred. My afternoon visits to women
neighbours, ostensibly to comb through our families’ happenings over
several cups of tea, illuminated the infectious qualities shared by gossip
(ghaiebat) and “black magic” (kala jadu). Listening to gossip, or partaking
Enmities and Introspection 149

in it, invoked the same types of cosmological vulnerability accompanying

“magic”. “If you believe in it, it will happen to you!” Or, “If you listen to
your enemies’ gossip about you, you’ll fall into the hole that they have dug
for you!” It was nearly impossible to extricate myself from the vigorous,
ongoing mindfulness and self-surveillance necessitated by “magic” and
gossip, and which paralleled in many ways the routine self-monitoring
required by pardah (veiling, gender segregation) and other practices of
Islamic comportment such as prayer (namaz) and ritual purification
(wudhu). Moreover, as was the case for our everyday negotiation of
gossip, the local importance of “magic” required that we think through, re-
imagine, and predict our family and community relationships from our
enemies’ standpoints. Equally importantly, by discussing Cheshma’s
antipathies with me and mine with her, our fireside chats concerning
community and family-level disputes permitted us to contrast the “what is”
with “what should be”. Our assessments of sectarian violence and
women’s animosities relied heavily on inferred contrasts between the
conditions and characteristics of peace or conflict, social balance or
Women’s fights, like Gilgit’s sectarian skirmishes, were weighted with
a locally grounded sense of religiosity, Sunni identity, “fair” social play,
appropriate versus inappropriate revenge (badal) or response. In important
ways, women’s clashes and “black magic” involved the alienation,
dislocation, and forcible disintegration of “normative” Islamic concepts of
identity, boundaries, and belonging. The disruptive nature of women’s
magical acts also, however inadvertently, contradicted the Sunni
community’s responses to the Shia-Sunni “tensions”, which emphasised
community reintegration, and the enhancement, affiliation, and demarcation
of ideological, identity, community, and inter-sectarian boundaries.
Related insights were soon to follow; for the Shias, my predicaments
demonstrated the callous disregard of Sunnis for each other and upheld the
Shias’ tacit, sectarian prejudices and cultural stereotypes. In turn, my
acceptance of my Shia neighbour’s sympathies at a time when our Sunni
family had turned on us was indicative of how corrupted my supposedly
Muslim sentiments really were. A “real Sunni” would never acknowledge
how much a fellow Sunni had harmed then, especially not to Shias.
Through my quiet admissions to Wadood’s Shia relatives that Sunnis had
“wronged me”, I weakened the tactical front held by Gilgiti Sunnis at a
time when community unity was viewed as paramount.
During the worst of my “tension times”, I firmly subscribed to self-
quietening, and withheld my troubles from both my family and my
supervisor. But the more I learned, the more I felt these stories deserved a
150 Emma Varley

legitimate share of my ethnographic attentions. Like Camilla Gibb, who

mused on the centrality of manic miseries and erotic entanglements for her
fieldwork and their presumed unsuitability for her doctoral thesis, I found
it increasingly difficult to adopt the academic imperative, whereby we
“maintain the posture of ethnographic success, project the picture of
anthropological health” (Gibb 2005, 222). In my correspondence and early
thesis drafts, I debated whether to continue disguising the origins of my
insights. But because my story was so thoroughly interwoven with the
larger issues I had explored with my participants, separating critical
analysis from personal narratives proved to be taxing and ultimately
impractical. And because many of my fights had evidentiary potential, I
had to seriously rethink the value of sidelining the ethically ambiguous
events that motivated some of my more effective analytical leaps. My
prejudices and my antagonisms remained an irrepressible component of
my work’s focus, the questions I asked, and the narratives I pursued. More
unexpectedly, I discovered that sublimating germane and often-traumatic
fieldwork events entailed emotionally depleting practices of self-
suppression, as well as ongoing vigilance concerning what I shared with
whom, and its academic, ethical, or legal implications.
Even after agreeing to write about the wider scope of my Gilgiti
fieldwork, I have felt anxious about how exactly to write to a discipline
where our expositions are usually legitimated by precedent. In my search
to find a way to include the personal as evidence, Paul Stoller’s article
Ethnographies as texts/ethnographers as Griots (1994) was especially
beneficial. Stoller, too, struggled to co-align his more awkward
involvements with anthropological theorising and standard ethnographic
writing. Where Stoller paid poetic homage to his practice of West African
sorcery through “reconstructed dialogue” and a novel-based approach
(1994, 359), I have attempted instead to emulate what he calls the
authorial “middle-voice” (ibid., 356). Stoller draws on Roland Barthes
(1977) to argue for reflexivity to honestly appraise our positionality and
encompass the socio-economic, political, and historical forces surrounding
our fieldsites and analytical choices (Stoller 1994, 355). When Stoller’s
contextually attuned reflexivity is coupled with Kondo’s sense of
experience as “enacting” theory, self-reflection more effectively dovetails
with cultural analysis, while also guarding against the peripheral or
undirected self-indulgencies most heavily debated by critics of
Broadly inclusive approaches come with a caveat or two, and require
careful handling. Using my own story, for example, risks destabilising my
professionalism. Or, I might be interpreted as yet another postmodernist,
Enmities and Introspection 151

“life-damaged self-revealer” (Geertz 1988, 92). In his essay Excessive

witnessing: The ethical as temptation, Basque separatist and anthropologist
Joseba Zulaika cites Paul DeMan to caution us that “self-accusing [can
turn] into a mode of self-excusing which, in terms of truth, ‘ruins the
seriousness of any confessional discourse by making it self-destructive’”
(DeMan 1979, 280, cited in Zulaika 2003, 97). More hopefully, adopting a
more inclusive position affords a precarious intimacy, whereby even for a
few hopefully constructive moments, we are able to glimpse into, share, or
place our self “where the author situated himself” (Bourdieu 2001, 94).
Encouragingly, for those anthropologists who invoke deeper layers of
opinion or feeling, their use of reflexivity is distinguished by a beguiling
emotional immediacy (see Lambek 1997).

Reflecting on his fieldwork, where by his own admission he coerced
and played “various informants against one another in order to uncover
some village conflicts they didn’t want uncovered” (Geertz 1988, 97),
Rabinow stated:

To those who claim that some form of symbolic violence was not a part of
their field experience, I simply reply that I do not believe them. It is
inherent in the structure of the situation. (Rabinow 1977, 129-130, cited in
Geertz 1988, 98)

In some ways, my fieldwork was similar to other kinds of ethnographer-

perpetrated “violences”, which include research-participant power
imbalances. Like many anthropologists, for the duration of my fieldwork I
was often trapped in social situations that were beyond my effective
control. But in other ways, my entanglements were markedly different
from those of most ethnographers. Rather than being the stereotypically
alienated researcher (Gibb 2005, 225)—on the outside looking in—I
worked and lived at the hub of oscillating webs of ego-centred,
humiliating, and emotionally ruinous gossip and rumour. But instead of
detracting from my fieldwork, my troubles enriched it. Rather than simply
unearthing unproductive or outlandish vignettes of fieldwork gone awry,
my experiences demonstrate how through “self”, ethnographers might, if
only sometimes, authentically achieve perspectival insights into the
“other”. Or, how our mistakes and social failings act as evidence and drive
theoretical inquiry. If seen in this way, my story might work as a qualified
version of what Geertz called “ethnographic ventriloquism: the claim to
152 Emma Varley

speak not just about another form of life but to speak from within it”
(1988, 135).
Conceptualising my fieldwork fallibilities (whether inadvertent
weaknesses or disingenuous foibles, resolved or unresolved) as evidence
permits my re-exploration of even contentious aspects of “self”, such as
hatred or cultural myopia, as interpretive platforms. In light of my
neighbourhood fights and conflict-borne sectarian sentiments, I agree with
Christopher Long, who states that “despite the modern prejudice with
prejudice, there is no way to separate ourselves from the prejudgments that
condition us and through which understanding first becomes possible”
(2004, 12). Similarly, Martha Nussbaum argues that there is “no receptive
‘innocent’ eye in perception”, and that the imaginative component of
interpretation remains inextricably “bound up with [our] past, prejudices,
needs” (1985, 257). In order to resolve more fractious or morally
problematic aspects of interpretation, Long advises that the “only way to
ensure enabling prejudices do not calcify into blind prejudices” (2004, 12)
is to enter into dialogue with prejudice itself. By conscientiously
reconfiguring the acceptability and applicability of certain kinds of
disclosure, we might augment “regular” reflexivity, which according to
Michael Lambek often “mystifies the role played by prejudice” in forging
understanding (1997, 32). In light of my own trials, I suggest that the
ethnographer’s private “errors” and fieldwork disquiets can carry
considerable consequences for our knowledge generation and context-
dependent judgment. To this end, enmity’s emotional repercussions
sustain my interest in uncovering reflexivity’s “other” side, rather than
only upholding its ethical optimism.
At a primary level, my account represents those truths that resonated
both for my participants and me. Just as importantly, it stands alone and
apart as evidence of a new kind of post-colonial Gilgiti sociality. My
fieldwork came at a time when tourism, intrepid mountaineers, and a
regional explosion of Saudi-funded proselytisation movements, Shia
militants, and transnational non-governmental organisations had infused
Gilgiti socio-economic landscapes with eclectic elements of the world
“out-there”. It was in these spaces that I first discovered, and subsequently
re-applied, Gilgiti discursive and symbolic categories to understand and
make my own experience understood. And in the same ways that even our
fieldnotes require us to grapple with ego (Ottenberg 1990,114), our in-text
disclosures must be carefully combed through, tested against our personal
comfort levels, and crafted with a simultaneous intent for honesty,
readability, and productive applicability. Because reflexivity is a generally
unwieldy and unpredictable process, it is largely resistant to rigorous
Enmities and Introspection 153

protocols for deciding on either its inclusion or exclusion. But by

foregrounding the “local”, evidentiary weight of our fieldwork
problematics, we might avoid undue emphasis on our ethnographic
“selves”, which risks insufficient analysis of the peoples and places we
study. I hope to have shown that reflexive accounts of particularly
troubling fieldwork stories, when evaluated by their contextual relevance,
carry evidentiary merit. In the context of my own research, this was
achieved by their ability to reflect women’s everyday realities, and to
productively counter the ways women—and anthropologists—choose to
think about or represent themselves.

Without the love and encouragement of my husband Wadood, children
Kate, Nadeem, Imran and now Sofia, this chapter would not have been
possible. I also wish to thank my parents, Deborah and Christopher, and
Elizabeth Varley and Orian Hutton for their support. In Gilgit, I wish to
thank my research assistant Mrs Fazeelat Asif, my research participants,
and the family of Feroz and Hourima. I am also grateful for the support of
Dr Saba Gul Khattak and Ms Gulistan Ibadat. At the University of
Toronto, I wish to acknowledge the guidance of my supervisor, Dr
Michael J. Lambek, my co-supervisor Dr Janice Boddy and Committee
members Dr Bonnie McElhinny and Dr Richard Lee. My fieldwork was
funded by a Doctoral Fellowship with the Social Sciences and Humanities
Research Council (SSHRC), a Doctoral Research Award from the
International Development Research Centre (IDRC), a Lorna Marshall
Doctoral Fellowship in Social and Cultural Anthropology from the
University of Toronto, and support from Dr Lambek’s SSHRC Research
Programme on Medicine and Citizenship. My post-fieldwork research has
been funded by Doctoral Fellowships from the Ontario Women’s Health
Scholars Award (Ontario Council of Graduate Studies) and Health Care,
Technology and Place (HCTP, University of Toronto). The conference
presentation which precedes this chapter was supported through a Canada
Research Chair in the Anthropology of Ethical Life, held by Dr Lambek.
Finally, I would like to express my gratitude for Timm Lau’s meticulously
attentive editorial support, and the additional writing assistance of Liana
Chua and Casey High.
154 Emma Varley

This does not mean that reflexive modes have not already been thoroughly
problematised by a number of ethnographers. Following from the anthropologist-
power inequities suggested by Rabinow’s “symbolic violence” (1977), Clifford
Geertz describes the ethnographer’s textually-anchored “Self” as being akin to the
“unsolid ground upon which all [fieldwork] interaction inevitably rests, a tissue of
careerism, deception, manipulation, and micro-imperialism” (Geertz 1988, 95).
Many of my women neighbours encouraged me to vent my anger. It affirmed that
I was equally vulnerable to local forces and provided them a chance to reveal the
otherwise-concealed rages they held against their co-wives or their husbands’
former wives. Our emotional similarities were often reassuring to women, who had
initially imagined that our education and financial differences meant I was unable
to understand or commiserate with their everyday pains.
The 1999 Northern Areas Health Project’s baseline survey estimated the local
MMR (maternal morbidity and mortality ratio) as five hundred per hundred
thousand (Karim 2004, 11).
Gilgiti women, unlike men, were largely unable to call direct attention to
themselves and their needs without risking criticisms that they were “selfish” or
harmfully destabilising traditional emphasis on self-sacrifice, and women’s
prioritisation of family and honour (izzat) over “self”. Instead, Gilgiti women
obliquely expressed individual need through narratives concerning their social
“others”. By emphasising their contributions as mothers or wives, their
victimisation by “enemies”, or their sufferings due to health complaints, women
indirectly sought resolution and social attention for their individual needs or

Abu-Lughod, Lila. 1988. Fieldwork of a dutiful daughter. In Arab women
in the field: Studying your own society, ed. Soraya Altoki and Camilla
Fawzi El-Sohl. Syracuse: Syracuse University Press.
Aunger, Robert. 1995. On ethnography: Storytelling or science? Current
Anthropology 36 (1): 97-130.
Barthes, Roland. 1977. Image-music-text, trans. Stephen Heath. New
York: Hill and Wang.
Bourdieu, Pierre. 1977. Outline of a theory of practice. Cambridge:
Cambridge University Press.
—. 2001. Science of science and reflexivity. Chicago: University of
Chicago Press.
Enmities and Introspection 155

Bourgois, Phillipe. 1990. Confronting anthropological ethics:

Ethnographic lessons from Central America. Journal of Peace
Research 27 (1): 43-54.
D’Andrade, Roy. 1995. Moral models in anthropology. Current
Anthropology 36 (3): 399-408.
DeMan, Paul. 1979. Allegories of reading. New Haven: Yale University
Flyvbjerg, Bent. 2001. Making social science matter: Why social inquiry
fails and how it can succeed again. Cambridge: Cambridge University
Geertz, Clifford. 1988. Works and lives: The anthropologist as author.
Stanford, CA: Stanford University Press.
Gibb, Camilla. 2005. An anthropologist undone. In Auto-ethnographies:
An anthropology of ethnographic practices, ed. Anne Meneley and
Donna J. Young. Peterborough: Broadview Press.
Honigmann, John J. 1976. The personal approach in cultural
anthropological research. Current Anthropology 17 (2): 243-261.
Jenkins, Timothy. 1994. Fieldwork and the perception of everyday life.
Man, n.s., 29 (2): 433-455.
Karim, Nayat. 2004. Northern health program project–FPAP component:
“Family planning on the roof of the world” northern areas, Gilgit.
Gilgit: Family Planning Association of Pakistan.
Kondo, Dorinne. 1990. Crafting selves: Power, gender and discourses of
identity in a Japanese workplace. Chicago: University of Chicago
Kulick, Don. 1998. Travesti: Sex, gender and culture among Brazilian
transgendered prostitutes. Chicago: University of Chicago Press.
Lambek, Michael J. 1997. Pinching the crocodile’s tongue: Affinity and
the anxieties of influence in fieldwork. Anthropology and Humanism
22 (1): 31-53.
Long, Christopher. 2004. The ethics of ontology: Rethinking an
Aristotelian legacy. Albany: State University of New York Press.
Marcus, George E. 1994. What comes [just] after “post”? The case of
ethnography. In Handbook of Qualitative Research, ed. Norman K.
Denzin and Yvonne S. Lincoln. London: Sage.
Meneley, Anne, and Donna J. Young. 2005. Introduction to Auto-
ethnographies: The anthropology of academic practices, ed. Anne
Meneley and Donna J. Young. Peterborough: Broadview Press.
Narayan, Kirin. 1993. How native is a “native” anthropologist? American
Anthropologist 95 (3): 671-686.
156 Emma Varley

Nussbaum, Martha C. 1985. Aristotle’s De Motu Animalium. Princeton:

Princeton University Press.
Ottenberg, Simon. 1990. Thirty years of fieldnotes: Changing relationships
to the text. In Fieldnotes: The makings of anthropology, ed. Roger
Sanjek. London: Cornell University Press.
Quayson, Ato. 2003. Calibrations: Reading for the social. Minneapolis:
University of Minnesota Press.
Rabinow, Paul. 1977. Reflections on fieldwork in Morocco. Berkeley:
University of California Press.
Rapp, Rayna. 1999. Testing women, testing the fetus: The social impact of
amniocentesis in America. London: Routledge.
Rosaldo, Renato. 1996. Grief and the headhunter’s rage. In
Anthropological theory: An introductory history, ed. R. Jon McGee
and Richard L. Warms. London: Mayfield Publishing Company.
Scheper-Hughes, Nancy. 1992. Death without weeping: The violence of
everyday life in Brazil. Berkeley: University of California Press.
Scheper-Hughes, Nancy. 1995. The primacy of the ethical: Propositions
for a militant anthropology. Current Anthropology 36 (3): 409-440.
Stoller, Paul. 1994. Ethnographies as texts/ethnographers as Griots.
American Ethnologist 21 (2): 353-366.
Strathern, Marilyn. 1987. Out of context: The persuasive fictions of
anthropology. Current Anthropology 28 (3): 251-281.
Wardlow, Holly. 2006. Wayward women: Sexuality and agency in a New
Guinea society. Berkeley: University of California Press.
Van Maanen, John. 1988. Tales of the field: On writing ethnography.
Chicago: University of Chicago Press.
Young, Donna J. 1996. Remembering trouble: Three lives, three stories. In
Cultural essays in trauma and memory, ed. Paul Antze and Michael
Lambek. New York: Routledge.
—. 2005. Writing against the native point of view. In Auto-ethnographies:
The anthropology of academic practices, ed. Anne Meneley and
Donna Young. Peterborough: Broadview Press.
Zulaika, Joseba. 2003. Excessive witnessing: The ethical as temptation. In
Witness and memory: The discourse of trauma, ed. Ana Douglass and
Thomas A. Vogler. London: Routledge Press.
—. 2004. The anthropologist as terrorist. In Violence in war and peace: An
anthology, ed. Nancy Scheper-Hughes and Phillipe Bourgois. Oxford:


This chapter examines diasporic Tibetans’ emotional experience and
conceptualisation of shame, highlighting their role in the construction and
reconstruction of selves and sociality. The analytical starting point for this
exploration lies in my own emotional experience during a key episode in
my fieldwork in a Tibetan settlement in Northern India. It was caused in
the first place by my culturally specific disposition, through which I
valued interfering in a violent fight differently from my Tibetan hosts, thus
creating anger and confusion (see Tonkin 2005). Yet, this experience led
to valuable insights into the role of emotions in local social hierarchies,
sensitising me to the importance of hierarchical boundaries and the gaze of
others in my informants’ lives.
A number of authors have noted the importance of the ethnographer’s
emotional experience in yielding analytical insights. In his famous
discussion of his own experience of profound grief in the field and the
reactions of his informants, Renato Rosaldo (1984) described the insights
he gained into both; Unni Wikan (1992) has raised the issue of resonance
and its importance in ethnographic understanding; John Leavitt (1996)
attempted a synthesis of how emotional experience may aid an
anthropology of the emotions; and Lisette Josephides (2005) has recently
placed her investigation of the ethnographer’s self in fieldwork within a
discussion of resentment as a sense of self. The present analysis adds to
this discussion by demonstrating that anthropological fieldwork allows for
the inclusion of emotions in the ethnographer’s experience as important
contributors to our understanding of local social dynamics, even if the
emotions differ according to cultural dispositions. Hence, I do not claim
that my emotional experience was the same as my informants’, but rather
that it functioned as a heuristic device which enabled me to see and
158 Timm Lau

connect pieces of ethnographic evidence in new ways which may not have
become visible otherwise.
In overview, I will first outline the importance of social hierarchy both
in Tibetan political processes in exile and in the anthropological
description of Tibetan social organisation and experience more generally. I
will show that the Tibetan notions of shame, language, and respectful
behaviour are indicators of the presence and importance of hierarchical
levels in Tibetan society. I then illustrate my own sensitisation towards the
nexus of Tibetan shame and hierarchical sociality in a key ethnographic
episode in which I crossed hierarchical boundaries. My following
ethnography of socialisation, public gatherings, and the constraints on
romantic relationships inside a Tibetan refugee settlement shows that
shame is active in shaping subjective experience and the local construction
of social hierarchy. I argue that shame regulates behaviour according to
the visibility or “publicness” of Tibetan selves.1 Public situations are
central to the operation of social status, and shame demarcates what
accords to one’s status from what does not, in terms of behaviour or even
mere visibility. My analysis suggests that the processes by which Tibetan
social actors draw upon ngo tsha in social interaction at the same time
reproduce the structural features of their social hierarchy.

Tibetan exile politics and social hierarchy

Since the beginning of their flight from Tibet in the late 1950s,
Tibetans have joined numerous groups of displaced people living in India.
When the fourteenth Dalai Lama, political and religious leader of the
Tibetans, fled to India in 1959, he immediately established a Tibetan
administration in exile for the approximately eighty thousand Tibetans
who followed him. His establishment of the Central Tibetan Administration
(CTA), often called the “Tibetan government-in-exile”, was aided by the
fact that members of the Tibetan cabinet (bka’ shag) and other important
political and administrative figures had also fled to India. Despite their
sudden uprooting from Tibet, the Tibetans in India were therefore able to
found important, state-like administrative structures under the leadership
of the Dalai Lama (see Ardley 2003).
From its inception, the Dalai Lama’s administration advanced the
adoption of democratic elements in political processes, such as the election
of members of the legislature. Jane Ardley has pointed out that this
process of “democratisation” does not entail an unproblematic transition to
democracy as it is commonly understood in Western contexts, as both the
Tibetan leadership in exile and its Tibetan constituency maintain ideas and
Understanding Tibetan Shame and Hierarchy 159

practices that do not correspond with the latter. Tibetans, for example,
widely demurred at the inclusion into their charter of elements which
limited the personal powers of their unelected leader (Ardley 2003, 353-
54); the CTA itself does not allow the free formation of political parties,
preferring to organise political elections of individuals in constituencies
that are defined by regional origin in Tibet (ibid., 350, 355). In a recent
study of Tibetans in Nepal, Ann Frechette has used the contrasting terms
of “hierarchy” and “equality” in her discussion of conflicts between the
hierarchical framework of political organisation maintained in Tibet before
the flight, and norms and values connected to more recent democratic
reforms in the diaspora (Frechette 2004, 76). She also points out that
hierarchical notions of aristocratic leadership remain salient among
Tibetans in the diaspora (ibid., 77-78). Together with my own research,
these studies demonstrate that specifically Tibetan attitudes to hierarchical
relations are reflected in Tibetan exile politics today.
For thousands of Tibetan refugees in the early years after arrival in
India, the foundation of the CTA did not immediately ease the burden of
their difficult livelihoods. Up to a quarter earned their livelihoods as road
labourers under harsh conditions in the northernmost Indian states (Kharat
2003, 288), before taking up petty trade in sweaters during the winter
season (Lau 2007, chapter 4). Many Tibetan refugee settlements were also
not established until some years later. The Tibetan settlements near the
Indian village of Bir in Kangra District, Himachal Pradesh, for example,
were constructed in the early 1960s. They were built independently of the
Dalai Lama’s administration by three different groups of Tibetan refugees
whose leaders hailed from the eastern region of Kham. The three
settlements are politically distinct units with their own administrative
structures, but spatially at such close quarters that it is impossible for the
uninitiated to tell where one ends and another begins. Each of the three
groups retains a distinct political identity today, although the
administration of the settlements and their approximately two thousand
inhabitants has since been surrendered to the CTA. Their distinct identities
are in part based on regions of origin in Tibet, which are reflected in the
names of two of the three adjacent settlements.2 But the idiom of regional
identity glosses over an existing diversity in the make-up of the
settlements’ divisions, in which Tibetan families with origins in different
parts of Tibet also live side by side.
Elsewhere, I argue that political processes shaped by specifically
Tibetan cultural dynamics were central to the way in which the Bir Tibetan
settlements have been formed (Lau 2007, chapter 2). I demonstrate that
besides the significance of regional identities, notions and practices of
160 Timm Lau

social hierarchy in Tibetan government influenced the formation of the

settlements and their social groups. Key aspects of characteristically
Tibetan socio-political processes, explicated by Cassinelli and Ekvall
(1969), were reproduced in the foundation of the settlements. These
authors use the term “allegiance” to describe both a political relationship
between the hierarchical positions of governor and governed, and the
“proper attitude” of the subjects toward their governors, which importantly
included the responsibility of compliance with their directives. The
formation of new local identities in Tibetan settlements, based as it were
on place and territory, was therefore also enabled by prevalent hierarchical
forms of leadership. These are still important in the settlements today, for
example in the organisation of local settlement elections. Although these
elections are not based on regional identities in the same way as the
elections of the Tibetan legislature in India, they nevertheless entail a
strong tendency to refer to elements of Tibetan regionality. Tibetans
approach local elections with a logic of regionalism, illustrated in
statements such as: “my father is U-Tsang so I have to vote for the U-
Tsang (candidate)”, given to me on one occasion by a young informant.
Carole McGranahan makes the important observation that such allegiance
to regional identity implies allegiance to specific leaders, when she points
out that exiled Tibetans continue to elect their representatives “on the basis
of regional and sectarian affiliations”:

[D]espite homogenizing and hegemonizing efforts, regional affiliations and

allegiances to both lay and religious district leaders, such as village chiefs
and lamas, retain both practical and symbolic importance in exile.
(McGranahan 2005, 573)

In connecting the election modalities based on regional identities with

allegiances, McGranahan implicitly makes an important point: Tibetan
elections are inherently connected to hierarchical relationships based on
status. The allegiances expressed in the election process are often based on
the inherited hierarchical status of the leadership figures, to whom Tibetan
refugees entered into allegiance in Cassinelli and Ekvall’s sense of the
On closer inspection, the extant literature on Tibetan societies more
generally prefigures these research results on post-flight Tibetan politics.
The emerging picture, from early missionary and British colonial literature
on Tibet (e.g. Bell 1928; Fillipi 1937) to later academic publications which
attempted holistic descriptions of its society and history (e.g. Goldstein
1971; Richardson 1984; Shakabpa 1967; Snellgrove and Richardson 1968;
Stein 1972), is one of social stratification based on hereditary aristocratic
Understanding Tibetan Shame and Hierarchy 161

and monastic rank. This basic hierarchical structure is confirmed by

studies of Tibetan communities and work focused on socio-political
structure in Tibetan society (e.g. Aziz 1978; French 1995, 111-12; Gombo
1983). Notably, this emphasis on Tibetans’ “pervasive and unique notions
of hierarchy” (French 1995, 108) also occurs in the literature on Tibet after
Chinese annexation: Heidi Fjeld’s recent study (2005) of hereditary social
divisions in contemporary Lhasa argues for the deep-rooted and persistent
importance of Tibetan social hierarchy and hereditary family background,
in spite of the vast social changes effected by Chinese efforts at building a
socialist society.
But Tibetan hierarchy is described by anthropologists as more than
merely the outward form of Tibetan social structures: it is often portrayed
as an intrinsic part of Tibetan world view and cosmology. Toni Huber, for
example, notes that Tibetan systems of representation in general are
hierarchically ordered, systematically arraying persons on the basis of
ranked properties (Huber 1994, 360). Rebecca French claims that

[t]he Tibetan word view has a pronounced concept of hierarchy, of levels

and zones moving from the inferior to the superior and from the secular to
the sacred…Higher and lower levels appear in multifarious guises
throughout the Tibetan cosmological and social world. (French 1995, 109)

This strong characterisation of Tibetan cosmology and sociality as

hierarchically informed begs the question of how it is constructed and
experienced in social interaction. In this chapter, based on my fieldwork
with Tibetans in the diaspora in Northern India, I will explore how
hierarchy is ingrained in Tibetan sociality by discussing the Tibetan
concept of shame (ngo tsha) in relation to the perception and operation of
status. I will discuss ngo tsha as both Tibetan emotional experience and
concept, and argue that it is a central element in both the subjective
experience of Tibetan social hierarchy and its reproduction.

The Tibetan notion of shame

Ngo tsha literally translates as “hot face”, a reference to the flushing of
one’s face when feeling shamed. As Hildegard Diemberger notes for
Tibetans in Khumbo, children are not thought of as moral beings fully
responsible for their actions until they have acquired a sense of shame
(1993, 88-89). The centrality of shame as a moral-emotional concept in
Tibetan sociality has a long history, dating to at least the fifteenth century.
This is illustrated by the following excerpt from the biography of Chokyi
Dronma,3 the princess of Gungthang who lived from 1422-1455 and
162 Timm Lau

married around 1440 into the house of the rulers of southern Latoe (La
stod) in Western Central Tibet:

The princess always performed virtuous deeds. She always respected the
doctrine, the great lamas, the monastic community, (23b) her parents-in-
law and her husband like the jewel of the crown of the head. She never
spoke anything that was untrue. She enjoyed repeating good sayings as
ornaments for the ear. She never spoke any rude words and any words that
would hurt other people. Her throat wouldn't utter anything that could not
be trusted. She greatly enjoyed giving donations. She would never indulge
in wearing female ornaments such as bracelets. She would never loosen the
golden belt represented by the awareness of what is shameful and what are
the rules of proper behaviour. (Diemberger 2007, 162, emphasis added)

For the present context, what stands out in this description (dated circa
1460) is its referencing of shame as moral knowledge. Chokyi Dronma’s
“awareness of what is shameful” is described as the explicit summation of
her knowledge of proper behaviour as a perfectly well-behaved daughter-
in-law. I will argue that this awareness, and the socialisation into it, remain
essential for Tibetan sociality.
Although the concept of ngo tsha is very significant for the emotional
context of what it means to be a Tibetan person, little ethnographic
material on it has been presented. Emily Yeh notes that the self-definitions
emphasised by diasporic Tibetans in North America include the exhibition
of ngo tsha, along with the use of Tibetan language and the practice of
compassion (2002, 236). The fact that she recorded such emphasis among
Tibetans in North America also speaks for the wider significance of ngo
tsha and against the interpretation of the ethnographic material presented
here as particular to those Tibetans in my fieldsite. In the extant work on
Tibetan cultural areas in Nepal, Charlotte Hardman’s discussions of
ngesime and niwa as central notions among the Lohorung, and Philippe
Sagant’s description of the “head held high” and its counterpart of “losing
face” among the Limbus, come closest to a sustained discussion of the
subject of shame (Hardman 1981, 2000; Sagant 1996, 9-49).4 My
particular focus on the role of ngo tsha in Tibetan sociality will address
this gap in the existing literature.

Tibetan language and respectful behaviour

It is of the essence for Tibetans to understand hierarchy and to express
this understanding by behaving respectfully. Someone showing respect
and humbleness is called ya rabs, a person of good character—by contrast
Understanding Tibetan Shame and Hierarchy 163

to ma rabs, a person of bad character. The Tibetan syllables ma and ya

indicate “lower” and “higher” statuses. Therefore, the terms ma rabs and
ya rabs themselves exemplify the hierarchical nature of Tibetan social
concepts. The exhibition of ya rabs through giving respect is very
important in the presence of elders, teachers, monks, lamas, and other
persons who are in some respect one’s superior. The principle of seniority
as hierarchical superiority and thus as demanding respect is highly salient
for Tibetans, for example vis-à-vis parents and older siblings; acting in
accordance with it is thought to be very important in avoiding familial
The presence of at least three different honorific registers in the
Tibetan language serves to illustrate the importance of hierarchy for
Tibetan social forms. As Leo Howe has pointed out for Balinese, language
is a prime site for the enactment of hierarchy because it embodies a
hierarchical code, such that speaking entails making statements about
status relationships (Howe 2001, 84; see also Fjeld 2005, 103; French
1995, 109; Shakya 1994, 158). Honorific Tibetan language (zhe sa) is
normally required to express respect in verbal discourse with people of a
greater age or social, political, or religious position. Many words have a
completely different honorific form to express the high status of the agent
or addressee, such as gzigs instead of lta (to see) or mdzad instead of byed
(to do).6 Tibetans can become anxious and ashamed if they cannot speak
well enough in front of high-status persons, provoking strong emotional
reactions and shame. In the Tibetan refugee settlements in North India in
which I did fieldwork, the ability to speak honorific Tibetan is seen as an
achievement of some distinction. Many adults can read the official
publications of the CTA in Dharamsala only with great difficulty, and feel
some anxiety about their ability to behave with ya rabs in the presence of
superiors. They nonetheless praise the virtues and beauty of honorific
Tibetan, and derive enjoyment from the use of honorific and “educated”
language because it indicates cultural as well as linguistic excellence—this
point being all the more pronounced since Tibetan culture is perceived as
threatened in the diaspora.
Honorific language and ya rabs both reproduce social difference, and
the stratification inherent in both is internalised during childhood (Fjeld
2005, 134). Legitimation of hierarchy is thus further predicated on the fact
that the hierarchical and coercive nature of language is learned before it
can be questioned, and in combination with other aspects of social
interaction such as bodily comportment, gestures, and eye movements that
create “habitus” (Bourdieu 1977; Field 2005; Howe 2001, 93). However,
the existence of hierarchical levels of Tibetan speech is not only restricting
164 Timm Lau

but also enabling and rewarding for its participants: the use of higher
registers, for example, implies the understanding that one is respected in
return.7 Fjeld describes this process in an argument loosely inspired by
Mauss’s (1967) description of the gift as something that creates an
expectation of something to be returned:

When [nobles] are met with humble behaviour, they as receivers are
expected in return to recognize the giver as a “good person”, and a social
relation of exchange is established. (Fjeld 2005, 134)

In sum, the central importance of respectful behaviour and language

indicates the presence of levels of hierarchy among Tibetans. Non-
compliance with the norms governing honorific language and ya rabs is
considered to be shameful because it would have one act “out of one’s
place”, whereas having ngo tsha (shame) is thought of as an adornment,
enabling a person to act morally.

Shame and hierarchy: an ethnographic encounter

In his article Fields of Shame, the psychoanalyst Benjamin Kilborne
describes shame reactions of the ethnographer and his or her informants as
an important resource for fieldwork and ethnography. He tells us that
shame describes affective states such as feelings of painful exposure,
humiliation, and failure (1992, 245), and suggests that shame reactions
“can be used as one of the most important resources we have in
understanding our fieldwork and in providing faithful accounts of the
peoples we study” (ibid., 246). In his analysis, shame is an essential part of
the field experience and of the logic of what does and does not appear in
fieldnotes. He notes that published ethnographies do not use shame
dynamics as a way of understanding field situations, writing:

We need to know far more than we do about experiences in the field, what
portions of them are packaged in the form of ethnographies and what
portions escape descriptions, remaining secret because of a desire to hide
them out of shame. (ibid., 244)

In the following, I will show how on a specific occasion, my overstepping

of status boundaries led to an emotional experience that proved insightful
into the dynamics connecting shame and Tibetan social hierarchy.
On a warm evening in May 2004, I returned with a young monk from a
basketball tournament put on by local Tibetan youths. We exited the taxi
and walked up towards the settlement’s bazaar, essentially a small
Understanding Tibetan Shame and Hierarchy 165

esplanade facing the settlement’s only paved road. This space was a
popular meeting place and separated from street level by a number of
concrete steps. Immediately, I saw that in the street in front of the shops, a
Tibetan man and an Indian man were engaged in a vicious fight, watched
by a number of people. They wildly swung at each other with heavy iron
rods of about three feet in length taken from a nearby building site. Soon,
the Indian man was hit on the side of the head by a full blow and sank to
the ground. I spontaneously jumped between the two men and took hold of
the Tibetan man’s weapon with my hands. In hindsight, I think I felt the
need to intervene because the fight was potentially life-threatening.
Holding his weapon, I managed to speak to the Tibetan man and we
moved away from the Indian. I turned to see that the Indian man had
disappeared, probably taken away by other Indians who work in the
settlement as taxi drivers and congregate on the road below the shop-
But my sense of relief that the fight was over was very quickly
replaced with shock and bewilderment as I was being heavily scolded
from the esplanade atop the steps. A highly respected and authoritative
local figure was yelling at me very angrily in English: “What are you
doing? Why are you interfering? Mind your own business!” I was
surprised and confused about this. Why was he scolding me? Still
vulnerable from the violence I had been exposed to, my reaction was to
walk up the steps towards him. As I reached the top of the steps and
attempted to explain myself, I realised that his wrath was increasing. He
angrily confronted me, shouting at me in a threatening manner to “be
careful”. In this tense moment, I experienced a cacophonous mixture of
confusion, fright, intimidation, and embarrassment. Mercifully, the monk
with whom I had arrived led me away from the esplanade and the market
area towards my home. Trying to calm me down, he told me that I should
not have interfered, as this was how things were done around here: the two
had taken up weapons and the rest was up to them. Finally, he told me to
forget about what had happened and instead think about our next
basketball game.
Later, my emotional experience increasingly mixed with anger as I
reflected upon what had happened. I had no experience of such fights in
Bir and felt highly upset at the men involved, at the influential man who
scolded me, and even at my saviour because he had remarked about the
violent clash that “this is how we do things here”. I was effectively upset
at the whole Tibetan environment I had immersed myself in, and now
seriously questioned whether that had been such a good idea to begin with.
Yet, on a close informant’s advice, I paid a visit to the house of my
166 Timm Lau

antagonist the next morning in order to apologise and explain that I had
not meant to offend anyone. At his house, I erroneously expected
acceptance of my apologetic explanation. But I was greeted in the coldest
of manners and not even asked to enter the house, something which I had
learned enough about to understand to be disrespectful. Instead, I was
given a lecture about Tibetan manners of fighting, and the emotional
experience familiar from the night before continued.
In intervening in the fight, I had crossed a boundary into territory that
was not mine to claim. This became more apparent to me later, when it
was explained that the man who had scolded me was known to be a local
mediator of conflicts to such an extent that he bore the nickname “the
lawyer”. He was a highly respected man, in part because of his father’s
legacy as one of the founding figures of the settlements, and because he
had succeeded him as leader for a short time after his death. In intervening
in the conflict in his presence, I had not only done what no other than he
was “allowed” to do, I had also committed a further transgression when,
instead of falling silent when rebuked by him, I literally elevated myself to
his level by walking up the steps toward him. This was easily interpretable
as “shameless” behaviour on my part, and had further escalated his wrath.
This ethnographic encounter involved being chastised in the strongest
tones and a continuation of the associated experience of humiliation the
next day. Although I could not at first make much sense of it, I later
realised that the series of events had had a highly sensitising effect on me.
It was, in a sense, an introduction to the “field of shame” in my fieldwork.
The unfolding events led me to feel surrounded by strangers whom I did
not understand, and whose behaviour I found highly alienating. The
associated sense of being exposed developed further, as my actions had
become the talk of the settlement by the next day. To some, it seemed as
though I had helped the Indian, and thus supported the “wrong side”. I was
told by the owner of a tea shop and restaurant on the esplanade that many
people were saying that what I did was wrong. She added that she herself
knew that I had meant to do a good deed, but that others did not know.
This only added to my anxiety. Had I ruined my chances of doing
fieldwork in this settlement? At the time, I felt an acute anxiety that
existed alongside other feelings about the incident, and was connected to
feeling exposed to the gaze of others.
In Kilborne’s characterisation of the fieldwork situation, anthropologists
are “those who look” and their informants “those who are looked at”
(Kilborne 1992, 246). Kilborne holds that in this constellation,
anthropologists as intruders feel ashamed of the fact that they are looking,
and do not realise that the people they study are actually looking at them
Understanding Tibetan Shame and Hierarchy 167

all the time (ibid.). In my own experience of fieldwork, however, I was

very often highly conscious of the fact that I was being closely watched by
my informants, whose homes I entered as a stranger. The episode I
recounted represents an extreme occasion of this awareness of being seen.
The associated experience was intensely poly-emotional, and although I
am unable to apply a single label to it, it involved an acute sensitisation to
the way in which I was positioned socially. In the process, I was sensitised
toward the way in which Tibetans relate to exposure and to the
overstepping of hierarchical boundaries: with a sense of shame which they
call ngo tsha. With increasing fieldwork experience and growing
sensitivity, I realised that the creation of a sense of shame is by no means
an isolated or extreme occurrence, but a significant part of Tibetan

Shame as enabling sociality

As mentioned above, ngo tsha is thought by Tibetans to enable
children to become fully social beings. In my fieldwork, I observed that
the provocation of shame in children is an essential part of their
socialisation. This often takes three forms: (1) Telling the child explicitly
to be ashamed. This is done with toddlers before they fully comprehend
language, by stroking one’s cheek with a finger and saying “shame,
shame!” in English.8 (2) Telling the child that one is going to tell a relative
or friend of his or her misbehaviour. In this socialisation technique, the
person to whom one will tell is preferably someone the child is most fond
of, thus creating anxiety and often causing the child to burst into tears. (3)
In group situations, the child is also sometimes made afraid by adults
playfully threatening to do something which he or she would find terrible,
for example hitting their favourite relative with a stick. The desired effect
is to make the child cry in front of the group. We may safely infer that the
child will then immediately feel ashamed of being exposed to the gaze of
everybody present.
In schools, shame is invoked by teachers in the classroom and through
punishment. During lessons in the local Tibetan primary school, I often
observed teachers dwelling on individual students’ mistakes in front of the
assembled class, for example by playfully mimicking the error the child
had made, but sometimes also in more serious tones. One of the
punishments I was told about by young adult informants from their school
days was to stand at the wall facing the main market street. As the wall
was then merely the height of a fence, this meant exposure of one’s face
“in the corner” to all of the settlement’s busy market, a highly shameful
168 Timm Lau

position for my informants. I was told that punishment was generally

meted out in view of an audience, most prominently during the general
assembly of all school children, engendering an intense feeling of shame
in those being punished. But more playful interaction in classrooms also
incorporates shame. I observed that games played in the classroom of very
young children ended on playful humiliation of the losing team. In one
case, the boys had lost and were made to do a handstand at the wall while
the winning girls gleefully called out “donkeys, donkeys!” Thus, losing
meant that one was to be made a little ashamed of oneself.
Generally, then, shame is deeply involved in the very construction of
Tibetan persons through socialisation. The Tibetan child’s developing
experiences of shame at home and in school are closely linked with
experiences of the “public gaze”, where others are looking at the child.
Shameful situations are sought out and indeed created to induce what
Kilborne calls “feelings of painful exposure” and humiliation (1992, 245).
We need to keep in mind here that this is done out of a concern to raise a
moral person, as ngo tsha is described by Tibetans in very positive terms
as knowledge that imparts the potential of good moral behaviour.

Shame and status performance

Tibetan gatherings are a striking example of social interaction moulded
by notions of status performance and visibility. During meetings of an
official nature, such as those called by the local settlement office or
school, there is a clear embodiment of roles which are heavily imbued
with status. It is, for example, abundantly clear that speaking is highly ngo
tsha unless it is the position of the speaker to do so in public, in which
case he is performing his status. Thus, politicians, teachers, and other
officials speak in public, while the audience does not.9 Generally,
members of the audience are performing their status by being
inconspicuous, in contrast to the speakers. This is taken very seriously by
Tibetans. Once, at a large gathering of Tibetans at the inauguration
ceremony of a monastery, I was asked by a woman to take her young son
to the bathroom. On our return, she apologised for not being able to do it
herself because it would have been highly shameful (ngo tsha chen po) for
her to get up in front of all these people. Thus, mere visibility in a public
space in which being visible as an individual does not accord with one’s
status provokes shame. During meetings in the settlement, in which a
discussion with members of the audience was necessary in order to settle
the business of the day, the raising of one’s voice as an audience member
is often accompanied by great nervousness and nervous laughter of those
Understanding Tibetan Shame and Hierarchy 169

seated close to him or her. I witnessed a number of occasions when the

speaker became visibly upset while speaking and rushed out of the
assembly after finishing a heated contribution. The action of leaving in
anger underlines the fact that inconspicuousness is generally expected of
the audience, as the shame that comes with being “visible” is first
overridden, and the situation then terminated by departing altogether.
In the settlement, young Tibetan men and women in the age group of
potential sex and marriage partners seldom interact with each other in the
open. I was generally told that “too much shame” would be the
consequence of all the attention and gossip it would entail. This is in part
due to the tension between long-standing Tibetan practices of arranging
marriage and new ideas of romantic marriage among younger Tibetans
(Lau forthcoming a). But relationships were not only ngo tsha vis-à-vis
elders or the public eye, but also in front of one’s peers. Even the
appearance of being friendly with someone could provoke troublesome
talk, and young informants were extremely worried about being the
subject of rumours about romantic liaisons. These norms are gendered:
young women are talked about rather quickly, and their status as “good”
and virtuous could quickly be damaged. For young Tibetan women, giving
the impression of being a “good girl” is status-performing, while giving
the impression of having a romantic relationship works to the contrary.
Although this is especially poignant for girls and young women, ngo tsha
and romance are also connected for boys and young men, who related their
concerns of shame in this context.
As a consequence, the preferred mode of conducting these types of
relationships is either in complete secrecy, or in times spent away from the
settlement, for example when working in Delhi or during the sweater trade
season in Indian towns. In these times, the gaze of parents, family, and
neighbours was less intently focused on young people’s activities. They
thus had “more freedom” to conduct secret rendezvous and romantic
meetings. Romantic relationships, then, can only be had in private, not in
public, because they are not status-performing for young Tibetans, and
only that which performs status ought to be done publicly. Girls,
especially, are talked about in terms of being “bad” or “loose” if they do
not comply with the performance of being chaste. This ensures that a
“good” Tibetan girl moves up in the ranks of potential wives, and one who
does not perform her status as good marriage material by keeping
relationships private descends.
My ethnographic descriptions of gatherings and of relationships
between young adults of both sexes are examples of adults’ concern with
shame as a type of knowledge about what is proper—a type of knowledge
170 Timm Lau

they initially learned through socialisation in childhood. In these

examples, and in the generally important concern of Tibetans with
reputation, what others perceive of oneself takes central place. That this
effectively leads to an emotional aesthetic of shame is illustrated by an
instance in which a young man in his twenties told me about a recent
occasion when conflict with his father had brought him to tears. He
described the emotion he felt in those moments as “the same as thinking:
what are my neighbours thinking about me right now?” Because he was
arguing, the young Tibetan was confronting his father’s dominant position
in the order of Tibetan kinship relationships, as well as being confronted
with breaking social norms of harmonious relationships in the household.
His emotional reaction of shame and simultaneous orientation to the gaze
of others emerged forcefully in that moment in which he confronted (and
was confronted by) hierarchy. This brings to the fore the degree to which
shame and concern with others’ perceptions of oneself were central to
Tibetans’ emotional subjectivity. But it also locates emotions within “the
contours of power within which the person lives”, as Lutz and White
emphasise (1986, 429). Tibetan social hierarchy provides these contours,
which become experiential through the relevant emotions.
The ethnography I have presented illustrates that status-performing
activities are vital for Tibetan social forms, and that ngo tsha indicates the
boundary of public (status-performing) activities and private (non status-
performing) activities. To be more precise, ngo tsha appears especially
when what should be private appears as visible and exposed in public, the
realm of status by definition. Tibetan sociality is implicitly connected to
the regulation of behaviour according to what I have termed public and
private realms. I have shown that this process is mediated through ngo
tsha which is drawn upon by Tibetan social actors as they create everyday
social interaction, such as behaviour at meetings and interaction between
the sexes.
Additionally, Tibetans often have recourse to ngo tsha as a principle,
using it to explain principles and instances of sociality. As we have seen, it
explicitly carries the positive quality of bestowing knowledge necessary to
be a moral social being: Tibetans understand ngo tsha to be a necessary
feature of social personhood and inculcate it in their children. Thus, in
Tibetan social interaction as well as in my analysis, ngo tsha refers to an
emotional concept as well as an emotion. This answers the possible charge
against my analysis that it represents an analytical overextension of the
emotion of shame, since emotions only have limited life-spans during
which they are experienced and could thus not carry out all the work I
ascribe to them.10 As a concept, ngo tsha is not only a referent to an
Understanding Tibetan Shame and Hierarchy 171

emotion experienced by Tibetans, but a rich notion of which they avail

themselves as they construct sociality.
The present ethnography shows that shame is deeply involved in the
construction of persons, both in the sense of children’s socialisation and in
the sense of an ongoing process in which persons interacting in everyday
sociality are continually constructed. Among other things, Tibetan social
actors construct their attitudes towards status, their own position and
visibility, and what constitutes right conduct with reference to ngo tsha.
This enables them to “put themselves into place” according to an
internalised environment of attitudes. At the same time, for example in
delineating status positions and apposite behaviour during meetings, social
forms are recreated in this process. In Giddens’ terms, Tibetan ngo tsha is
involved in the duality of structure in the constitution of interaction, in
which it is

drawn upon by actors in the production of interaction, but at the same time
[is] the medi[um] of the reproduction of the structural components of
systems of interaction. (Giddens 1979, 81)

However, this description is not an all-embracing explanation of

Tibetan sociality, nor is it intended to be. There are well-known Tibetan
cultural institutions pointing at the inversion or transgression of
hierarchical boundaries which I have described as being tied to ngo tsha.
One of these institutions is the immensely popular Tibetan folk-figure Aku
Tonpa (a khu ston pa), a sage who purposely wears a mask of stupidity, all
the while playing ribald pranks on other lay people and monks that are
lessons in the true nature of human existence. The figure of Aku Tonpa is
similar to another Tibetan cultural institution, that of the saintly madman
described by Ardussi and Epstein (1978). He displays a rejection of
customary behaviour and “a disregard for the niceties of interpersonal
behaviour, particularly with regard to social status, modes of address,
deferential behaviour, and so forth” (ibid., 332). The saintly madman puts
into practice the Tantric insight that the dichotomy between noumenal
truth and relative truth is a false opposition:

By behaving in a manner contrary to social norms, the saintly madman

attempts to reach a perfect understanding of the MƗdhyamika thesis that
there is no distinction between good and bad, that everything has a
“uniformity of flavour” (ro-snyoms). (ibid., 335)

In transgressing conventional rules, which would normally arouse ngo tsha

in the transgressor, the saintly madman demonstrates his spiritual
172 Timm Lau

attainment, “slash[ing] away at the rules of interaction and the hypocrisies

which they engender—excessive, insincere deference, politeness, and
humility” (ibid., 336). These are contexts in which transgressing social
hierarchies and the normal self-regulation of shame are culturally
elaborated and positively, rather than negatively, evaluated. The very
existence of these contexts and the way they work, however, presuppose
and in a sense underline the importance of ngo tsha as an emotional
medium of everyday social hierarchies.

In my fieldwork, the crossing of local hierarchical boundaries and
being publicly “in the wrong place” engendered a key emotional
experience in me. My own socialisation during fieldwork carried similar
elements to Tibetan children’s socialisation into shame in the sense that it
involved intense emotional experience induced by others and learning
about hierarchical boundaries. This enabled me to see the similarity to the
way in which ngo tsha is connected to my Tibetan informants’ awareness
of social position and status in social interaction. I was sensitised towards
the ways in which feelings of shame are drawn upon and elaborated by
Tibetans, at the same time establishing and reproducing the hierarchical
sociality out of which they are generated.
I have demonstrated that emotions can have a sensitising effect on the
ethnographer, making it possible to recognise other ethnographic data as
“evidence for” certain local meanings. In my example, my own emotional
experience of exposure and transgressing hierarchical boundaries shed a
new and distinctive light on ethnographic data on socialisation, behaviour
during meetings, and relationships between the sexes in the Tibetan
settlements of my fieldwork. It enabled me to become aware of the way in
which meaning is attached locally to notions of shame vis-à-vis
hierarchical superiority. Admittedly, in my own description I appear as the
initially naïve ethnographer who learns through mistakes—an oft-used
trope in ethnographic writing (see Varley, this volume). There is, however,
some truth to this trope as mistakes are a very effective way of social
learning, not only during fieldwork. Moreover, rather than presenting a
moral story of ethnographic growth, I have emphasised the way in which
the emotional experience generated by my “mistake” fostered an
analogous understanding of Tibetan shame
Finally, I wish to reflect on anthropological practice by drawing an
analogy and making a distinction between anthropologists and my Tibetan
informants. At the outset, I mentioned Kilborne’s contention that
Understanding Tibetan Shame and Hierarchy 173

experiences associated with shame are generally excluded by

anthropologists from their publications. The dynamic of public and private
realms at work in both academic anthropology and Tibetan sociality are
similar. As we have seen, status-performing activities are vital for Tibetan
social existence, and ngo tsha indicates the boundary of public (status-
performing) activities, and private (non status-performing) activities. Thus,
what would be shameful is best hidden from view and not made public.
For anthropologists, publishing (and thus publicising) is the status-
performing exercise par excellence. This performance of status is vital for
the anthropologist to exist in the social realm of those who do
anthropology. A non-performance of status in academia is a form of social
death: publish, we are told, or perish. And as Kilborne notes, the shameful
is left out of this performance of status in the public realm and instead kept
private through the anthropologist’s selection of material for publication.11
In this, the social practices of both Tibetans and anthropologists are
structurally similar.
From here, we can ask the following question: if the Tibetan ngo tsha
dynamic serves, from an analytical perspective, the structuration of
Tibetan sociality, what can we learn from this analogy for an analysis of
the social practice of anthropology? Both Tibetan and academic
anthropological forms of sociality contain the structural features of private
and public realms, and share an interest in separating the types of
behaviour appropriate to each realm. Keeping what does not perform
status from the public realm is a common structural characteristic. The
difference lies in the fact that Tibetans name this principle of their
sociality and ascribe to it the positive quality of bestowing the knowledge
necessary to be a moral social being. Conversely, as anthropologists we do
not have a similarly integrated place for “our shame”, and tend to avoid its
acknowledgement in academic sociality. Shame remains unnamed and
unrepresented; the very fact lamented by Kilborne.
It is my suggestion, however, that the anthropological hiatus with
regard to acknowledging the role of shame (and perhaps of emotions more
generally) is an enabling feature of anthropology as we know it. Just as
Tibetan ngo tsha is one of the elements that maintain Tibetan forms of
sociality and hierarchy, anthropological elision of shame is one of the
elements that enable anthropology to exist in its present form. In both
cases, the treatment of shame helps secure existing hierarchies of
knowledge which engender forms of sociality. Acknowledging the
important analytical role which emotions can have in evidentiary terms
may contribute to a change of such hierarchies of knowledge in future
174 Timm Lau

This article is based on fieldwork with Tibetans in the Bir Tibetan
settlements in Himachal Pradesh, India, carried out from March 2004 until
July 2005. It developed out of a paper presented at the conference
“Emotions in the Field”, 14-15 September 2006 at Lincoln College,
Oxford University. Research was supported by a Dissertation Research
Grant of the Wenner-Gren Foundation; a Research Studentship of the
Economic and Social Research Council, UK; a Reginal Smith Studentship
of King’s College, Cambridge; a Cambridge European Trust Bursary; a
Wyse Trust Grant of Trinity College, Cambridge; and a Ling Roth
Scholarship of the Department of Social Anthropology, Cambridge
University. I thank these institutions for their financial support; Dimitrina
Mihaylova and James Davies for their support before and during the
conference in Oxford; James Laidlaw for comments on the first draft;
Hildegard Diemberger, Caroline Humphrey, and Giovanni da Col for
comments on a later version of the paper given 23/10/07 at MIASU,
Cambridge University; Liana Chua and Casey High for inclusion into the
volume and editorial comments; and my informants and friends in Bir for
their generosity.

I use the terms public and private throughout this paper in their meaning of
“marked/unmarked” or “performed/muted”. Michelle Rosaldo’s work on the
Ilongot critically engages the cross-cultural validity of the “popular Western view”
of shame constraining “desires and fears as remain real inside the self, although in
public life denied” (1983, 149, emphasis added; see also Rosaldo 1980). My
analysis does not present Tibetan shame thus, i.e. as circumscribing the private as
the real and “hidden” self. Instead, it presents Tibetans as concerned with status,
and shame as involved in a structuring process which includes selves and social
Their names are direct references to the areas of Nangchen (nang chen) and Dege
(sde dge), separate political units in the Kham region of eastern Tibet (khams),
each consisting of subunits with their own rulers before the restructuring of the
political landscape of Tibet was initiated by the Chinese.
Folio 23a-23b of Ye shes mkha' 'gro bsod nams 'dren gyi sku skyes gsum pa rje
btsun ma chos kyi sgron ma'i rnam thar, an incomplete manuscript of 144 folios.
(Diemberger 2007).
Steven Parish’s description of lajya (1991, 1994) presents a careful analysis of
shame among Newars in the Hindu city Bhaktapur in Nepal. His work provides
some interesting parallels from a non-Tibetan cultural area of Nepal.
Understanding Tibetan Shame and Hierarchy 175

Elsewhere, I describe the ill-effects of the latter and the importance of ‘cham po
(“friendly”) as another normative emotional concept effective in Tibetan sociality
(Lau forthcoming a).
For a description of how Tibetan honorific nouns are intrinsically categorised
according to the parameters of social interaction which expresses respect towards
individuals in reference to status, see DeLancey (1998).
In this, Tibetan honorific language is again similar to Balinese (Howe 2001, 91).
Interestingly, even those Tibetans who do not know English used it in this way,
which was wide-spread in my fieldsite and may have been acquired from Indian
television, where interspersed English words are common.
This does not mean that private conversations are not indulged in: in fact, they
frequently are, especially as the meetings (which can be quite lengthy) drag on.
What is significant here is the act of “public speaking”, of raising one’s voice and
speaking in a manner that reaches, and is supposed to reach, the assembled
See also Lau (forthcoming b) for the emotion of fear. I am thankful to Vincent
Crapanzano for raising this question at the “Emotions in the Field” conference in
Oxford, where I presented the original version of this paper.
Documents of non status-performing characters do of course exist. What can
happen when they do enter the realm of published accounts is amply illustrated by
the academic scandal created by Malinowksi’s A Diary in the Strict Sense of the
Term (1989; Kilborne 1992, 243-44).

Ardley, Jane. 2003. Learning the art of democracy? Continuity and change
in the Tibetan government in exile. Contemporary South Asia 12 (3):
Ardussi, John and Lawrence Epstein. 1978. The saintly madman in Tibet.
In Himalayan anthropology: The Indo-Tibetan interface, ed. James F.
Fisher. The Hague: Mouton.
Aziz, Barbara Nimri. 1978. Tibetan frontier families. Reflections of three
generations from D'ing-ri. New Delhi: Vikas.
Bell, Charles. 1928. The people of Tibet. Oxford: Clarendon Press.
Bourdieu, Pierre. 1977. Outline of a theory of practice. Cambridge:
University Press.
Cassinelli, C.W. and Robert B. Ekvall. 1969. A Tibetan principality. The
political system of Sa sKya. Ithaca, New York: Cornell University
Diemberger, Hildegard. 1993. Blood, sperm, soul and the mountain. In
Gendered anthropology, ed. Teresa Del Valle. London: Routledge.
176 Timm Lau

—. 2007. When a woman becomes a religious dynasty. The Samding

Dorje Phagmo of Tibet. New York: Columbia University Press.
DeLancey, Scott. 1998. Semantic categorization in Tibetan honorific
nouns. Anthropological Linguistics 40 (1): 109-23.
Fillipi, Fillipo de, ed. 1937. An account of Tibet: the travels of Ippolito
Desideri of Pistoia, S.J., 1712-1727. London: Routledge.
Fjeld, Heidi. 2005. Commoners and nobles. Hereditary divisions in Tibet.
Copenhagen: NIAS.
Frechette, Ann. 2004. Tibetans in Nepal. The dynamics of international
assistance among a community in exile. New York and Oxford:
French, Rebecca R. 1995. The golden yoke: the legal cosmology of
Buddhist Tibet. Ithaca, NY and London: Cornell University Press.
Giddens, Anthony. 1979. Central problems in social theory. London:
Goldstein, Melvyn C. 1971. Serfdom and mobility: an examination of the
institution of “human lease” in traditional Tibetan society. Journal of
Asian Studies (30) 3: 521-34.
Gombo, Ugen. 1983. Cultural expressions of social stratification in
traditional Tibet: “caste” and casteism in a non-Hindu context.
Anthropology 7 (1): 43-72.
Hardman, Charlotte. 1981. The psychology of conformity and self-
expression among the Lohorung Rai of east Nepal. In Indigenous
psychologies. The anthropology of the self, ed. Paul Heelas and
Andrew Lock. London, New York: Academic Press.
—. 2000. Other worlds. Notions of self and emotion among the Lohorung
Rai. Oxford and New York: Berg.
Howe, Leo. 2001. Hinduism and hierarchy in Bali. Oxford: James Currey.
Huber, Toni. 1994. Why can’t women climb pure crystal mountain?
Remarks on gender, ritual and space in Tibet. In Tibetan studies:
Proceedings of the sixth seminar of the International Association for
Tibetan Studies, ed. Per Kvaerne. Oslo: Institute for Comparative
Research in Human Culture.
Josephides, Lisette. 2005. Resentment as a sense of self. In Mixed
emotions. Anthropological studies of feeling, ed. Kay Milton and
Maruªka Svaªek. Oxford and New York: Berg.
Kharat, Rajesh. 2003. Gainers of a stalemate: the Tibetans in India. In
Refugees and the state: practices of asylum and care in India, 1947-
2000, ed. Ranabir Samaddar. New Delhi, Thousand Oaks, and London:
Understanding Tibetan Shame and Hierarchy 177

Kilborne, Benjamin. 1992. Fields of shame: anthropologists abroad. Ethos,

20 (2): 230-253.
Lau, Timm. 2007. The Tibetan diaspora in India: Approaching itinerant
trade, popular cultural consumption and diasporic sociality. PhD
thesis, University of Cambridge.
—. Forthcoming a. Hindi film’s romance and Tibetan notions of harmony:
Emotional attachments and personal identity in the Tibetan diaspora in
India. Journal of Ethnic and Migration Studies.
—. Forthcoming b. Tibetan fears and Indian foes: Fears of cultural
extinction and antagonism as discursive strategy. Vis A Vis:
Explorations in Anthropology.
Leavitt, John. 1996. Meaning and feeling in the anthropology of emotions.
American Ethnologist 23 (3): 514-39.
Lutz, Catherine and Geoffrey M. White. 1986. The anthropology of
emotions. Annual Review of Anthropology 15:405-436.
Malinowski, Bronislaw. 1989. A diary in the strict sense of the term.
Stanford: Stanford University Press.
Mauss, Marcel. 1967. The gift: Forms and functions of exchange in
archaic societies. New York: Norton and Company.
McGranahan, Carole. 2005. Truth, fear, and lies: Exile politics and
arrested histories of the Tibetan resistance. Cultural Anthropology 20
(4): 570-600.
Parish, Steven M. 1991. The sacred mind: Newar cultural representations
of mental life and the production of moral consciousness. Ethos: 19
(3): 313-351.
—. 1994. Moral knowing in a Hindu sacred city. An exploration of mind,
emotion and self. New York: Columbia University Press.
Richardson, Hugh. 1984. Tibet and its history. Boston: Shambhala
Rosaldo, Michelle Z. 1980. Knowledge and passion: Ilongot notions of
self and social life. Cambridge: Cambridge University Press.
—. 1983. The shame of headhunters and the autonomy of self. Ethos 11
(3): 135-51.
Rosaldo, Renaldo I. 1984. Grief and a headhunter’s rage: on the cultural
force of emotions. In Text, play and story: The construction and
reconstruction of self and society, ed. Edward M. Bruner. Prospect
Heights, IL: Waveland Press.
Sagant, Philippe. 1996. The dozing shaman: The Limbus of eastern Nepal.
New Delhi: Oxford University Press.
Shakabpa, Tsepon W.D. 1967. Tibet: A political history. New Haven: Yale
University Press.
178 Timm Lau

Shakya, Tsering W. 1994. Politicisation and the Tibetan language. In

Resistance and reform in Tibet, ed. Robert Barnett and Shirin Akiner.
London: Hurst and Company.
Snellgrove, David L. and Hugh Richardson. 1968. A cultural history of
Tibet. London: Weidenfeld and Nicolson.
Stein, Rolf A. 1972. Tibetan civilization. London: Faber.
Tonkin, Elizabeth. 2005. Being there: emotion and imagination in
anthropologists’ encounters. In Mixed emotions. Anthropological
studies of feeling, ed. Kay Milton and Maruªka Svaªek. Oxford and New
York: Berg.
Wikan, Unni. 1992. Beyond the words: The power of resonance. American
Ethnologist 19 (3): 460-82.
Yeh, Emily. 2002. Will the real Tibetan please stand up! Identity politics
in the Tibetan diaspora. In Tibet, self, and the Tibetan diaspora. Voices
of difference. Proceedings of the ninth Seminar of the International
Association for Tibetan Studies, ed. Christiaan P. Klieger. Leiden:


Poetry is the spontaneous overflow of powerful feelings: it takes its origin

from emotion recollected in tranquillity: the emotion is contemplated till
by a species of reaction the tranquillity gradually disappears, and an
emotion, kindred to what was before the subject of contemplation, is
gradually produced, and does itself actually exist in the mind.
—William Wordsworth, Preface to Lyrical Ballads, 2nd ed, 1802.

This chapter examines the impact of fieldwork “recollected in

tranquillity” on what is considered to be ethnographic evidence. In my
reflexive critique of the epistemological conditions that gave rise to that
evidence, and their shifts over the years of fieldwork, I follow a simple
trajectory. I take two books, from the two ends of my career as an
anthropologist, and examine them for evidence of the influences that went
into their making. Then I consider the space between them, that of returns,
both actual and metaphorical. Because actual returns are suffused with
metaphorical ones, both can be seen to some extent as “virtual”. My
argument itself is a metaphorical journey from one monograph to another,
tracing the different stages of ethnographic practice and the different
conceptions of evidence.
My fieldwork among the Kewa of the Papua New Guinea (PNG)
Highlands utterly captivated me, becoming an enduring aspect of my self-
definition. I arrived in Mt Hagen with barely suppressed excitement,
having just immersed myself in the ethnography of the area prior to
writing a dissertation about suppressed and overt antagonism—suppressed
among men, overt between the sexes (Josephides 1982a). The pungent
smell of wood fires as it wafted from people’s clothes and skins
overpowered my senses at the airport, permeating my whole being and
becoming a permanent memory-trigger, causing me to relive that moment
at the slightest whiff of a wood fire. At the time it brought to life my
earlier iconic encounter with PNG Highlanders. This was on a promotional
poster by PNG’s national airline showing the plumed head of a man, an
enormous bird of paradise, and in the background the figure of a woman
180 Lisette Josephides

with a netbag on her head surrounded by gambolling pigs. The inscription

read: “Papua Niugini: The Last Place You’ve Never Been.” I had come to
the last place on earth, yet nothing here seemed foreign to me.

The Production of Inequality

My first monograph, The Production of Inequality, focused on how
inequalities developed in these famously egalitarian societies. According
to my analysis they developed mainly along gender lines through a process
of mystification, when men siphoned off the products of the domestic
sphere into the exchange sphere. I characterised exchanges as
smokescreens concealing the domestic labour that went into the
production of pigs as exchange items. My argument had to contend with
the distinction between the economic concepts of “use value” and
“exchange value”, which hid behind another distinction, that between
thing and commodity. To quote from the book:

Within the capitalist economy, where labour power is itself a commodity,

the wage formally buys the workers’ claims to the products of their labour.
Since … the labour power thus bought produces more exchange values
than are necessary for its own reproduction, and since it is on the cost of
the latter that the wage is calculated, exploitation takes place when a
portion of the commodity produced by the worker is not paid for by the
capitalist. This is the basis of the theory of surplus value. But a Kewa
woman has not sold her rights in her product, which she can still claim.
Her labour power is not a commodity, so alienation of the product is not
automatic. Precisely because pigs are not produced as alienable items, their
alienation and conversion into exchange values must be carried out in a
cloud of mystique. They are not produced for alienation, yet they are
alienated. People produce use values, yet exchange exchange values.
(Josephides 1985, 209)

The argument my book sought to demonstrate was that although Kewa

production is geared to use values, another activity, that of ritualised
exchange, converts these use values into exchange values—and in the
process conceals the relationship between prestige and work. Wealth is
drawn into the exchange sphere under the guise of serving the interests of
the group but then becomes reallocated as prestige (symbolic capital) to a
few persons. In this way exchange is fetishized as creating wealth,
whereas in reality wealth is merely appropriated there by being
transformed into prestige. Ceremonial exchanges were smokescreens for
this extraction. Since men monopolised the convertibility of values from
Virtual Returns: Fieldwork Recollected in Tranquillity 181

the domestic sphere into the political sphere, women had to be seen as
While my fieldwork recorded the observed surface of things, my aim
was to uncover a hidden structure—one that covered up or mystified social
relations. Thus my monograph explained the problems of a complex social
reality through the use of a theoretical model that posited a contradiction
between ideology and practice. With Marx and Bourdieu as guides I
looked for (and found) alienation, exploitation, and smokescreens
disguising relations of production. In this view, “culture” (always
conceptually in quotation marks) was always somebody else’s unfair
advantage. Fieldwork provided the necessary evidence for these
conclusions, dazzling me with theoretical concepts springing fully-formed
from ethnographic observations like Athena from the brow of Zeus. A
good example was the twin concepts of “sojourner” and “peripheral” to
describe women (Josephides 1985, 65). An examination of how I arrived
at these concepts sheds light on the role of evidence, or what I took to be
evidence, at every stage of theory-formation.
Kewa women lived virilocally. When they were divorced or widowed
they might remarry and move away. These possibilities—of divorce or
widowhood followed by remarriage and departure—allowed men to
describe women as temporary residents of any clan settlement. The
putative “transitoriness” was then given as a reason for considering
women to be ideologically peripheral to any clan. From being seen as
“physical sojourners” in one clan women thus became “ideologically
peripheral” to all clans. What this sleight of hand covered up (or
“masked”) was the reality that most women, once their marriage was
established, were unlikely to leave their husbands’ settlement, so their
residence was as permanent as any man’s. Thus the composite concept
“peripheral sojourner” was exposed, as soon as invented, as self-referential
or circular, a false syllogism that provided evidence for my theoretical
perspective. It was based on trick and deception, yet informed real-life
attitudes about women.

The nature of evidence

While I was in the field I never doubted what counted as evidence on a
prima facie level. Potentially everything was evidence—but evidence for
what? Before collecting the building blocks for an edifice the ethnographer
must first imagine its shape. The interpretative act is needed to link
evidence to edifice as cause to effect, and this act itself had a prior
motivation in the intentionality, worldview, and state of receptivity of the
182 Lisette Josephides

interpreter. For me at that time, economic and political edifices were at

once the most basic and the most determinant, and the strongest evidence
was activity deemed to serve their ends. Though I had rejected the
reification of culture, I espoused ideas of the structural dominance of
“relations of production”, whose mystifications and smokescreens I first
discovered and then exposed.
On the way to interpreting evidence I sometimes privileged one of its
explications because of its fit within the larger picture I had already
formed in my understanding. Thus I universalised a particular rationality.
Two concerns cropped up again and again during fieldwork and writing
up: the temptation to interpret observations through a familiar logic, and a
preoccupation with coherence. My first two pursuits in the field, to collect
genealogies and match them with clans and to document clan stories of
origin, immediately came up against my concern with “logical
connections”—or lack thereof. In the Yala clan story of origin the apical
ancestor was a woman called Rimbanu, who had a son named Kaima. My
informant Rimbu agreed that if his subclan had been called Kaima rather
than Paripa, or if Rimbanu’s child had been called Yala, the genealogies
would have made better sense. But unfortunately this was not the case, and
nobody could tell me how the clan name “Yala” originated. I called this
untidiness “plain ignorance of facts”, and Rimbu was distressed at not
being able to satisfy me on this score. (With some inconsistency, I
accepted the presence of classificatory kinship as a legitimate form of
organisation identified by anthropological inquiry.)
Related to the self-imposed demands of this stringent logic was my
concern with “reality”. I made distinctions between actions done for effect,
in order to achieve a goal beyond their ostensible one, and therefore
reversible; and “real” actions—by which I understood those with direct
face-value only—which “spoke for themselves”. When Pastor Rorea broke
shells at the pig kill and withdrew in high dudgeon, only to return, having
made his point, I recorded in my fieldnotes: “Is nothing real, is everything
done for effect?” On re-reading my notes in tranquillity I asked myself:
Why must I always be looking for a more enduring reality than this effect,
this rhetoric and the making of a point? Instead, I began a more radical
questioning of basic concepts, such as reality, logic, and coherence. The
answer, as I now saw it, was not to look “deeper” or get a thicker
description, but to be freed from a single constraining logic.
As for coherence, I considered it a basic requirement for intelligibility.
Bourdieu had written that ritual practice “owes its practical coherence
(which may be reconstituted in the form of an objectified diagram of
operations) to the fact that it is the product of a single system of
Virtual Returns: Fieldwork Recollected in Tranquillity 183

conceptual schemes immanent in practice” (1977, 118). Taking my cue

from Bourdieu I attempted a diagrammatic modelling of the Kewa pig kill,
with the aim of capturing a particular exchange in its entirety (Josephides
1985, 192-194).
But as I progressed in this I began to question the likelihood that such a
diagram could constitute the full meaning of a ritual practice. I took the
“practical coherence” of these practices to mean that they made sense, that
they can be read off as a logical series in the form of an objectified
diagram of operations. Yet on secondary writing I found that my neat
tables simplified transactions as viewed from one perspective. In this form
they could be shown to correspond to a single system of conceptual
schemes—but it was I who had constructed this system. It was possible to
have many different authentic pictures, according to who is describing the
operations. A man will say, “I gave this leg of pork to this partner for that
reason, I gave to my wife’s kin for another”; but his wife may give
different accounts and different reasons, giving rise to a different
conceptual scheme.1 Are these different interpretations of the same
transaction? If so, the transaction cannot be described or diagrammed
objectively because it has no objective meaning beyond practice, as a
transaction that happened with no further explanation of its implications.
The only meaning it can have is interpreted, positional meaning.
Bourdieu, as the master explicator of strategies, acknowledges the
positionality of meaning when he writes that objective meanings are “in
truth” past regulated improvisations (1977, 78). (My concepts of
“rehearsed and rehearsing talk” in Melanesian Odysseys also develop this
point.) He invokes the “genesis amnesia” of the objectivist apprehension:
it is only when apprehended in pure synchrony that the product of history
appears as objective meaning (ibid., 79). Coherence, then, reveals position
(there is more than one way of putting things together to make sense),
reduces complexity, and excludes data that resist classification or cannot
be beaten into shape. Yet Bourdieu’s coherence model treats evidence as
the stuff that “reconstitutes”, in a diagram or representation, “a single
system of conceptual schemes immanent in practice”, the diagram and
practices reflecting each other’s coherence. Though the acknowledgement
that “synchrony makes objectivists of us all” exposes this impossible
bird’s eye view, Bourdieu clings to the model of mutual reflection as
evidence for the immanence of coherence.
Yet there is no escaping this ethnographic reality: that even when I
have a wealth of data that can potentially be used as evidence, it is only
after selecting those that fit in the diagram (in order to create models,
structures, or theories) that evidence has a role to play in producing
184 Lisette Josephides

meaning and making a cultural practice intelligible. Models replace the

world, but there is no other way to understand it. For Wagner (2001, 253),
all such models for understanding are “miniatures” said to replicate the
reality they are supposed to represent; but in the process of their
construction they reduce reality to their scale, so that representation is
controlled by its own shadow. It is not a question here of positionality or
even selection of data, but of scale, a lessening of complexity rather than
the adoption of a point of view.2 In striving to understand our models of
the world, we lose sight of the world.
The remainder of this chapter develops a different conception of
evidence, one that avoids this type of modelling. Starting from
“recollections in tranquillity” and moving on to an examination of
phenomenological perspectives and the role of emotion in understanding,
it gives precedence to reality or “life as lived” and people’s actions and
intentions, rather than objective meaning or intelligibility. In the
conclusion I suggest that ethnographic evidence and theorisations in my
later work derive their legitimacy not from abstract “modelling” but from
the “dialoguing” of people’s own discourses in everyday life, as
understood within the larger context described.

Recollections in tranquillity
What happens to evidence after the ethnographer has left the field for
the first time? I opened this chapter with a quotation from Wordsworth. In
his autobiographical poem The Prelude, Wordsworth talks of his boyhood
as an apprenticeship, when he was “[f]ostered alike by beauty and by
fear”, physically experiencing the joys that formed his moral education
and became the wellspring of his poetry. As a boy he loved the sun, not as
a pledge of our earthly life but because he “had seen him lay / His beauty
on the morning hill”. Years after these formative experiences, when the
physical joy was forgotten but the scenes of that joy remained, “[d]epicted
on the brain, and to the eye / Were visible, a daily sight”, he outlines the
technique of recollecting in tranquillity. “[Slackening his] thoughts by
choice”, he settles “into gentler happiness”; then (as he explains in the
Preface to the Lyrical Ballads) he contemplates the emotion “till by a
species of reaction the tranquillity gradually disappears, and an emotion,
kindred to what was before the subject of contemplation, is gradually
produced, and does itself actually exist in the mind”. In this recollection
without “bodily eyes” the soul remembers how it felt but not what it felt,
thus retaining “an obscure sense / Of possible sublimity”.3
Virtual Returns: Fieldwork Recollected in Tranquillity 185

For Wordsworth recollection in tranquillity is one in which the

experience is distilled to its essence qua experience, remembered and
relived in the mind, with an attenuated connection to an actual concrete
event. Emptied of specific content, it relates directly to us as meaningful,
and this direct relation is what is experienced as sublime. This process
appears to be halfway between Kant’s transcendental approach and
Husserl’s phenomenological reduction.4
But for the anthropologist, matters cannot be left in this transcendental
realm if “virtual returns” are to be the basis of ethnographic evidence.
Purged of the content that gave rise to it, “unplaced” memory becomes
wonder, or wonderment. It has transformed the experience into a sense of
possibility. Wordsworth takes us this far. But this possibility does not
remain a pure and empty form. In my analysis it seeks out and becomes
suffused with other memories, perhaps similarly processed, thus becoming
recontextualised within a broader experience. Virtual returns add more to
that original experience rather than simply stripping it of its particularity.
The relation to the broader context follows from recollection in
tranquillity, in the making of connections in the web of meanings learned
and constructed. This process of construction through memory can then be
informed by the near sublimity experienced in the tranquil process.
To sum up, these are the steps by which recollection in tranquillity
yields up new insights (as evidence): memory purges experience, making
it direct and sublime; this sublimity makes available a sense of possibility;
the experience thus transformed as a result becomes suffused with all other
memories and recontextualised within a broader experience; the near-
sublimity of the tranquil process then informs the expanded experience,
establishing and reinforcing connections with other memories. For the last
two steps I must go beyond Wordsworth’s contemplation, from the
sublimation of powerful experiences to the linking of experiences and the
understanding of the other of ethnography. The inquiry then entails an
elaboration of “returns” as providing new ethnographic evidence.

From sublimity to the hermeneutical circle

Wordsworth’s contemplation of his experience, though opening onto
“magic casements” (to switch from Wordsworth to Keats) of possible
sublimity, is a personal quest and is not designed to lead to the
understanding of other persons.5 For this understanding, which in the case
of my fieldwork was developed in relations with people over several
actual, physical returns, I turn to Ricoeur’s theory of subjectivity,
according to which “the self is constituted by the ‘matter’ of the text”
186 Lisette Josephides

(1981, 143). With the slight but crucial substitution of “culture” for “text”,
I paraphrase Ricoeur’s description of the hermeneutical circle as follows:

The world of the other culture is not hidden behind arcane practices but
unfolds in front of me, through a series of strategies of making explicit.
Rather than distort it by imposing upon it my finite capacity for
understanding, I understand it by a process of appropriation. To
appropriate is to make into one’s own what was initially “other”. I can
achieve this only by shedding the uncritical and illusory understanding
which I always believed I had of myself prior to being instituted as a
subject by the other culture which I thought I only interpreted. (Ricoeur
1981, 37)

I “exchange the me, master of itself, for the self, disciple of the text”
(ibid., 113). This distanciation of self from itself destroys “the ego’s
pretension to constitute itself as ultimate origin”, thus closing off the
possibility of a “secret return of the sovereign subject” (ibid.). The term
“appropriation” underlines the struggle against cultural distance, in the
fusion of cultural interpretation with self-interpretation. This is the
hermeneutical circle; an interpretation in which the interpreter becomes
transformed (ibid., 178).
At the fieldwork and writing-up stages, Ricoeur’s “text” may be
understood as the culture as lived; ethnography is the hermeneutical
interpretation of that culture, in the manner just described. Ricoeur defines
interpretation as being midway between construction and description,
describing and constructing the possible worlds of an “open and infinite
horizon” (ibid., 126).6 Ethnography offers precisely such a horizon, and
my own concepts of “making explicit” and “eliciting” act as Ricoeur’s
“interpretation”. According to this understanding I should not look for “the
other” in my existence prior to the fieldwork encounter, but only in the
existence characterised by the open and infinite horizon which we now
share. In this shift from understanding the other to understanding the world
of the ethnography (or the encounter), what we appropriate is no longer an
alien experience, but the horizon of a world towards which the work
(ethnography) directs itself—thus a possible mode of being-in-the-world
opens up (ibid., 177). It is not a question here of fusion of consciousness,
of empathy or sympathy, recognition of another person, or projection of

To understand oneself in front of a text is quite the contrary of projecting

oneself and one’s beliefs and prejudices; it is to let the work and its world
enlarge the horizon of the understanding which I have of myself. (ibid.,
Virtual Returns: Fieldwork Recollected in Tranquillity 187

We learn about ourselves only through our acts, the exteriorisation of

our lives and the effects they produce on others (ibid., 52). The experience
of the other parallels and develops my being, but this being is more than
myself, having “a potentiality of meaning which exceeds the gaze of
reflection” (ibid., 128), reflection itself being nothing without the
mediation of “signs scattered in the world” (Ricoeur 1970, 46). Being
itself situated, reflection must always begin by interpreting the cultural
products of a specific tradition; thus it is concrete.
I follow Ricoeur as he develops the phenomenology of Husserl and
Heidegger, bringing it back from a sterile ontology to a world with others
who may be known.7 Ricoeur reinjects epistemology and ethics into
phenomenology, by describing the experience of the other as expanding
my being, and the gaze of reflection as requiring the mediation of “signs
scattered in the world”. With Ricoeur as guide, we return to the world of
human beings.
Applied to ethnographic fieldwork this analysis might yield the
following narrative: as ethnographers we do not appropriate an alien
experience, or the culture of the people we study, so that we may invent it
or them; what we make our own is a projection of the world, a mode of
being-in-the-world which our fieldwork discloses. (The possibility that
incompetence, cowardice, confusion, or bad faith may distort such
projection, debasing it into paltry interpretation, does not invalidate the
observation.) Thus the ethnography is always a projection of the world,
not an alienation. It is an interpretation of the fieldwork experience, and
for Ricoeur an interpretation is authentic only if it makes its own what was
initially alien. This is not to be understood as a mutual appropriation
between two subjectivities (the ethnographer’s and the informant’s), but a
displacement of the hermeneutical circle “from the subjectivist level to an
ontological plane”, where the possibilities opened up by the world
projected by the fieldwork enlarge my mode of being. For Ricoeur this
disclosure of the world is synonymous with Aristotelian mimesis, “a future
horizon of undecided possibilities” (1981, 187). In that projection there
can be no alienation, only a finding.
This insight responds to a dilemma in the problematic alternative
between “alienating distanciation and participatory belonging” (ibid., 131).
As an ethnographic problem it can be stated as the antinomy implied in the
compound phrase “participant observation”. The distanciation, for Ricoeur
as for me, is not alienating, but the very condition even for self-
understanding. All consciousness of meaning involves a moment of
distancing from “lived experience” (ibid., 116); our very use of language
exemplifies this, as it consists of signs (words) for things that stand for
188 Lisette Josephides

things they are not (ibid., 153). Thus the speaking subject must be able to
retreat to an “empty space” from where the use of signs can begin.
While it is true that empathy does not follow from transformation
through the hermeneutical circle, such transformation in the ethnographer
was needed before emotions could achieve this sort of understanding. In
Melanesian Odysseys, which I wrote using ethnography “recollected in
tranquillity”, I sometimes scramble time periods to show the process in
which empathy developed between me and several Kewa individuals. I
reproduce notes in the words of earlier days of the encounter and comment
on them from a later and changed perspective, revealing how certain
understandings forced themselves on me following an awkward episode;
thus tracing the process of anthropological knowledge as it developed by
using personal relations as evidence. Like Fabian (1990), I found that
cultural knowledge is not transferable as information of pre-existing
messages via signs, nor accessible through a question-and-answer method
(Favret-Saada 1980). Rather, the quality of the relationship was a measure
of the process of understanding.
The ethnographer whose being is resituated in the field at a certain
point becomes aware of the potentiality of meaning not only beyond her
being, now enlarged by the field, but also beyond the gaze of that enlarged
reflection. Reflection can only come into play when it encounters “signs
scattered in the world”, which are products of specific traditions. But as
“signs” they require interpretation, thus they are beyond mere gaze. As an
ethnographer writing The Production of Inequality, I encountered
“exchange” as such a sign, the product of a specific tradition, and
interpreted it beyond the gaze of my reflection, discovering the hidden
meanings of extraction and its masking. But in this interpretation I still
kept aloof, looking from the outside. Later, following virtual returns and
the suffusion of experiences, I realised that letting the field “enlarge the
horizon of the understanding which I have of myself” (Ricoeur 1981, 178)
was precisely what was happening in those conditions when
understandings forced themselves on me against my inclination. I became
embedded in the local “economy of emotion” when I understood the full
implications of having been made Rimbu’s sister. Empathy, following
from transformation through the hermeneutical circle and the subsequent
work of emotion, emerges as evidence, because on it depend the
relationships through which the ethnographer will understand local life
and people.
Virtual Returns: Fieldwork Recollected in Tranquillity 189

Emotions as evidence
My attitude to my bulging ethnographic collection on my first virtual
return was ambivalent. I recollected on going back to it:

Why is it that while my fieldnotes always make interesting reading and

rarely embarrass me, reading my diaries can be excruciatingly painful and
depressing? They often present different angles and insights, missing from
my fieldnotes, yet whenever I read them I am overwhelmed by feelings of
loss, regret, desolation; a fear that my fieldwork has been incomplete and
my own commitment inadequate. The problem was not lack of insight,
sensitivity, preparation, intelligence, imagination or even empathy. I felt I
knew my Yala friends through affective bonds. Nonetheless, a disquieting
feeling persisted that I held something back, and this impeded a total
immersion into the local culture. A sullen resentment, at times even
defensive aggression, permeated those entries. I wanted to retain myself, as
an active agent with powers of interpretation and control. (On rereading
my fieldnotes and diaries in 1991)

The “sullen resentment” was read into the text years later, in tranquil
contemplation away from the field, and was experienced as a source of
disquietude now. It was now that I felt regret for “holding back”, now that
I recognised this secret reserve as a wish to “retain myself”. A vestigial
resistance to incorporation always tempered my desire to count for
something in the village. In writing the passage quoted above, I recognised
the regret searing my soul as being existential, because it concerned
behaviour that risked self-loss yet had to be ventured. Potential self-loss
was of two kinds: of the transformation of the self I had known all my life,
and the loss of control. Now, again, several years from this reading of the
situation, I realise that the angst expressed above also revealed the extent
to which I was already caught and transformed. The hermeneutical circle
does not work in a straightforward or absolute way, but leaves a residue of
feelings locked in various vaults of memory and consciousness.
Several dreams since leaving the field reveal a preoccupation with the
moral aspects of my relations with my local hosts. Emotions acted here as
moral judgements.8 The fact that dreams, “like fell cruel hounds, evermore
pursue me”, can be seen as evidence of a subconscious or repressed
understanding. In 1984, while living in Port Moresby in between field
visits, I wrote the following:

I dreamt we returned to our highlands Kewa village, to find it built up out

of all recognition. There was a supermarket and a department store, and
scores of brightly-clad Europeans ambled about leisurely. … Our old
190 Lisette Josephides

house was completely run down, dwarfed by a luxurious mansion that

Rimbu, my Kewa patron and adoptive brother, had built alongside it.
Rimbu told us that he often put up European guests in his palatial home. I
felt a tightness in my throat and a hollowness in the pit of my stomach.
We’d been cheapened! With so many Europeans around we’d go
unremarked, and with the glut of luxury goods we would lose our
superiority. In a flash our relationship with Rimbu and other villagers
appeared in stark outline: it was based on inequality, and once
“democratisation” took place we would lose our structural advantage.
(Josephides 2003, 55, from which the paragraph that follows is also

Since it was not my conscious ideal to live in unequal relations in the

field, I always thought of them as being symmetrical on some level. If I
was treated as being in a superior position, I considered this their model.
Then, to my horror, a dream revealed that I had nurtured an unconscious
or repressed model of superiority9. Worse, the dream exposed a wishful
misconstruction of my standing in the village: when I began to rummage
through my fieldnotes for concrete displays of the anthropologist’s power,
I discovered only instances of feeling thwarted and marginal, of frustration
at not being informed of important events which were never delayed for
my sake.
At the time of my first monograph (Josephides 1985) I treated
emotions as mere facilitators (or a hindrance) to good fieldwork. As I
recorded after a virtual return,

Certain talk only is to be recorded, analysed to provide support structures

for my arguments, bridges for my understanding: important talk,
“culturally salient”, affecting those structures which it helps me to
construct. Other talk I feel only through the various conditionings of my
senses and emotions; it is to facilitate fieldwork, not be the object of it.
Thus facilitations and enablements are already divorced from objectives,
theoretical conclusions, cultural representations and understandings, the
“meat” and final academic message of fieldwork. (2008, 4)

My rereading pursued this point, asking:

Yet how does the talk appear to those who participate in it, how does it
construct that social life which I thought adequately understood without it
in this role? What do people talk about all the time, seemingly casually and
disinterestedly, sometimes dully and tonelessly, at other times excitedly
and angrily, vacillating between laughter and tears? (ibid.)
Virtual Returns: Fieldwork Recollected in Tranquillity 191

On rereading my fieldnotes I found, unsurprisingly, that my own

emotional involvement played a part in their recording. From time to time
I noted down my feelings, especially in outrage at unfolding events or
disparagement of decisions taken. This disapproval itself is instructive,
serving as a reminder of how independent of my feelings these accounts
were, how powerfully they established themselves, almost against my will
and inclination. I grumbled, but wrote them down.
But first and foremost, the ambivalence of my feelings concealed a fear
of loss of self and identity, a desire to keep myself somehow intact. This
fear revealed itself in several incidents. One was the occasion when Rimbu
put a padlock on the tap of the water drum I had installed for the use of the
whole village. I considered this my gift to the people, and Rimbu’s action
caused me to experience an acute sense of self-loss. The experience
sensitised me to similar fears on the part of Kewa people. When Kiru
returned from a hunting trip and hung a possum outside the house, his wife
Liame took it in with the intention of preparing it for cooking, but Kiru
told her angrily that it was a gift for another female relative. Liame was
affronted; her sense of self and her position as a wife were impugned, and
following angry exchanges she left the house to do her own hunting. She
returned triumphant the following day, insisting she had killed and eaten
several rats. My perception of the relationship between emotion and
personhood developed from such experiences, leading me to theorise
resentment as a sense of self (see Josephides 2005). Though I was
predisposed to see Highlanders as pragmatists, I soon realised that it was
not easy to distinguish between sensible, practical, logical explanations
and emotional ones (see, for instance, my initial readiness to accept
economistic reasons over psychological or emotional ones).
While in the field I also wanted to be acknowledged as a person, in the
way I saw people constantly demanding such acknowledgement. They
understood their culture as a living space where they would engage in such
negotiations and elicit such acknowledgements, because their sense of self
required it; this was how their culture was lived and how people’s relation
to their culture was conceived. And they judged me, too, by how I played
in this field. But I felt guilty about desiring this acknowledgement. I was
an ethnographer, a student learning from the people; I had no rights to my
own insistent personhood. My business was to keep myself out of the
picture and not interfere with the unfolding of their culture. Away from the
field, and in virtual returns, such fears no longer exerted devious control
over my emotions. Rather than understanding through those emotions, as
facilitators, I now tackled them head-on—they became themselves
ethnographic data, not only clarifying how I understood Kewa persons but
192 Lisette Josephides

also how they understood their “culture” and even me. My emotions
became data when I recognised them in equivalent Kewa strategies, as
people acted to define themselves by insisting on particular cultural
understandings. Through my emotions I understood theirs. Evidence, then,
can be found long after fieldwork is over, offering new ways to understand
our fieldwork and its variously recorded materials.
The fear of self-loss with which this section began is an existential
condition. What the anxieties and dreams underline is the sense of the
“unnaturalness” of the situation. The natural way to live is as a person, not
a participant-observer. Despite this, the awkward relations, as fumbling
attempts to interpret evidence, require the development of some empathy.
For even to talk of misunderstanding suggests a crossing over to the
opposite shore. Whether conveyed by intentional or unintentional
scenarios, the crossing always entails a personal transformation.

Actual returns
After my initial doctoral fieldwork, which started in 1979, I continued
to visit the field, first from Port Moresby where I lived until 1986 and later
from the United States. A story grew up around me, especially at the time
of my 1993 return (with Marc Schiltz), that I was “asples”, a local person
who spoke the local language. Proudly my Yala hosts enumerated my
visits, declaring to other Kewa clans that my returns would not cease while
I or Rimbu lived. The returns, more than the length of stay, established
relations of trust; I could be relied upon to come back again and again, as I
had demonstrated. In slicing time, these returns established a shared
history that transformed the relationship. Young people who had known
me as children and in whose background I featured confided in me as an
aunt, and this shared history established my presence in the village as a
right. Different expectations began to be expressed of me, and different
demands were made of me. I was struck, during our 1993 visit, by how
much more closely Marc and I were interrogated about the wider world
which was making inroads into village life. For the first time we noted
cargo-like attitudes and expectations, including speculations about Jesus’
second coming in the year 2000. Questions were pointed, evincing a
concern that some information may have been withheld from them.
But the phrase “while I or Rimbu lived” immediately identified the
parameters within which the claim of a special status was made. It
concerned a relationship, and I returned because of this relationship. The
Virtual Returns: Fieldwork Recollected in Tranquillity 193

quality of the relationship was revealed to be of paramount importance to

the sort of understanding necessary for the construction of an ethnography.
Nor was this a one-way relationship: it had been made clear to me on a
previous return that filtering my relationships with local people through
Rimbu was unacceptable. The outcome of the dispute following Rimbu’s
attempt to control my gift-giving tied me publicly in a reciprocal
relationship with all the Yala of the area. I learned from this dispute that I
was not faced with a dualistic relationship with a homogeneous set of
people with identical interests acting in accordance with specific norms,
with whom I must establish relations by learning acceptable cultural
behaviours and responses. Rather, I had to work at individual relationships
with many different people, who had their own agendas, strategies, ideas,
and goals (compare Metcalf 2002). None of these are inflexible, and the
indelible achievement of the encounter lies in the extent to which it marks
those agendas, ideas, and goals.
In sum, the importance of physical returns is two-fold: they establish
reciprocal relations in which people think of the ethnographer as a reliable
person with whom they share a history, and they encourage the
ethnographer herself to develop new ways of looking at local conditions.

Virtual returns
The benefits of virtual returns are realised in combination with actual
returns, but are important even in their own right, for several reasons.
First, when reviewing my fieldwork materials, the data I had amassed on a
fragmented day-to-day basis suddenly appear in a linear, synchronic
reading, revealed as structures, stories, texts in Ricoeur’s (not Geertz’s)
sense, which called for an understanding “in front of the text”. I was not
aware while writing up my fieldnotes that I had such systematic
information on a topic, which now confronted me as an edifice quite
independent of my artful theorising and conscious data-collecting. Second,
while away from the field I became immersed in comparative reading of
monographs and theoretical writings, whose insights and methodologies
entered my consciousness and provided me with new filters and
perspectives through which to view my own material. This is the
“suffusion” mentioned earlier, which led, thirdly, to a change in my ideas
of what counts as evidence. Fourth, I began to reflect on how I had
understood others.
It is in this context that empathy develops a sort of evidence. As
discussed earlier, the scrambling of time periods, the juxtaposition of
fieldnotes, the critical commentary from later perspectives, the reluctant
194 Lisette Josephides

impositions of understandings, all point to the use of personal relations as

evidence. Virtual returns, mediated by rereadings, emotions, and dreams,
have a potential for a transformative influence on the interpretation of
ethnographic evidence. They affect both emotional and epistemological
understandings giving rise to evidence, discovering new evidence long
after fieldwork is over and colouring subsequent fieldtrips.
My awareness that the Kewa will read my writings about our
encounter has a profound influence on my attitude to the ethnographic
enterprise, forcing me to consider the possibility of alternative depictions
and methodologies. (A dog-eared copy of my published collection of
Kewa stories and songs [1982b] was still passing from hand to hand in
1993.) This is not to claim that ethnographies are the ethnographers’
inventions alone; my agency was affected by how Kewa people placed me
in the field situation and how they chose to reveal themselves. Kewa
inventions of me underwent many shifts. First as a missionary, then a
classificatory sister, then the reincarnation of Rimbu’s sister who had died
in infancy, then as mediator with the powerful outside world, especially as
the birthplace of Jesus; and throughout and simultaneously as a returned
ancestor. These multiple inventions, which are both metaphorical and
empirical levels of understanding, suggest that the Kewa themselves go
through equivalent processes of “virtual returns” in their relationship with
me; but this angle cannot be developed here.
A description of my arsenal of ethnographic materials from the field
makes a good entry into the discussion of Melanesian Odysseys,
completed at the other end of my career (“completed” rather than
“written” because the writing project was underway for more than a
decade). This collection included (to name a few) fieldnotes, diaries,
emotions, dreams, tape recordings, videos, photographs, language
materials, songs and poems, objects, unrecorded memories, and mental
snapshots. It bulged with imponderabilia (some of them, such as dreams,
continued to be collected post-fieldwork) that would later provide the
seeds for new “imaginative horizons” (Crapanzano 2004). This hidden
promise is what makes virtual returns so rewarding. Early on in my
fieldwork I began to collect the stories that became the backbone of
Odysseys, without giving a thought at the time to their possible future use.
Though my theoretical perspective determined how I interpreted evidence
at the time, as well as what I took to be evidence, those limiting conditions
did not apply to my fieldwork or recording.
These heterogeneous materials are by no means analogues, nor are they
a purposive collection of data, though I have processed them all, whether
consciously or unconsciously. Nonetheless, the more spontaneous (or
Virtual Returns: Fieldwork Recollected in Tranquillity 195

unwilled) data suggest a shift in the balance of power, taking control out of
my hands. The operation of emotions here is not unlike the process
Wagner describes, when he writes that “we do not learn a ‘culture’ or its
reprojection within the ‘given’ or natural world of fact, or even learn about
them, so much as we teach ourselves to them” (2001, xiii). Wagner’s
conclusion—that a “virtual anthropology of the subject ‘understood’ the
anthropologist better than the anthropologist could understand, or even
formulate, it” (ibid., xviii)—can be used to explain what is happening with
some of my own data. I “taught myself” to the culture—became
predisposed to certain emotions and dreams, through which in turn I
struggled to understand myself better.

Melanesian Odysseys
How, finally, does Melanesian Odysseys position itself differently
from my own earlier ethnography? In terms of ethnographic interests, it
explores strategies of making explicit rather than of covering up, depicting
people as they create their lives in complex negotiations within specific
situations. Having set the scene by first evoking Kewa lives and then
contextualising their stories within the larger human story of strategies of
self construction, the book unfolds in a series of epic self-narratives, from
traditional accounts where narrators epitomised the “cultural persons”
passing through life-cycles, to Christian epiphanies, and from picaresque
adventures to the traumas and transformations of lives in a changing
world. Expositions of how social knowledge was made explicit in the
process of negotiation and elicitation cast doubts on claims that
institutions, ideologies, or tradition (“rehearsed talk”) determine such
knowledge, or even social practice. When people’s multi-layered accounts
were allowed to dialogue, they revealed instead that evidence for what
becomes accepted as socially and culturally significant is gleaned from the
interaction of particular claims (“rehearsing talk”).
While discussion of self theory stressed communication and mutual
construction, deproblematising “otherness” to allow a space for
understanding, another discussion reproblematised the relationship by
focusing on the more recalcitrant aspects of the ethnographic encounter:
misunderstandings, awkwardness, ambivalence, resentment, resistance to
incorporation, asymmetric expectations. They reveal the real issues behind
the disjunctures as arising from fumbling attempts to establish rapport in
uncharted areas, but also from different perspectives and agendas. The
ethnography showed that intentionality and deviousness as well as
differences in personality had to be taken into account: people often found
196 Lisette Josephides

it as difficult to work out each other’s meanings as to negotiate the strange

anthropologist’s more predictable peculiarities. Key contexts defined the
encounter by providing insights into two operations: the emergence of a
subjective portrait of myself in terms of the relations I tried to establish in
the field, and the contribution (as I perceived it) of my local interactants in
defining these relations from the perspective of their own interests.
Virtually returning to my fieldsite while writing Melanesian Odysseys
became a consuming personal engagement. Now, rereading the first
evocative ethnographic chapter, which I have written, brings my
experience back to me, enriched. Evidence here is seen in everyday things,
made visible by the quality of the relationships. This is what authenticates
ethnography, as it depicts the reality in the field, where it was played out.
Without the ethnography, the story would have been lost.

I sought in this chapter to show how ethnographic evidence can be
found in recollections of fieldwork in tranquillity, through rereadings,
comparative readings, philosophical frameworks, dreams, subconscious
feelings, and images enclosed in memories (or material artefacts). These
discoveries lead to recontextualisations, which in turn reveal how
something at first considered part of a facilitating methodology (such as
emotions and personal relations) may also be part of ethnography.
Ethnographic collections then emerge as middens for perennial siftings.
Data and “facts” become evidence through interpretation, and interpretation
needs a framework, which moulds the facts; thus the decision to use a
particular framework becomes part of the evidence.
In this respect, the importance of physical returns is two-fold: they
establish reciprocal relations and impose on the ethnographer new ways of
looking at the field. In virtual returns, fieldwork materials are seen in a
synchronic form and acquire a systematic structure; ideas of what counts
as evidence change in the review, influenced by other readings and in turn
yielding new realizations of how others are known. My statement that
Kewa narratives are used to “exchange experiences of practical wisdom, to
make claims and to seek feedback” does not arise simply from Kewa
ethnography as a locally specific observation, but is a general statement
that reflects my expanded insight into the ethnography resulting from the
various “returns” discussed here. At the same time this synchronic picture
from afar captures and reprojects the ethnography as a “projection of a
world”. But this world has now been internalised by the ethnographer.
Even so, it never loses its character as a complete world, the product of a
Virtual Returns: Fieldwork Recollected in Tranquillity 197

sustained experience that describes other experiences by following their

interactions and logic.
The difference in the use of evidence in the two books is stark. While
the arguments and interpretations in The Production of Inequality were
authenticated by facts in an integrated synthesis, in Melanesian Odysseys I
need no such authority. Or rather, I do not have to speak authoritatively. I
am reproducing people’s talk and my engagement with them as a
narrative, a story that happened and relations that were established. It may
be argued that the latter are “facts” just as much as the former, and that my
endeavour has been to devise a way of approaching relationships and the
qualities inherent in them as data and thus as ethnographic “facts”. But this
would be erroneous. In the second case, as opposed to the first, I am not
“modelling” a social reality; I am presenting a description of actual life,
however much my understanding of it was informed by virtual returns.
The main conceptual tools employed in this presentation are not “modelled
on” Kewa practices; they actually are those practices. Eliciting,
negotiating, using emotions such as resentment to construct the self, are
not my models or my heuristic devices, but originate in fieldwork findings.
In so far as I talk of morality, I treat it as the result of emotions as
empathy, not the deontological ethics of a moral “ought”. What continues
to draw me back to my field materials, aside from sheer fascination and
the art of forgetting, is the hidden promise of discovering new
perspectives, rather than the search for a blueprint with which to model

I wish to thank the editors of this volume for their thoughtful
comments on the first draft of this chapter. The Kewa people, as ever, are
to be thanked for making all stages of my understanding possible.

At the time I argued that wives giving pork at pig kills were “distributing not
transacting”, because I had defined transaction as an activity that creates debts
towards oneself and concluded too quickly that women were not creating such debts
(Josephides 1985, 196).
There is no space here to do justice to Wagner’s “holographic modelling”, in
which the model becomes part of the reality it is modelling by reprojecting the
representation. In Wagner’s example (2001, 253), anthropological models for
198 Lisette Josephides

representing kinship take a key aspect of kinship—the proscription of incest—as its

basis, then reproject the artifice of modelling itself (how we “picture facts to
ourselves”). Hence, Wagner concludes, “kinship is not so much based on the
proscription of incest as upon its reprojection through legitimate means” (ibid.).
What these legitimate means may be, Wagner does not specify. Is he referring to the
empirical fact of such a proscription (that it is found to exist), the “legitimate”
reasons for its existence (social, biological, political…), or does he simply mean that
having already been selected as a fact makes it legitimate?
All poetry citations are taken from The Prelude. Wordsworth’s experiences, being
so dramatically and organically linked to nature, cannot be seen as deriving from a
philosophy that leads to scepticism about the external world. Nevertheless, David
Hume’s views on perceptions may shed some light on Wordsworth’s account of his
experiences. In his major work, A Treatise of Human Nature (1897), Hume
elaborates the view that the immediate objects of the mind are perceptions. He
distinguishes two kinds of perceptions: “impressions” and “ideas”. Impressions in
turn are of two kinds, of sensation and of reflection. Ideas are always causally
dependent on impressions (copies of impressions, as Russell [1946, 636] put it), and
though complex ideas (for instance, of a winged horse) need not resemble
impressions, the constituents of the idea are derived from impressions. Despite
being “faint images” of impressions, ideas combine memory (of impressions) and
imagination and have the potential to become very powerful. While Hume
acknowledged the existence of unknown external causes, he found it impossible to
establish reliable links between thought and reality.
There is an important difference, however, in that Wordsworth’s concern is with
his own spiritual development, with understanding the moral lessons of nature in
order to improve himself. It is also worth remembering that while Wordsworth treats
“recollection in tranquillity” as the higher good, a distillation of the sublime, in his
ode on Intimations of Immortality it is the original experience—from childhood and
youth—that is seen as the wellspring of glory, with whose passing a glory departs
from the earth.
Crapanzano seems to take this perspective in his critical discussion of how
Wordsworth creates “horizons” and a “sense of the beyond”. He concludes that
Wordsworth “teases” us with symbolic significance, but the tease often fails,
“because it is reduced, often crudely, to the personal” (2004, 28).
Crapanzano believes that whenever a horizon and what lies beyond it are given full
articulate form, “they freeze our view of the reality that immediately confronts us”
(2004, 2). “Ethnographies are themselves constructions of the hinterland”, that
imaginative horizon (ibid., 23); but once articulated, this hinterland becomes “fixed
and constraining”, pushing us to seek new possibilities, new hinterlands (ibid.). My
articulation of the lifeworld of the Kewa from the imaginative horizons in The
Production of Inequality, though novel when first viewed, had become constraining
following my virtual returns. It pushed me to a new horizon in Melanesian
Odysseys, when the Kewa became evidence for an existential ethnography of social
and cultural change. Emotions and dreams are part of the landscape of “imaginative
Virtual Returns: Fieldwork Recollected in Tranquillity 199

According to Husserl’s phenomenological reduction, all empirical existential
considerations and all a priori assumptions about entities external to experience are
suspended so that we may focus on what is immanently given in our own stream of
experience. This phenomenological reflection transforms my own consciousness
into a stream of “transcendentally purified” experiences (Husserl 1931 [1913], cited
in Pivcevic 1970, 66). Husserl also believed that “[b]etween consciousness [res
cogitans] and reality [res extensa] there yawns a veritable abyss of meaning” – and
thus that subject and object are distinguished ontologically (Husserl 1931, 153, cited
in Heidegger 1982 [1975],124).
Recent research on the emotions investigates how emotions can be used as
ethnographic evidence (Josephides 2005; Milton and Svašek 2005).
See Josephides 1997 for a discussion of the relationship between the field
encounter and the anthropologist’s epistemological commitments.

Bourdieu, Pierre. 1977. Outline of a theory of practice. Cambridge:
Cambridge University Press.
Crapanzano, Vincent. 2004. Imaginative horizons. An essay in literary-
philosophical anthropology. Chicago: Chicago University Press.
Fabian, Johannes. 1990. Power and performance: Ethnographic
explorations through proverbial wisdom and theatre in Shaba, Zaire.
Madison: University of Wisconsin Press.
Favret-Saada, Jeanne. 1980. Deadly words: Witchcraft in the Bocage.
Cambridge and Paris: Cambridge University Press and Editions de la
Maison des Sciences de l’Homme.
Heidegger, Martin. 1982 [1975]. The basic problems of phenomenology.
Trans. Albert Hofstadter. Rev. ed. Bloomington and Indianapolis:
Indiana University Press.
Hume, David. 1897 [1888]. A treatise of human nature. Reprinted from
the original edition in three volumes, 1739-1740. Oxford : Clarendon
Husserl, Edmund. 1931 [1913]. Ideas pertaining to a pure phenomenology
and phenomenological philosophy. Trans. W.R. Boyce-Gibson.
London: Macmillan.
Josephides, Lisette. 1982a. Suppressed and overt antagonism: A study in
aspects of power and reciprocity among the Northern Melpa. Research
in Melanesia Occasional Publication No. 2. Port Moresby: University
of Papua New Guinea.
—. 1982b. Kewa stories and songs. Oral History 10 (2).
200 Lisette Josephides

—. 1985. The production of inequality: Gender and exchange among the

Kewa. London: Tavistock.
—. 1997. Representing the anthropologist’s predicament. In After writing
culture: Epistemology and praxis in contemporary anthropology, ed.
Allison James, Jenny Hockey, and Andrew Dawson. London:
—. 2003. Being there: The magic of presence or the metaphysics of
morality. In The ethics of anthropology: Debates and dilemmas, ed.
Pat Caplan. London: Routledge.
—. 2005. Resentment as a sense of self. In Mixed emotions:
Anthropological studies of feeling, ed. Kay Milton and Marushka
Svašek. Oxford: Berg.
—. 2008. Melanesian odysseys. Oxford: Berghahn.
Metcalf, Peter. 2002. They lie, we lie: Getting on with anthropology.
London: Routledge.
Milton, Kay, and Marushka Svašek, eds. 2005. Mixed emotions:
Anthropological studies of feeling. Oxford: Berg.
Pivcevic, Edo. 1970. Husserl and phenomenology. London: Hutchinson
and Co.
Ricoeur, Paul. 1970 [1965]. Freud and philosophy: An essay on
interpretation. Trans. Denis Savage. New Haven: Yale University
—. 1981. Hermeneutics and the human sciences. Ed. and trans. John B.
Thompson. London and Paris: Cambridge University Press and
Editions de la Maison des Sciences de l’Homme.
Russell, Bertrand. 1946. History of western philosophy. London :George
Allen and Unwin.
Wagner, Roy. 2001. An anthropology of the subject: Holographic
worldview in New Guinea and its meaning and significance for the
world of anthropology. Berkeley and Los Angeles: University of
California Press.
Wordsworth, William. 1802. Preface to Lyrical Ballads, 2nd ed. In
Selections from William Wordsworth. London: Methuen, 1966.


One Saturday morning in 1980, I found myself addressing a large,

mainly black working class audience in Detroit on the subject of the
Atlantic slave trade. I was claiming that, before the industrial revolution,
relations between Europeans and Africans on the Guinea Coast were
relatively equal. The Africans supplied the slaves and the Europeans
bought them. At this point, a huge man sitting on the front row suddenly
stood up:

Man: You sayin’ we slaved our own people, motherfucker?

Keith: Well, yes, I suppose … yes.
Man: How you know?
Keith: Oh, historical records—Portuguese, Dutch, French, British records.
Man: Any black people write them records?
Keith: Ah yes, good point. We do rely mainly on what white people wrote

He sat down and, rather shaken, I continued with my lecture. I start

with this story because, although the contributors to this volume have
brought up the scientific, forensic, and self-referential dimensions of
evidence in a variety of media, I miss an explicit engagement with the
politics of evidence, outside as well as inside the universities.
Why does evidence matter? It clearly does to the young
anthropologists who convened the workshop on which this book is based
and to others who did much the same thing not long before (Engelke
2008). In this Afterword I will try to explain why, for reasons of
institutional politics, a group of apprentice anthropologists from around
the world, temporarily inhabiting the shell of what was once British social
anthropology, might be pre-occupied with questions of evidence at this
time. But there are more general considerations too. What links the two
levels is the problem of intellectual authority in a world where respect for
it is dwindling fast (Grimshaw and Hart 1993).
I was once asked to launch a series of Key Debates in Anthropology
(Ingold 1996) by proposing the motion “Anthropology is a generalizing
202 Keith Hart

science or it is nothing”. Wanting to win and knowing my audience, I cast

around for the most ecumenical definition of “science” I could find, since
any hint of the men in white coats would be a disaster. (Our side lost in
any case.) I argued that the eighteenth-century Enlightenment invented
modern anthropology as a means to a revolutionary political end, the
institution of “democracy” at the expense of the Old Regime. People
power could only succeed if it was based on sound knowledge of the
world; so “science” was promoted as an alternative to religion, myth,
superstition, etc., all of which sustained unequal society. The first reason
evidence matters is that it is central to the scientific project; this in turn is
indispensable to social and material improvement of the human condition.
Science is about getting something right over and over again within an
acceptable range of error; and that means systematically juxtaposing the
rules of knowledge and the evidence of practice. Otherwise the bridge
collapses, you miss the moon, or whatever. The consequences for human
life of getting medical research right are obvious and this does sit uneasily
with conventional ethnography (Kelly, this volume).
My Detroit encounter was an object lesson in the truism that the rules
of engagement normally observed in the academy don’t hold outside it.
My life as an anthropologist has been shaped by the politics of race, often
in situations where I was not only unprotected by my academic expertise,
but actually suspect because of it (and hopelessly outnumbered). Evidence
matters under these circumstances because a claim to objectivity made in
ordinary language may be the best means of self-defence. (Convincing
people of one’s humanity helps too.) The founders of twentieth-century
ethnography—Boas, Rivers, and Mauss—all deliberately engaged with
political questions of the day through their anthropological knowledge.
Mauss in particular was absolutely clear about how sociological science
supported his career as a political activist (Hart 2007). It is symptomatic of
what happened next that the leaders of British social anthropology
declined to take part in the UNESCO book on race (UNESCO 1956) on
the grounds that their discipline’s academic standing as a social science
would be compromised by getting involved in “controversial” topics
(Verena Stolcke personal communication).
The forensic uses of evidence often have legal consequences that can
be particularly severe for individuals. At the same time, this rhetorical
practice has wide application inside and outside the courtroom, with
effects across a continuum ranging from the catastrophic to the trivial.
Marilyn Strathern (this volume) explores, through examples like academic
promotion procedures and women’s credibility as witnesses in Mt Hagen,
how evidence actually works in discourse and social practice. Here I
Afterword: Malinowski’s Heirs 203

would just make the point that the burden placed on evidence is relative to
the consequences of the decision made or the story told. For instance,
readers of a newspaper story about “man bites dog” will demand a
stronger line of evidence than for the commonplace reverse.
The same goes for issues implicating professional authority. Thus, as
part of his unsuccessful campaign to impress the Freudians, Malinowski
(1929) asserted that the Trobrianders were ignorant of physiological
paternity. This was later refuted in Man by a colonial administrator who
had spent sixteen years in Papua and cited various kinds of evidence,
including the Trobrianders’ practices of contraception and animal breeding
(Rentoul 1931). Malinowski replied with fury, mentioning his being
described as a “visiting anthropologist” seven times in five pages and
complaining haughtily that Rentoul “sweepingly disposes of the
qualifications of the scientific specialist” (Malinowski 1932, 34). At least
the world cared about the truth value of Malinowski’s story. The vast bulk
of ethnography produced today is of interest only to professional insiders
and their students. I think this should be made explicit when discussing
questions of evidence.
But of course life carried on within academic seclusion has its politics
too (Cornford 1908); and this, it seemed to me then and now, played a
significant part in the dynamics of the original workshop. The editors’
introduction to this volume shows how their idea of the event evolved into
something more complex and eventually into this book. In short, a small
network of anthropology PhD students and postdocs in London’s
hinterland (Cambridge in particular) conceived of a one-day workshop
organised by and for people like themselves, with one or two established
scholars as invited guests. The response of young and established scholars
alike was larger than expected and the workshop expanded into a two-day
weekend event. By the time my slot arrived at the very end, the audience
had reverted to its original expected character. So I tried to address the
question at issue from the perspective of people writing a doctoral thesis in
anthropology or having just completed one. I was particularly sensitive
throughout the workshop to clues about why evidence mattered to them.
Almost all the younger participants had been forced by departmental
convention to write a thesis employing the methods of fieldwork-based
ethnography. The founders of the British school they had joined so
recently produced a stream of exemplary monographs in this genre
between the wars and just after. I have long been convinced that, when the
history of twentieth-century thought is written, the ethnographies of
Evans-Pritchard, Firth, Fortes, Lienhardt, Nadel, Richards, Turner, and
many others will be celebrated for the extraordinary achievements that
204 Keith Hart

they were. Their successors—trained in the 1960s and early 1970s, blessed
with the golden age of university expansion and still broadly in charge,
although now on the edge of retirement—maintained the tradition, while
also branching out into several other genres of research and writing. (I
must confess that I never managed to complete an ethnographic
monograph myself.)
Since the 1980s, the institutional context of anthropology in Britain
and the world more generally has been transformed in ways that are both
too familiar and complex to summarise here (Grimshaw and Hart 1995;
Hart 2005). What anthropologists do professionally today is almost
unrecognisable when compared with those classical monographs. Yet they
still cling to a guild identity founded on fieldwork and an increasingly
loose invocation of the term “ethnography”. I believe that a concern for
“evidence” arises from a shared experience of the dilemmas these young
anthropologists encounter when trying to establish themselves within such
a tradition.
Let me be clear from the outset. Choosing to live with the people in
order to discover what they do and think was a radical political move, fully
compatible with anthropology’s birth in the Enlightenment’s drive to
apply science to democratic ends. Read again the Introduction to African
Political Systems (Fortes and Evans-Pritchard 1940) to get a sense of this
politics. Hitherto, they say, political philosophy has been the province of
European elites with no direct experience of how the mass of humanity
lives. We have to abandon our libraries, laboratories, and seminar rooms
(if only temporarily) to discover what is really going on out there and to
develop concepts honed in empirical investigation. This is anthropology’s
deep political purpose, to shake up ruling ideas in the name of a wider
human agenda. Marx (in Capital, 1867) refers to ideology as a camera
obscura that turns the image upside down. Ideas come from life, but it is
the task of ideologues to persuade people that it is the other way round,
that their lives depend on ideas produced and controlled by an expert elite.
It was Malinowski’s supreme achievement to embody a model of
ethnographic practice where ideas have to be derived from life, from
systematic observations made in all corners of the world. That is why the
classical British school has a unique part in the history of modern
anthropology. No-one has ever done it better. The premise is of course
naïve and epistemologically problematic (Grimshaw and Hart 1995). But
as a riposte to the lingering legacy of the Old Regime, the move was
priceless and should be conserved at all costs.
Writing a doctoral thesis has a claim to being the worst job in the
world. Its demands throw students onto their own resources in lives of
Afterword: Malinowski’s Heirs 205

lonely isolation; yet they are subject to a formal hierarchy that might
literally stop them from ever submitting their work or might reject it
utterly when they do. There is nothing new about that. What is new is that
the intellectual practice anthropology students are expected to follow is
now contradicted by the conditions of the job market they hope to enter
and by the prevailing activities of the profession at large. For the same
teachers who invoke the legacy of Malinowski and Firth routinely advance
their own careers by literary means that owe almost nothing to that legacy.
“The field” for most anthropologists occupies a small part of our
working lives. The number of truly ethnographic monographs we produce
is even smaller. We spend any time spared from teaching and
administration cranking out conference papers, journal articles, and book
chapters, usually with an upper limit of less than ten thousand words. The
whole aim of this writing is to come up with a new argument relevant to
“the literature” we presumptively have in common. There is precious little
room here for anything like ethnographic description. Instead the camera
obscura of ideology rules in an inverted pyramid: scattered soundbites
drawn from life illuminate baroque arguments of unfounded generality, all
delivered with a long list of references to shore up the author’s claim to
erudition. Even while students are labouring to write a thesis on the
Malinowskian model, they are being encouraged to emulate this
institutional practice, since, without publications, their prospects of
academic employment would be even lower than they already are.
The literary turn in contemporary anthropology (Clifford and Marcus
1986) feeds into how students approach writing a thesis. Many of them
believe they are writing a book, whereas the highest value of a thesis is its
methodological transparency. The few people who ever read it will want to
extract bits they can use for their own research purposes. They need to
know if they can trust these snippets, which means that how they were
gained should be made unusually explicit. This privileges bad writing over
good and it would never do for a book, where literary convention favours
elegance. I was told as a graduate student that a good thesis or monograph
should not be driven by its theoretical ideas. Rather the reader should be
able to reach conclusions independent of those of the author, drawing on
an abundance of empirical material that has not been shaped merely by the
need to illustrate an abstract argument. According to Bruce Kapferer
(2005), Max Gluckman took this principle even further to a pedagogical
practice where students were encouraged to critique the published
ethnographies of their teachers, who would then be forced out into the
field to make good the damage and try again. However idealised a portrait,
206 Keith Hart

this does stand in contrast to the spirit of anthropological scholarship

The importance of evidence for apprentice anthropologists operating
within the leftovers of the British school should perhaps be seen in this
light. Why should they sacrifice their abstract intellectual ambitions to a
realist paradigm when none of their teachers do? Here again, Marilyn
Strathern’s chapter illustrates what we all do when we reach a position of
some security. If asked to contribute to a theme, we assemble a few ideas,
string them together into an argument, and illustrate it with examples
drawn from our whole life experience, whether based on “fieldwork” or
not. This Afterword is another example of the genre. I actually think that
the idea of ethnography as a practice drawing on all of life, not just some
artificial compartments labeled “research”, has much to be said for it.
Strathern’s essay is proof of the genre’s potential. But we would never
allow a PhD student to try the same thing, and postdocs won’t get far by
employing this strategy either. Could this contradiction be one source for
the current fascination with questions of evidence?
The workshop’s organisers, Timm Lau in particular, let the cat out of
the bag when he spoke in some introductory remarks about a dialectic of
“evidence” and “protocols”. This last term is significant. It comes from the
politics of the internet and computing practice. Some now argue (e.g.
Lessig 1999) that control is no longer established by making the rules, but
by writing the code, the “protocols” that invisibly shape what practitioners
can actually do. The “evidence/protocols” pair seemed to me a metaphor
for the student/teacher dialectic. There was a time when any ethnographer
had a huge chunk of the world to him- or herself. The world these students
encounter is crowded with burdensome precedents, established ways of
“knowing” the Dogon or whoever. Their teachers are closer to these
traditions, more retentive of the past than the students. Lau was explicit in
raising the possibility that “evidence” (“I have been there more recently
than you and it isn’t like that any more”) could be used to break out of
restrictive “protocols” that often seemed arbitrary and unjust. The
workshop was dotted with other examples of this move. It is linked to the
emphasis on reflexivity and the use of new media. If truth is in fact
continuously renewed on a reflexive basis or transformed by different
technologies, the power of tradition to constrain present action is thereby
reduced. But can they get away with it? An appeal to the evidence,
whatever that is these days, might be one way of fighting the underdog’s
I make this inference partly because my own doctoral research
involved studying the Tallensi made famous by the head of my
Afterword: Malinowski’s Heirs 207

department, Meyer Fortes (e.g. Fortes 1949). He began his fieldwork in

Northern Ghana in 1934 and I in 1967, as part of a rural-urban study of
migrants. Apart from being shamed by the sheer virtuosity of his fieldwork
technique, I found it almost impossible to place my findings in some
framework of comparison with his. Suppose I observed a ritual that he
described in one of his writings and found a difference. Was this to be
accounted for as a change in practice through time, or as empirical error or
bias on the part of either ethnographer? Such questions were fascinating in
a recursive way, but I found them to be ultimately fruitless (but see Hart
1978). They raise the thorny issue of the relationship between ethnography
and history, as touched on in my opening anecdote (see also the
Introduction and High, this volume). Anthropologists do far more than
fieldwork (Hart 2004) and it is time that we abandoned the ostrich-like
pretence that our methods have not moved on since Malinowski founded
the modern British school.
This admirable event and the book that flows from it confront squarely
what we might consider to be good ethnographic practice. Our methods
have always been substantially occult. The practice of hoarding
“fieldnotes” in private places where they can never be inspected by other
professionals is indefensible. We all know it, but we keep quiet. The
pretence that we “learn the local language” in a year or so covers up tacit
acknowledgment that our linguistic competence might reach that of a nine
year-old, but rarely more. Those who commit to a discipline called
“ethnography” live in constant fear of being found out; and this is
exacerbated by a revolution in transport and communications that is
rapidly collapsing the walls anthropologists once erected between their
academic base and “the field”. How much more intolerable it is then for
apprentice ethnographers to learn to live with the contradictions of this
inherited norm under social conditions that by most measures are
deteriorating, when compared with the heyday of the universities in the
1960s and 1970s. I offer, without much evidence, the suggestion that one
impulse to this collective effort lies in such pragmatic dilemmas.
However we explain the metaphysics of contemporary anthropological
practice, the factors that sometimes lend urgency to the search for reliable
evidence were not the main focus of concern on this occasion. Rather the
object of the exercise seems to have been a mixture of concern for
professional authority and for existential truth. The former is defensible,
even laudable in apprentice scholars; but I find it less interesting than the
latter. There has always been an element of the “lone ranger” in
anthropologists, a sense of making a personal journey in the world that
offers infinite scope for reflection. We move inwards as we get older and I
208 Keith Hart

would certainly not deny the value of such an exercise. Edward Said once
suggested (in a 1992 television programme) that life gives us so many
cultural fragments and our task is to make a story out of them. It is not
quickly done, as Lisette Josephides shows (this volume). My gloss on this
sentiment would be Durkheimian: fieldwork is one of many opportunities
to internalise experience of society and our job is to get some of it out
eventually in a coherent form that we can share with others. My own
explorations of the “informal economy” (Hart 1973) came from interacting
with development economists who were then obsessed with the problem of
“urban unemployment”. This term did not sit well with my fieldwork
experience in an Accra slum, but it took some time to figure out why and
even longer to find an alternative vehicle to convey the results of my
ethnography to economists. Questions of evidence were definitely
secondary in this process of excavation.
A preoccupation with self and the world has its roots in humanism, but
it is not science and it lacks the political force of good science. I have
argued here that anthropologists should be more explicit in confronting the
politics of evidence. Maybe it would help if this generation of
Malinowski’s heirs tried being more honest than their predecessors. In the
process they might renew anthropology’s mission to engage with the
deepest political questions concerning humanity’s past, present, and

Clifford, James, and George Marcus, eds. 1986. Writing culture: The
poetics and politics of ethnography. Berkeley: University of California
Cornford, Francis. 1953 [1908]. Microcosmographia academica: Being a
guide to the young politician. London: Bodley Head.
Engelke, Matthew. 2008. The objects of evidence. In The objects of
evidence: Anthropological approaches to the production of knowledge,
ed. Matthew Engelke. Journal of the Royal Anthropological Institute
Special issue:S1-S21.
Fortes, Meyer. 1949. The web of kinship among the Tallensi. London:
Oxford University Press.
Fortes, Meyer, and Edward E. Evans-Pritchard, eds. 1940. African
political systems. London: Oxford University Press.
Grimshaw, Anna, and Keith Hart. 1993. Anthropology and the crisis of the
intellectuals. Cambridge: Prickly Pear Pamphlets No. 1.
Afterword: Malinowski’s Heirs 209

—. 1995. The rise and fall of scientific ethnography. In The future of

anthropology: Its relevance to the contemporary world, ed. Akbar S.
Ahmed and Cris Shore. London: Athlone.
Hart, Keith. 1973. Informal income opportunities and urban employment
in Ghana. Journal of Modern African Studies 11 (3): 61-89.
—. 1978. The economic basis of Tallensi social history in the early
twentieth century. In Research in economic anthropology, Vol. 1, ed.
George Dalton. Greenwich, CT: JAI Press.
—. 2004. What anthropologists really do. Anthropology Today 20 (1): 3-5.
—. 2005. The hit man’s dilemma: Or business, personal and impersonal.
Chicago: Prickly Paradigm Press.
—. 2007. Marcel Mauss: In pursuit of the whole. Comparative Studies in
Society and History 49 (2): 473-485.
Ingold, Tim, ed. 1996. Key debates in anthropology. London: Routledge.
Kapferer, Bruce. 2005. Situations, crisis and the anthropology of the
concrete. Social Analysis 49 (3): 87-125.
Lessig, Lawrence. 1999. Code: And other laws of cyberspace. New York:
Basic Books.
Malinowski, Bronislaw. 1929. The sexual life of savages in North-western
Melanesia. London: Routledge and Kegan Paul.
—. 1932. Pigs, Papuans, and police court perspective. Man 32:33-38.
Marx, Karl. 1970 [1867]. Capital: The critique of political economy,
Volume 1. London: Lawrence and Wishart.
Rentoul, Alex. 1931. Physiological paternity and the Trobrianders. Man
UNESCO. 1956. The race question in modern science. New York:
UNESCO and William Morrow.

Anna-Maija Aguilera Calderòn is a PhD student in Social Anthropology

at Brunel University, working on Animism and Modernity in a Southern
Beninese Village.

Liana Chua is a Research Fellow in Social Anthropology at Gonville and

Caius College, Cambridge University, where she completed her PhD in
2007. She works on new ethnic formations, religious conversion,
materiality, place and memory in Sarawak, Malaysian Borneo.

Carlo A. Cubero received his PhD in Social Anthropology using Visual

Media from the University of Manchester in 2007. His dissertation was
about social processes that lead to the constitution of Caribbean island
identities. The ethnographic documentary that he made as part of his
doctoral project has been screened at various film festivals in Europe and
the Americas. He is currently an independent researcher and filmmaker
based in Luxembourg.

Keith Hart lives in Paris; he is Honorary Research Professor in the

School of Development Studies, University of Kwazulu Natal, Durban
and, at the time of writing, Professor of Anthropology, Goldsmiths,
University of London. He contributed the concept of the informal
economy to development studies. His books include The Memory Bank:
money in an unequal world (2000) and The Hit Man's Dilemma: or
business, personal and impersonal (2005). Many of his recent writings
may be found at

Casey High is a lecturer in anthropology at Goldsmiths College,

University of London. In 2006 he completed his PhD at the London
School of Economics, entitled From Enemies to Affines: Conflict and
Community among the Huaorani of Amazonian Ecuador. His research
interests include violence, historical anthropology, indigenous rights, and
How Do We Know? Evidence, Ethnography, and the Making 211
of Anthropological Knowledge

Lisette Josephides teaches anthropology at Queen’s University Belfast,

having previously taught at the London School of Economics and the
Universities of Papua New Guinea and Minnesota. Her research interests
are broadly in philosophical approaches in anthropology, especially
phenomenology and moral philosophy, and include human rights,
emotions, narrative genres and theories of the self, interactive
communicative practices, and local and anthropological knowledge. She
has published widely on the Kewa people of the Southern Highlands of
Papua New Guinea, including the monographs Suppressed and overt
antagonism (1982), The Production of Inequality (1985) and Melanesian
Odysseys (2008).

Ann Kelly is a Wellcome Trust Postdoctoral Fellow in the Health Policy

Unit at the London School of Hygiene and Tropical Medicine. She is
currently conducting a comparative ethnographic investigation of two
large-scale malaria control projects in The Gambia and Tanzania. Recent
publications include “Research and the Subject: The practice of informed
consent”, (PoLAR 2003); “Pragmatic Clinical Research: C.S. Peirce and
the Convergence of Truth” (Kilifi Conference Volume, in press), “‘He is
now like a brother, I can even give him my blood’—relational ethics and
material exchanges in a malaria vaccine ‘trial community’ in The Gambia”
(Social Science and Medicine, in press); and “Global Clinical Trials and
the Contextualisation of Research” (Collaborators inside Collaboration:
Comparability and the Ethics of Collaborative Knowledge Relations
across Worlds of Research-in-Action, in press).

Timm Lau studied Ethnology in Hamburg, Germany, and Social

Anthropology at the London School of Economics and Political Science
(MSc 2000) and Cambridge University (PhD 2007). His doctoral research
discusses Tibetan sociality in India as it arises from the specifically Indian
diasporic context. It provides an in-depth analysis of how itinerant trade
markets and pop-cultural media such as Hindi film influence identities and
social practice among Tibetan exiles in India. His research interests
include changing notions of romance and marriage in the diaspora,
itinerant trade markets, emotional subjectivity, and the nature of evidence
in anthropological inquiry.
212 Contributors

Fraser McNeill recently received his PhD from the University of London
for a thesis entitled An Ethnographic Analysis of HIV/AIDS in the Venda
Region of South Africa: Politics, Peer Education and Music. He currently
holds a post-doctoral fellowship in the Department of Anthropology at the
London School of Economics, and continues to conduct ethnographic
research in South Africa.

Marilyn Strathern: PhD (Cambridge); Fieldwork: Papua New Guinea.

Currently Professor of Social Anthropology at the University of
Cambridge and Mistress of Girton College. Publications include The
Gender of the Gift (1988), After Nature (1992), the co-authored
Technologies of Procreation (Edwards et al, 1993), Property, Substance
and Effect (1999), the edited volumes Audit Cultures (2000), and
Transactions and Creativity (Hirsch and Strathern 2004), and most
recently Kinship, Law and the Unexpected (2005).

Emma Varley received her Master’s degree in Anthropology at the

University of British Columbia (2002), and is completing her Doctoral
dissertation in Anthropology at the University of Toronto. Her chapter
“‘Halaat Kharab’/Tension Times: The Maternal Health Costs of Gilgit’s
Shia-Sunni Conflict” was published in the edited anthology Missing Links
in Sustainable Development: South Asian Perspectives (Sustainable
Development Policy Institute’s, Islamabad, 2008). She has co-authored the
course textbook Working in International Development (IHHS, College of
Health Disciplines, University of British Columbia, 2000), and written
numerous policy and impact studies during overseas consultancies.
Recently, she has worked as Co-Investigator on a qualitative research
project on the impacts of the December 2004 Tsunami for health service
providers working in southern Thailand (“Health Care, Technology and
Place” grant, University of Toronto), and contributed to upcoming book
projects on the confluence of Islamic conservatism, women’s rights, civil
and sectarian conflict in Pakistan (Sustainable Policy Development
Institute, Islamabad).

!Kung, 63 analogy, 10, 11, 13, 15, 16, 25, 27,

abortion, 51, 147 29, 30, 31, 32, 87, 88, 99, 100,
academia, 173, 201, 202, 205 109, 110, 111, 143, 172, 194
career in, 16, 179, 194, 205 analogic reasoning, 23, 24, 29,
politics in, 201, 203 30, 31, 100
publishing, 173 ancestors
university promotions exercises, in Papua New Guinea, 24
10, 20 in repatriation debates, 26, 27
Africa, 63, 78, 126 in Venda, South Africa, 36, 39,
capitalism in, 126, 127, 128, 45
129, 130 Andes, the, 79
Guinea coast, 201 anthropologists, 4, 6, 7, 13, 14, 15,
African Political Systems, 204 16, 17, 20, 21, 22, 30, 59, 60,
Aga Khan, 137 62, 64, 66, 73, 74, 77, 78, 79,
Aku Tonpa, 171, See also Tibetans 83, 86, 88, 90, 97, 98, 99, 100,
Amaqabane, 43, 44 106, 111, 127, 133, 134, 135,
Amazonia, 8, 78, 79, 80, 89, See 136, 137, 151, 153, 166, 172,
also Ecuador, Huaorani 173, 181, 182, 184, 185, 187,
Amazonian studies, 8, 78, 79, 188, 190, 191, 193, 194, 196,
80, 83, 84, 89 199, 201, 202, 204, 205, 206,
ethnographies, 79 207, 208
ethno-historical research, 79 selfhood, 13, 133, 134, 135, 136,
theoretical models, 78, 79, 145, 151, 152, 157, 186, 188,
80, 87, 88 189, 191, 192, 196, See also
ontological predation, 8, emotions, reflexivity,
80, 84, 86, 87, 89 subjectivity
ontology, 78, 79, 80, 88 anthropology, 1, 2, 3, 4, 5, 6, 9, 10,
symbolic cannibalism, 80, 11, 12, 13, 14, 16, 17, 18, 19,
84, 89 62, 65, 74, 78, 80, 85, 90, 98,
Christianity in, 79 100, 106, 110, 111, 150, 172,
colonialism, 79, 89 173, 202, 203, 204, 205, 206,
development in, 79, 83 207, 208
history, 79, 80, 83, 88, 89, 90 applied, 10
cannibalism, 89 British social anthropology, 201,
indigenous groups, 8, 76, 78, 79, 202, 203, 204, 206, 207
80, 83, 88, 89, 90 comparison, modes of, 21, 22,
Yanomami, 83 30, 31, 32
state formation, 79 doctoral work, 203, 204, 205,
violence and warfare, 79, 80, 83, 206
87, 89 ethnography, 1, 2, 3, 4, 5, 6, 7, 9,
11, 12, 13, 14, 15, 16, 17, 18,
214 Index

35, 38, 62, 66, 75, 77, 88, 89, See also ARC SCOOP clinical
97, 98, 99, 100, 101, 106, trial
108, 109, 110, 111, 118, 121, atherosclerosis, 100, 106, 107, 108,
127, 134, 135, 136, 154, 155, 109
156, 158, 164, 170, 171, 179, Atlantic slave trade, 201
185, 186, 187, 188, 193, 194, audiovisual media, 7, 9, 58, 59, 60,
195, 196, 198, 202, 203, 61, 62, 63, 64, 73, 74, 194
204, 207, 208 film. See film
evidence in. See evidence music. See music
fieldwork, 1, 2, 4, 5, 6, 7, 9, 11, audit, 106
13, 14, 15, 16, 17, 29, 32, 54, Balázs, Béla, 63
58, 59, 60, 61, 62, 64, 66, 67, Bali, 61
69, 70, 72, 73, 74, 77, 78, 79, hierarchical language in, 163
80, 85, 88, 90, 99, 102, 106, Barthes, Roland, 150
112, 118, 119, 120, 121, 122, Bateson, Gregory, 61, 62, 63, See
129, 133, 134, 135, 136, 137, also Margaret Mead
138, 143, 146, 148, 149, 150, Benin, 12, 118, 119, 120, 122, 123,
151, 152, 153, 155, 157, 161, 124, 129, 130
163, 164, 166, 167, 172, 174, Cotonou, 124, 125
179, 181, 182, 184, 185, 186, gold mining, 128
187, 189, 190, 192, 193, 194, government, 119
195, 196, 197, 203, 204, 206, Guen-Mina people, 122
207, 208 history of trade, 128
recollected in tranquillity. See le Bureau National de Vodou.
recollections in See vodou in Benin
tranquillity Marxist-Leninist regime, 118,
history of, 4, 17, 60, 62, 202, 204 119, 121
salvage ethnography, 61 oil palm plantations, 128
knowledge. See knowledge pomé (ancestral family), 122
linguistic, 37 presidential elections, 123, 124
medical, 97, 98, 106, 111 slave trade, 128, 129
methodology, 4, 10, 16, 17, 31, tradition, 118, 119
37, 58, 59, 61, 66, 99, 100, vodou in. See vodou in Benin
106, 134, 137, 187, 194, 196, Western influences in, 118, 123,
204, 207 126, 128, 130
political engagements, 11, 202, biomedicine, 98, 99, 102, 108
208 Bloch, Maurice, 21
theory. See theory BMG, record company, 43
visual, 7, 65 Boas, Franz, 202
ARC SCOOP clinical trial, 11, 99, Body Multiple, The, 100, 107, 109
103, 104, 105 Botswana, 51
archive, 9, 90 Bourdieu, Pierre, 107, 134, 181,
Aristotle, 187 182, 183, See also practice
art, 90 habitus, 107, 163
Arthritis Research Campaign British Broadcasting Corporation,
(ARC), UK, 11, 99, 103, 104, 27
How Do We Know? Evidence, Ethnography, and the Making 215
of Anthropological Knowledge

broadband, 2 in the United States, 76, 77, 82

Burkina Faso, 126 missionary discourse, 8, 77, 78,
Buthelezi, Mangosuthu, 43 80, 82, 85, 86, 87, 88, 89, 90
Caliph Ali, 139 cinema. See film
Cambridge, University of, 20, 29 cine-trance. See Rouch, Jean
1898 Expedition to the Torres civil society, 111
Strait, 61 clinical practice, 11, 98, 101, 102,
Canada, 140, 141 103, 105, 108, 109, 110, See
capitalism, 128, 129 also medical practice
Caribbean, 2, 6, 60, 65, 68 “real clinic”, 103, 104
Culebra, 6, 60, 65, 66, 67, 68, intervention, 99, 102, 103, 105,
69, 70, 72 106, 107, 108, 109, 110
culebrense, 68 clinical research, 99, 100, 102, 110,
development on, 6, 66, 67, See also medical research
68, 69, anthropologists in, 99
discourses of mobility, 65, 72 clinical trials, 2, 11, 12, 99, 103,
filmmaking on, 6, 66, 67, 69 110, 111, See also evidence-
Mangrove Music, 60, 70, based medicine (EBM)
72, See also film pragmatic, 2, 99, 100, 101, 102,
government and politics, 66, 103, 105, 106, 107, 110, See
67, 68 also ARC SCOOP clinical
identity, 65, 68, 72 trial
land struggles, 1970s, 68 randomised controlled trials
landscape, 60, 66, 67, 68, 72, (RCT), 101, 102, 105, 108
73 Cochrane, Archie, 101, See also
music, 60, 72 clinical trials, evidence-based
The Wiki Sound, 70, 71, medicine (EBM)
72 Comaroff, John and Jean, 127
tourism, 67, 68 computing, 206
eco-tourism, 67 consumerism, 118, 126, 129, 130
Puerto Rico, 66, 67, 68, 70 context, 22, 90, 111, 130, 152, 184
colonial politics, 67 contraception, 51, 52
landscape, 67 Creationism, 21
San Juan, 70 Cucapá curse-words, 2
child psychologists, 27 culture, 182, 186, 187, 191, 195
childhood concept of, 22, 181
Tibetan. See Tibetans Dalai Lama. See Tibetans
Chokyi Dronma, princess of dance, 27, 40
Gungthang (1422-1455). See data, 6, 22, 102, 104, 121, 137, 183,
Tibet 184, 191, 192, 193, 194, 195,
Christianity, 195 196
in Ecuador. See Ecuador, death, 39
Huaorani DeMan, Paul, 150
in End of the Spear, film. See democracy, 202
End of the Spear
in Europe, 76
216 Index

DEXA scan, 103, 104, 110, See also Ecuador, film, Huaorani, United
ARC SCOOP clinical trial, See States
also ARC SCOOP clinical trial Every Tribe Entertainment, 82
Diemberger, Hildegard, 161 filmmakers, 77
disease, 100, 104, 106, 107, 108, Huaorani reaction to. See
109, 110 ONHAE
materiality of, 107, See also The martyrdom narrative, 82, 85, 86,
Body Multiple 89
dreams, 2, 16, 189, 192, 194, 195, Engelke, Matthew, 9
196 Englund, Harri, 127
Durkheim, Emile, 208 Enlightenment, the, 202, 204
eco-tourism, 67 epistemology, 5, 9, 100, 107, 110,
Ecuador, 8, 76, 81, See also 111, 179, 187, 194
Amazonia, Huaorani ethnographer. See anthropologists
Christian missionaries in, 8, 76, ethnography. See anthropology
77, 78, 81, 82, 83, 89, See Eurocentrism, 127
also Huaorani Europe, 76, 82
colonialism, 83 Christianity in, 76
Curaray River, 81 Evans-Pritchard, E.E., 203
indigenous groups, 76, 78, 81, 83 Every Tribe Entertainment. See End
Huaorani. See Huaorani of the Spear
oil companies in, 81, 83, 84 everyday practice, 12, 99, 100, 101,
Organisation of Huaorani 103, 105, 106, 107, 108, 109,
Nationalities of Amazonian 110, 111
Ecuador (ONHAE), 85, 86 evidence, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10,
Palm Beach killings. See 11, 12, 13, 14, 15, 16, 17, 18,
Huaorani 20, 21, 22, 23, 25, 26, 27, 29,
plantations, 83 31, 32, 36, 38, 47, 53, 60, 62,
Quichua, 81 64, 69, 77, 78, 79, 80, 87, 89,
Eisenstein, Sergei, 63 90, 97, 98, 99, 101, 102, 104,
emotions, 2, 5, 6, 15, 16, 17, 134, 105, 106, 108, 110, 111, 118,
143, 144, 151, 152, 157, 170, 119, 121, 128, 130, 134, 136,
184, 188, 189, 190, 191, 192, 137, 145, 150, 151, 152, 158,
194, 195, 196, 197 172, 179, 181, 182, 183, 184,
ethnographer's, 15, 16, 133, 135, 185, 188, 192, 193, 194, 195,
142, 144, 145, 150, 157, 164, 196, 197, 201, 202, 203, 204,
165, 166, 167, 172, 173, 191, 206, 207, 208
See also anthropologists, evidentiary conventions, 6, 9, 10,
reflexivity, subjectivity 11, 12, 17, 88, 98, 100, 103,
in Papua New Guinea, 23, 24, 28 105, 206
Tibetan. See Tibetans politics of, 21, 201, 208
encompassment, 8, 10, 11, 12, 20, pragmatic, 99, 100
22, 26, 27, 29, 32 evidence-based medicine (EBM),
End of the Spear, 8, 76, 77, 80, 81, 11, 21, 101, 102, 108, See also
82, 85, 86, 87, 89, 90, See also clinical trials
evidence-based policy, 11
How Do We Know? Evidence, Ethnography, and the Making 217
of Anthropological Knowledge

evolutionism, 4 Gell, Alfred, 40

exchange, 129, 130, 180, 188 General Practitioner (GP). See
exemplary centre, 12 healthcare
Fabian, Johannes, 188 Ghana, 123, 125, 207
festival, 38 Accra, 208
fieldnotes. See writing rural-urban migration, 207
fieldwork. See anthropology Gibb, Camilla, 149
film, 2, 6, 7, 8, 9, 10, 17, 58, 59, 61, Giddens, Anthony, 171
62, 63, 64, 73, 77, 80, 90, 194 Gift, The, 12, 164
as knowledge, 58, 59, 61, 63, 64, Gilgit. See Pakistan
73 global health, 111
documentary, 63 globalisation, 2, 13, 118, 121, 127,
End of the Spear. See End of the 129
Spear anthropological conceptions of,
equipment, 59, 60, 61, 62, 63, 126, 130
66, 69, 73 Gluckman, Max, 205
ethnographic, 6, 7, 58, 59, 60, Granada Centre for Visual
64, 65, 66, 69, 70, 72, 73 Anthropology, University of
feature film, 8, 90 Manchester, 65
filmmaker, 7, 60, 63, 64, 69, 77, habitus. See Bourdieu, Pierre
See also anthropologists Haddon, Alfred Cort, 61
filmmaking, 6, 7, 8, 58, 59, 60, Hardman, Charlotte, 162
62, 63, 64, 65, 66, 69, 70, 71, Hastrup, Kirsten, 87
72, 73, 74, 85 healthcare, 98, 99, 102, 103, 108
footage, 6, 59, 60, 61, 62, 64, 66, in Gilgit. See Pakistan
69, 70, 72, 73 practitioners, 99, 101, 102, 103,
in Culebra. See Caribbean 104, 105, 107, 109, 111
observational, 65, 66, 73 Primary Care Trusts, UK, 103
of Margaret Mead and Gregory public benefit, 99, 102
Bateson. See Mead-Bateson public health policy, 98, 99, 108
project Heidegger, Martin, 187
popular, 8, 9, 77, 78, 88, 89 hermeneutics, 186
reflexivity, 63 hermeneutical circle, 185, 187,
theory, 63 188, 189, See Ricoeur, Paul
viewing, 59, 60, 63, 64, 73 history, 6, 77, 80, 90, 192, 207
film-truth. See Vertov, Dziga historical sources, 8, 76, 77, 201
Firth, Raymond, 203, 205 oral, 8, 90, 122, See also
Fjeld, Heidi, 161 Huaorani
Flaherty, Robert, 63 HIV/AIDS, 39, 48, 49, 51, 53
Flyvbjerg, Bent, 136 in Uganda, 47
Fortes, Meyer, 203, 207 in Venda, South Africa, 7, 36,
France, 123, 124 37, 38, 39, 48, 49, 50
Frechette, Ann, 159 associated sexual illnesses,
functionalism, 4 50, 52
Gabon, 122 biomedical discourse, 36, 39,
Geertz, Clifford, 136, 151 41, 47, 52
218 Index

deaths from, 37, 39 Palm Beach killings, 76, 81, 82,

folk models, 36, 49, 51, 52 85, 86, See also End of the
international donor agencies, Spear
36, 39, 40 Mincayani, 81, 82
music, 37, 39, 40, 41, 47, 52, relatives of victims, 82, 86,
53 87
peer education projects, 39, Steve Saint, 82
40, 41, 46, 47, 48, 53 report in Life magazine, 81
peer group educators, 36, 37, victims, 76, 81, 82, 86
38, 39, 40, 41, 42, 44, 46, Nathan Saint, 81, 82
47, 52, 53 predation, notions of, 77, 87
secrecy about, 37, 39, 48, 49, relations with Christian
52 missionaries, 8, 76, 77, 78,
Hollywood, 8, 77, 80, 82, 90 81, 82, 86, 88, 89, See also
homosexuality, 82 Ecuador
Hormone Replacement Therapy self-isolation, 81, 83, 89
(HRT), 104 victimhood, notions of, 8, 77, 78,
hounon (priest or priestess). See 83, 84, 85, 86, 87, 88, 89
vodou in Benin violence and spear-killing, 8, 76,
Howe, Leo, 163 77, 78, 80, 81, 83, 84, 85, 86,
Huaorani, 8, 76, 77, 78, 80, 81, 82, 88, 89
83, 84, 85, 86, 87, 88, 89, See Western characterisations of, 8,
also Amazonia, Ecuador 76, See also End of the Spear
auca, derogatory term for, 81 witchcraft accusations, 84
Christian mission, 81, 83, 85, 86, Huber, Toni, 161
87, 88, 89 human remains
conflicts with oil companies, 81, as scientific evidence, 26
83 repatriation debate, 25, 26, 27
conflicts with other groups, 81, human rights, 111
83, 85 Husserl, Edmund, 185, 187
conversion to Christianity, 81, ideology, 204, 205
85, 86 India
cosmology, 77 Dharamsala, 163
ethnography of, 77 Himachal Pradesh
household groups, 83 Bir, 159, 165
identity, 77, 84, 86, 88 Tibetan settlements in, 159,
missionary representations of, 165, See Tibetans
77, 82 New Delhi, 169
oral history, 8, 77, 78, 80, 82, 83, Tibetan diaspora in. See Tibetans
84, 85, 87, 88, 89 industrial revolution, 201
Christian imagery in, 78, 86, informants, 14, 15, 29, 64, 69, 73,
87 77, 78, 85, 86, 90, 119, 121,
Organisation of Huaorani 129, 130, 134, 143, 150, 152,
Nationalities of Amazonian 157, 164, 166, 167, 168, 172,
Ecuador (ONHAE), 85, 86 182, 187, 188, 193, 196
interdisciplinarity, 2, 10, 11
How Do We Know? Evidence, Ethnography, and the Making 219
of Anthropological Knowledge

interest groups, 32 Le Bureau National de Vodou,

internet, 206 Benin. See vodou in Benin
interpretation, 181, 182, 183, 186, Leach, James, 127
187, 188, 194, 196 Leavitt, John, 157
intersubjectivity, 60, 64 Lévi-Strauss, Claude, 78
intervention Lienhardt, Godfrey, 203
anthropological, 99 Life magazine, 81
clinical, 99, 102, 103, 105, 106, literary turn in anthropology, 205
107, 108, 110 literature, 90
surgical, 109 Long, Christopher, 151, 152
therapeutic, 47, 108, 110 Luthuli, Albert, 43
Isicathamiya. See South Africa Madagascar, 21, 46
Islam, 144, 145, 146, 147, 148, 149 Madzivhandila, Mmbangiseni, 49
in Pakistan. See Pakistan Makhene, Blondie, 43
Qu’ran, 143, 144 makwaya. See South Africa
Ivory Coast, 122, 129 Malawi, 51
Jesus, 192, 194 Malinowski, Bronislaw, 4, 203, 204,
Jojo Tshilangano. See South Africa 205, 207, 208
Josephides, Lisette, 157 Mami Sika. See vodou in Benin
Kalahari Desert, 63 Mami Wata. See vodou in Benin
Kant, Immanuel, 185 Mandela, Nelson, 43
Keats, John Mangrove Music. See Caribbean
magic casements, 185 Marcus, George, 99, 111, 145
Kewa. See Papua New Guinea marriage
Key Debates in Anthropology, 201 in Gilgit. See Pakistan
Khumbo, Nepal. See Tibetans, in Papua New Guinea, 23, 24
diaspora in Nepal Marshall, John, 63
Kilborne, Benjamin, 164, 168, 172, Marx, Karl, 181, 204
173, See also shame reactions on ideology, 204, 205
kinship, 182 materiality, 107, 109
knowledge, 3, 5, 6, 9, 13, 15, 17, 21, of disease, 107, See also The
22, 26, 30, 36, 37, 38, 45, 46, Body Multiple
52, 58, 59, 60, 61, 62, 63, 64, Mathase, Solomon, 36, 39, 46, 49,
65, 73, 78, 80, 87, 88, 90, 97, 50, 51
99, 107, 111, 127, 133, 145, Mauss, Marcel, 12, 164, 202
152, 188, 202 Mead, Margaret, 6, 61, 62, 63
Kondo, Dorinne, 145, 150 Mead-Bateson project, 61, 62
Lambek, Michael, 136, 152 medical ethnomusicology, 47, 48
language, 187, 207 medical knowledge. See
Balinese hierarchical language, biomedicine
163 medical practice, 98, 108, 111, See
Tibetan hierarchical language. also clinical practice
See Tibetans medical research, 99, 202, See also
Larry King Live Show, 82 clinical research
Latoe, Tibet. See Tibet medical technologies, 98
medicine, 111
220 Index

Melanesian Odysseys, 183, 188, ontology, 9, 80, 100, 107, 110, 187,
194, 195, 196, 197 See also Amazonia
memory, 5, 16, 179, 185, 189, 194, Organisation of Huaorani
196, See also recollections in Nationalities of Amazonian
tranquillity Ecuador (ONHAE), 85, 86, See
menstruation, 51 also Ecuador, Huaorani
mental illness, 98 osteoporosis, 11, 103, 104, 105, See
Merina, 21 also ARC SCOOP clinical trial
mimesis, 187 Pakistan
Mincayani. See Huaorani army, 138, 139
modernity, 13, 88, 118, 121, 127, Gilgit, 14, 133, 137, 138, 139,
128, 129 140, 142, 144, 146, 147, 148,
anthropological conceptions of, 152, 155
118, 121, 126, 127, 128, 129, Cheshmagula, 140, 141, 142,
130 143, 144, 145, 146, 148,
meta-narrative of, 127 149
Modise, Joe, 44 healthcare, 14, 138, 141, 146,
Mol, Annemarie, 100, 101, 106, 147, 148
107, 108, 109, 110, 112, 116 Ismaili, 137, 138
molecular biology, 26 magic, 143, 146, 147, 148,
money, 129 149
Mt Hagen. See Papua New Guinea marriage, 134, 140, 145, 146,
Muhammed, Prophet, 139 147
multimedia, 2, 6, 206 reproduction, 147, 148
multiple modernities, 127, See also Shia, 14, 137, 138, 139, 140,
Comaroff, John and Jean 141, 142, 147, 148, 149,
multi-sitedness, 5 152
music, 6, 7, 17, 27, 38, 39, 45, 47, Ashura ritual, 139
53, 194 Imam Syed Aga Zia-u’din
in Culebra. See Caribbean Rizvi, 138
in Venda. See South Africa Shia-Sunni hostilities, 14,
on film, 60, 70, 72 134, 138, 139, 140, 146,
performance, 6, 7, 37, 38, 39, 41, 147, 148, 149
60, 70, 71, 72 sociality, 14, 137, 143, 146,
Nadel, Siegfried, 203 147, 152
National Health Service (NHS), Sunni, 14, 137, 138, 139,
United Kingdom, 103 140, 141, 142, 146, 147,
Nepal 148, 149
Khumbo. See Tibetans tourism, 152
Tibetan diaspora in, 161 Wadood, 137, 140, 141, 142,
network, 5, 107 143, 144, 145, 147, 149
ngo tsha (shame). See Tibetans women, 14, 134, 137, 140,
non-governmental organisations, 40 142, 143, 144, 146, 147,
Nussbaum, Martha, 152 148, 149, 152, 153
objectivism, 183 Iranian political influences in,
objects, 194, 196 137
How Do We Know? Evidence, Ethnography, and the Making 221
of Anthropological Knowledge

Islam, 137, 146, 147, 148, 149 performance, 7, 27, 36, 37, 38, 47,
nationalist politics, 137 53, 61
Northern Areas, 137 performative ethnography, 38
police, 138, 139 pharmaceuticals, 98
Zia Ul-Haq, 137 phenomenology, 16, 184, 185, 187
Palm Beach killings. See Huaorani photography, 61, 62, 144
Panama, 82 placebo, 101
Papua, 203 poetry, 179, 184, 194
Papua New Guinea, 10, 23, 179 politics, 24, 37, 39, 42, 44, 48, 59,
administration, 23 60, 62, 66, 67, 85, 88, 89, 90,
ancestors, 24 110, 111, 123, 127, 133, 150,
courts of law, 10, 23, 24, 25 158, 159, 160, 182, 204, 208
emotions, 23, 24 academic, 201, 203
Kewa, 179, 180, 183, 188, 189, anti-Apartheid, 42, 43, 45, See
190, 191, 192, 194, 195, 196, also South Africa
197 contemporary debates, 202
emotions, 188 of evidence, 21, 201, 208
exchange, 180, 183 of race, 202
gender relations, 179, 180, political thought, 204
181 positivism, 97
genealogies, 182 postmodernism, 1, 4, 133, 135, 150
personhood, 191 practice, 107, 108, 110, 181, 182,
prestige, 180 183, See also Bourdieu, Pierre
production, 180 Production of Inequality, The, 180,
Rimbu, 182, 188, 190, 191, 181, 188, 190, 197
192, 193, 194 Puerto Rico. See Caribbean
wealth, 180 Questions of Evidence conference,
women, 181 January 2007, 201, 203
Yala clan, 182, 189, 192, 193 Rabinow, Paul, 151
male politics, 24 race
marriage, 23, 24 politics of, 202
Mt Hagen, 23, 24, 25, 27, 28, 29, reality
30, 32, 179, 202 concept of, 182, 184, 197
national airline, 179 recollections in tranquillity, 16, 182,
Port Moresby, 189, 192 184, 185, 188, 189, 196, See
warfare compensations, 24 also virtual returns,
women, 23, 24 Wordsworth, William
parenthood reflexivity, 4, 5, 6, 9, 10, 13, 14, 15,
in Britain, 27, 32 16, 59, 74, 77, 127, 135, 136,
in Papua New Guinea, 28 145, 150, 151, 152, 179, 189,
participant-observation. See 192, 206
anthropology ethnographic, 133, 134, 135,
patients, 98, 102, 103, 104, 105, 136, 137, 145, See also
107, 108, 109, 110, 111 anthropologists, emotions,
patient-activist groups, 98, 99 subjectivity
in film, 63
222 Index

tropes of, 133, 135, 136 anti-Apartheid politics, 42,

religion, 118, 202 43, 45
representation, 85, 97, 100, 107, Radio Freedom, 43
108, 110, 153, 184, 190 Ciskei, 43
Richards, Audrey, 203 Isicathamiya (musical genre), 42
Ricoeur, Paul, 185, 186, 187, 193 Johannesburg, 46
hermeneutical circle, 186 KwaZulu homeland, 43
theory of subjectivity, 185, 186, makwaya (musical genre), 42
187, 188 toyi toyi (dance), 43
Rimbu. See Papua New Guinea Venda, 7, 39, 40, 43, 45, 48, 49
ritual, 38, 182, 183 ancestors, 36, 39, 45
Rival, Laura, 84 cults of affliction (Ngoma),
Rivers, W.H.R., 4, 202 48
Rosaldo, Renato, 157 dances, 40
Rouch, Jean, 63, 69 death, 39
Sagant, Phillipe, 162 female initiation, 45, 51
Said, Edward, 208 Domba dance, 45, 46
Saint, Nathan. See Huaorani, Palm Vhusha, 45, 46
Beach killings HIV/AIDS in, 7, 36, 37, 38,
Saint, Steve. See Huaorani, Palm 39, 48, 50
Beach killings associated sexual
Sakpata spirits. See vodou in Benin illnesses, 50, 52
Scheper-Hughes, Nancy, 136 biomedical discourse, 36,
Schiltz, Marc, 192 39, 41, 47, 52
science, 97, 202, 208 deaths from, 37, 39
science studies, 107 folk models, 36, 49, 51,
senses, the, 59, 190 52
sexual illness, 50, 52 international donor
shame, 164, 166, 170, 173 agencies, 39, 40
in anthropology, 173 music on, 37, 39, 40, 41,
Tibetan notions of. See Tibetans 47, 52, 53
shame reactions, 164, See also peer education projects,
Kilborne, Benjamin, emotions 39, 40, 41, 47, 48, 53
Shia. See Pakistan peer group educators, 36,
Sisulu, Walter, 43 37, 38, 39, 40, 41, 42,
social structure, 181 44, 46, 47, 52, 53
social, the, 106, 109, 111 secrecy about, 37, 39, 48,
society, 22, 109, 110, 111, 208 49, 52
Sofo, 13, 121, 122, 123, 124, 125, music, 7, 36, 39, 40
126, 128, 129, 130, See vodou in anti-Apartheid, 41, 42, 43,
Benin 44, 45
South Africa, 7, 42 church, 41, 42
African National Congress, 43, communal songs, 39, 46,
44 47
Apartheid era, 43
How Do We Know? Evidence, Ethnography, and the Making 223
of Anthropological Knowledge

Tshilombe, Zwilombe, 36, pre-flight political organisation,

48, 49, 52, 53, See also 159, 160, 161
Mathase, Solomon regional identities, 159, 160
spirit possession cults Tibetans, 15, 157, 158, 159, 160,
(Malombo), 48 166, 167, 170, 172, 173
TshiVenda, 48 Aku Tonpa, folk-figure, 171
witchcraft accusations, 7, 37, diaspora in India, 15, 157, 158,
48 159, 161, 163, 164, 165
South America, 78, 90 education, 167, 168
status, 173 politics and leadership, 158,
Stoller, Paul, 150 159, 160
Strathern, Marilyn, 29 Central Tibetan
struggle songs. See South Africa Administration (CTA),
subjectivity, 13, 15, 16, 58, 61, 134, 158, 159, 163
137, 145, 157, 170, 182, 186, Dalai Lama, 158, 159
187, 188, 189, 196, See also democratisation, 158, 159
anthropologists, emotions, political parties, 159
reflexivity regionality, 160
Ricoeur's theory of. See Ricoeur, public gatherings, 168, 169,
Paul 172
Summer Institute of Linguistics romantic relationships, 169,
(SIL), 81 172
Sunni. See Pakistan diaspora in Nepal, 159, 161, 162
surplus value, theory of, 180 Lohorung Rai, 162
symbols, 38, 39 diaspora in North America, 162
Tallensi, 206 emotions, 157, 161, 162, 170,
Thambo, Oliver, 43 171, 172
theatre, 38 hierarchy, 15, 16, 157, 158, 159,
theory, 13, 17, 100, 133, 134, 145, 160, 161, 162, 163, 164, 167,
150, 151, 181, 184, 193, 194, 170, 171, 172
195 in language, 163
theoretical models, 6, 8, 9, 12, language, 158, 162, 163, 164
13, 17, 74, 78, 79, 87, 88, 97, honorific (zhe sa), 163, 164
98, 109, 119, 121, 127, 129, ma rabs, 163, See also ya rabs
130, 181, 183, 184 ngo tsha (shame), 15, 157, 158,
in Amazonian studies. 161, 162, 163, 164, 167, 168,
Tibet, 158, 159 169, 170, 171, 172, 173
bka' shag (cabinet), 158 respectful behaviour, 158, 162,
Chinese annexation of, 161 164
Chokyi Dronma, princess of saintly madman, 171
Gungthang (1422-1455), 161, selfhood, 157, 162, 170
162 socialisation in childhood, 163,
Kham, 159 167, 168, 170, 172
Latoe, 162 sociality, 15, 16, 157, 158, 161,
Lhasa, 161 162, 167, 170, 171, 172, 173
literature on, 160
224 Index

status, 158, 161, 164, 168, 169, hounon (priest or priestess), 120,
170, 171 121, 122, 123
ya rabs, 162, 163, 164, See also bo (spirit residence), 120,
ma rabs 121, 122, 123, 124, 125,
Togbe Yéhou (ancestral spirits). See 126, 128, 129
vodou in Benin bo pomési, 120, 124, 126,
Toren, Christina, 130 128, 129, 130
Torres Strait. See Cambridge bo yovoni, 120, 121, 126,
University 129
tourism, 152 human-spirit relations, 118, 120,
tradition, 13, 45, 47, 118, 127, 129 129, 130
anthropological conceptions of, le Bureau National de Vodou,
118, 121, 126, 127, 128, 129, 119, 123, 124, 125, 126
130 National Vodou Day, 123, 124
Trobriand Islanders, 203 reme (spirit companion), 126
Tshilombe. See South Africa shrines, 119, 122, 123
Turner, Victor, 38, 203 smoto (temple servant), 123
Uganda Sofo, 121, 122, 123, 124, 125,
HIV/AIDS in, 47 126, 128, 129, 130
UNESCO, 202 pomé (ancestral family) of,
United Kingdom, 11, 25, 27, 29, 32, 122
99, 102, 103 spirits, 119, 120, 121, 123, 125,
National Health Service (NHS), 126, 128, 129
103 Edan, 119
United States, 67, 76, 81, 82, 192 Heviosso, 119
Christianity. See Christianity kuku (dead ancestors), 119,
Detroit, 201, 202 120
Van Maanen, John, 135 Mami Wata, 118, 119, 120,
Venda. See South Africa 121, 123, 124, 126, 127,
Vertov, Dziga, 63, 69 128, 129
video. See film literature on, 121, 126,
virtual returns, 16, 179, 185, 188, 127, 128, 129
189, 191, 193, 194, 196, 197, Mami Sika, 120, 123, 124,
See also recollections in 125, 126, 128
tranquillity Mamino (priest of), 124,
Viveiros de Castro, Eduardo, 83, 84, 125, 126
89 worship, 118, 121, 126,
vodou, 12, 13 127, 128, 129
vodou in Benin, 2, 13, 118, 119, regional trade in, 128
121, 123, 126, 128, 129, 130 Sakpata, 119, 120, 121, 122,
ahwan (troop of spirits), 123, 123, 125, 126, 129
124, 125, 128 Togbe Yéhou (ancestral
as national heritage, 119 spirits), 119
botono (maker of the bos), 122, Vodou/Yéhoue, 119
See bo, See bo temples, 119, 120, 121, 122
categories, 118, 121, 129, 130 vossa (sacrificial gift), 120
How Do We Know? Evidence, Ethnography, and the Making 225
of Anthropological Knowledge

vodou in West Africa, 119, 126 172, 179, 181, 182, 183, 186,
Wagner, Roy, 195 188, 190, 193, 194, 197, 203,
Wardlow, Holly, 135 204, 205
West Indies. See Caribbean personal disclosure, 133, 134,
Wikan, Unni, 157 135, 136, 145, 152, See
Wiki Sound, The. See Caribbean also reflexivity
witchcraft accusations, 7, 37, 48, 84 fiction, 90
Wittgenstein, Ludwig, 135 fieldnotes, 7, 13, 16, 106, 152,
women 164, 182, 189, 190, 191, 193,
in Gilgit. See Pakistan 194, 207
in Papua New Guinea. See Papua writing culture debate, 1, 205
New Guinea Wycliffe Bible Translators, 81
Wordsworth, William, 16, 184, 185 Yanomami, 83
Lyrical Ballads, 179, 184, See Yeh, Emily, 162
also recollections in Zafimaniry, 21
tranquillity Zia Ul-Haq. See Pakistan
The Prelude, 184 Zimbabwe
writing, 6, 10, 194, 205 Karanga, 51
ethnographic, 10, 13, 17, 58, 60, Zulaika, Joseba, 135, 150
64, 66, 74, 90, 97, 133, 134, Zwilombe. See South Africa
135, 136, 137, 149, 150, 164,