Mosse 2006 AntiSocial

Anti-social anthropology?
Objectivity, objection, and the

ethnography of public policy
and professional communities*
D.vi u Mossv School of Oriental and African Studies, University of London
One legacy of Malinowskis ethnographic method is the separation of eld and desk. What
anthropologists know is inseparable from their relationship with those they study the epistemology
is relational but ethnographic writing breaks eldwork relations, cuts the network, and erects
boundaries: it is necessarily anti-social. As anthropologists turn their interest in what people believe,
say, and do (and the inconsistencies between these) to the inter-connected institutions that comprise
the modern world, to policy and professional communities of which they may also be members,
their method of entering and exiting social worlds becomes more difcult. Arguing for the particular
importance of an ethnographic perspective on the practices of powerful institutions, this article uses
recent research on international aid and development to show how inuential informants object to
ethnographic accounts, resist anthropological boundary-making, and attempt to unpack academic
knowledge back into relationships.
At the end of September :oo:, I completed the rst draft of a book on international
development policy and practice, taking as its focus a project in tribal western India
funded by Britains Department for International Development (DFID) (Mosse
:oo,a). This was an unusual piece of research, complex, long-term, multi-sited, and
initially unintentional, drawing as it did on insights as a participant-insider within
international aid; and its conclusions questioned prevailing assumptions about devel-
opment policy-making and project practice. The book manuscript provoked unusual
controversy. Objections were made by my co-workers and informants to the publisher,
to my university research ethics committee, my Department convenors, the Dean and
the academic head of my university, as well as to my professional association the ASA
(Association of Social Anthropologists of the UK and the Commonwealth) on the
grounds that the book was unfair, biased, contained statements that were defamatory
and would seriously damage the professional reputation of individuals and institu-
tions, and would harm work among poor tribals in India. Those of my project col-
leagues who raised these objections sought to interrupt the publication process and to
ensure that many parts of the book were rewritten. In April :oo I was called to defend
* Malinowski Memorial Lecture, :oo,.
Journal of the Royal Anthropological Institute (N.S.) :i, ,,,-,,o
Royal Anthropological Institute :ooo
my ethnography in front of angry informants international experts and Indian
project managers in the presence of professional colleagues. The move to publication
had strained and broken valued relationships of eldwork.
Malinowski might have been puzzled by such a scenario. The young Malinowski
might have wondered how anthropologists could turn ethnographic attention to the
schemes and policies of those authorities upon whose good ofce they and especially
he relied for eldwork, rather than conning comment to letters and diaries (Young
:oo); and the older Malinowksi might have been surprised that the practical anthro-
pology he so strongly promoted in the I,,os as the basis for a science of colonial social
policy (Kuper :oo,) could turn so controversial. However, what I want to suggest in this
article is that the scenario in fact arises from the fundamental structure of the ethno-
graphic method that Malinowski innovated nearly a century ago. That structure is the
relationship between eldwork and writing between, for Malinowski, the empirical
work of observation (of the actualities of Trobriand life) and the constructive work of
tabulation, inference, and theory (Malinowski I,::); a relationship now stretched on
Leachs (I,oI) famous rubber sheet to the point where entirely new problems in the
analysis of events are emerging. The challenge for anthropology today is not how to
rearrange eldwork that dubious category that has come to signify any shift in
location that is the ethnographic pretext (Gupta & Ferguson I,,,) or how to re-frame
writing, but how to get to grips with the changing relationship between the two:
change, rst, in how eldwork relations shape writing, and, second, in how writing now
alters relationships of the eld.
In many ways anthropologists are a lot closer to their other than they used to be.
Arguably Malinowskis eldwork was a method of dislocation rather than of immer-
sion. His now legendary social distance from his Trobriand subjects, his loneliness, and
his scientic isolation from the ow of social relations formed, Michael Young suggests,
the bedrock of his synchronic functionalism (:oo: ,:,). What Malinowksis succes-
sors lacked of his brilliant powers of observation and exhaustive description, they made
up for in forging closer relations with their subjects, greater identication, equity, and
dialogue, often through long-term and repeated eldwork. But with this, ethnogra-
phers became socially bound into their eld sites in a new way, or, as Parkin puts it,
became increasinglytemplated by the eld (:oooa: IoI). In parallel, anunbounding of
the eld so as to include webs of regional and transnational connections and commu-
nities means that all anthropologists now research to some degree as insiders or at
home. Furthermore, with the higher education funding squeeze of the I,8os (at least in
Britain), trained anthropologists (myself included) joined non-academic institutions
for example, in international development (Panayiotopoulis :oo:; Spencer :ooo) and
while meeting new professional obligations also began contributing to a growing body
of insider ethnography of organizations and public policy. As researchers, we resolved
the intractable problems of access to closed organizational worlds through member-
ship of the communities we ended up studying. But in doing we so substituted a set of
boundaries that kept us out (the problem of access) with another set that kept us in.
Those who made themselves professional insiders in this way faced the problem not of
entering a different world so as to be able to imagine or infer the taken-for-granted
(and therefore hidden) way in which individual action and collective illusions are
interlinked (Hastrup :oo: o,), but of exiting a known world for the same purpose.
In fact, closer relations in the eld, long-term and insider research have all made
exit rather than entry the signicant shift in location that is ethnographys pretext
David Mosse 936
including exit from the templates of our younger ethnographic or professional selves
(cf. Parkin :oooa), whenthe eld is a dislocation in time. And as other boundaries fade,
it is often the detachment of writing itself that has become the primary mode of exit.
While eldwork has changed beyond recognition becoming ever more intensely
social ethnographic writing (interpreting, objectifying, and textualizing) remains a
solitary process that disembeds knowing from its relationships, denying (to varying
degrees) the social its claim to power, to ownership, to negotiation.
1
For Malinowksi
himself, writing was a necessarily anti-social process incompatible with intense social
intercourse, frustrated by the conversational (Young :oo: ,) and [t]he unnecessary
communion of souls (Young :oo: ,,:, citing Malinowski I,o,). What we have inher-
ited is not so much a particular practice of eldwork, as an ethnographic method
premised on the division of eld and desk the social and the anti-social experienced
by every returning researcher. The changing nature of eldwork its closeness both
intensies this division and surrounds it with tension.
Closeness makes writing more difcult, not just because of an exponential sense of
incompleteness that David Parkin (:oooa: Io,) notes of accounts which are always
partial and provisional, but because ethnographic writing begins to have signicant
social effects of its own. The detachment of writing is now socially experienced by
others. Of course, those reading about themselves may be intrigued, amused, or
pleased;
2
but turning relationships into data, and placing interpretations in public, can
also disturb and break relationships of eldwork. It may be anti-social.
3
Those inter-
locutors neighbours, friends, colleagues, or co-professionals who directly experi-
ence ethnographic objectications now surround the anthropologist at her or his desk;
they raise objections, make new demands to negotiate public and published interpre-
tations. The relationships of the eld persist, the capacity to exit through writing is in
question, and ethnographic representations have become unavoidably part of the
world that is studied. When desk collapses into eld, something important has changed
in the structure of ethnographic practice. We are starkly confronted with the essentially
relational nature of anthropological knowledge,
4
no longer an object in our possession.
That is to say, what anthropologists know is inseparable from their relationship with
those they study. Consequently, the issue at hand is not just ethical but epistemological.
The narrative ethics of the I,8os that followed Edward Saids critique of Oriental-
ism and the writing culture debate left some anthropologists imagining the problem
could be solved by a retreat from representation altogether, allowing subjects to speak
in their own words through personal narratives. Such replacement of description with
evocation (Parkin :oooa) and the honouring of individual agency implies, Kapferer
suggests, a broad shift away from concerns with social relations and interactive struc-
tures (:oo: I,:) that has made the discipline anti-social in another sense. Recently the
metaphor has shifted again from dialogue to collaboration in ethnography (Lassiter
:oo,; Marcus I,,8). Of course, the notion that power inequalities between the inter-
preter and the interpreted can be dialogued away, or written out, is too obviously false;
and, as Spencer (I,8,) points out, the analytical cost of this pretence is considerable.
But equally problematic, in the politics of representation debates, and the professional
ethical guidelines that they inuenced, were the assumptions made about the gradients
of power across which research takes place. Relying on informant self-representation
and allowing subjects to speak in their own words are not self-evident solutions for
anthropologists of public policy whose informants are ofcials at the World Bank, the
International Monetary Fund, DFID, or any group with a strong organizational need
David Mosse 937
to produce and protect authorized views. Moreover, if they do critically investigate
organizations or public policy of one kind or another, anthropologists begin to face not
just personal unhappiness, but also public and formal reprimands, or even the threat
of defamation proceedings, for their ethnographic accounts;
5
threats which may none
the less still be framed in terms of research ethics set out in guidelines drawn up from
an earlier generation of ethnographic practice in order to protect those lacking
other means of redress.
Here I am concerned with this kind of ethnography, where eld/desk, self/other,
subject/object, here/there distinctions do not apply in the same way; where reexive
concern about epistemological privilege gives way to worry about epistemic capture or
co-option; but which is no less based on extended eldwork and participant observa-
tion of the social and symbolic transactions of communities, even if these are epistemic
communities (e.g. of policy experts) interacting electronically, organized transnation-
ally with their own public spheres, and with contractual or other exchange relations
with similar groups (Friedman :oo: Io). Why undertake such study? Well, ethnog-
raphy offers particular insights into relationships mediated by policy ideas within
contemporary network society (Castells I,,o). It also offers another means of public
engagement with powerful institutions whose knowledge systems constantly organize
attention away from the contradictions and contingencies of practice and the plurality
of perspectives. Or as Burawoy puts it, by highlighting the ethnographic worlds of the
local, [anthropology] challenges the postulated omnipotence of the global whether it
be international capital, neoliberal politics, space ows, or mass culture (I,,8: ,o). But
my concern here is not to explicate or justify the anthropology of policy, global gov-
ernance, or international development (see Mosse :oo,b) but to examine the dynamics
of such research in the light of the ethnographic practice of exit and objection.
In what follows I will rst explain my approach to a piece of insider research and
outline the main argument that resulted. Second, I will examine the nature of the
objections made to this ethnography, and the epistemological divide between an
anthropologist and his informant-readers that they reveal. Third, I will track the
unfolding controversy to see how the boundary between eld and desk is contested,
and how rightness is both countered and claimed in an ethnographic encounter. My
concern with representations in anthropology here points not to reexive poetics but to
the politics (and ethnography) of objection.
I worked as an anthropologist-consultant on a development project from its initial
design in I,,o until :ooI. Because of this rare continuity, and the particular importance
of this project as a agship within the I,,os British aid programme demonstrating
a new commitment to participatory and poverty-focused interventions the DFID
agreed to support a study of the project experience from my particular anthropological
perspective. This would be a critical analysis of policy and administrative rationality
and modes of expertise in aid and development including those of social anthropol-
ogy itself. It would be based on the best available evidence, but would not cease to be an
interested interpretation, a personal analytical account; an ethnography in which I was
myself a key informant.
Over ten years I was part of an extended project team including consultant col-
leagues (experts in forestry, crops, irrigation, soil and water conservation, or gender),
project managers, and eld staff, all with whom I spent time in meetings, on long
journeys, at their homes, with farmers in the scattered villages of this Bhil adivasi
David Mosse 938
(tribal) region of western India, and in drafting and redrafting collaborative reports.
These intense periods of interaction from which grew personal relationships of
understanding, trust, and respect were the rst source for my research. The second was
the series of studies and reports produced from our various engagements, and a large
body of contemporary project documentation (for I,,o-:ooI). As a third layer of
research, in :ooI I returned to India to carry out interviews with project workers and
ex-staff, as well as with India- and UK-based DFID ofcials and consultants. The aim
was to test and verify my understanding of project processes, to decentre my own view,
and to extend the analysis to the wider context of British aid in India (Mosse :oo,a: ix).
Now, there was another, fourth, methodological level, and this concerned the
response to my analysis from those who shared the experience and about whom I write.
Such ethnography courts controversy and is likely to produce objections. I would like
to suggest that these objections are themselves part of research which emerges from,
and reects on, relationships in development. And here I concur with Bruno Latours
view of objectivity, which derives not from standing above the fray or suppressing
subjectivity, but from maximizing the capacity of actors to object to what is said about
them (:ooo). So, I shared my writing with informants, collaborators, colleagues, and
friends, who possessed a capacity to object.
Summary of the argument
Before outlining these objections let me explain, at the risk of oversimplifying, the
nature of the ethnography at the centre of the controversy (see Mosse :oo,a). It is an
exploration of the relationship between international development policy and project
practice, and focuses on participatory approaches prominent in the I,,os. The argu-
ment unfolds around ve general propositions.
The rst is that policy in development functions to mobilize and maintain political
support as much as to orientate practice. I show how, in this case, the work of project
design served to negotiate relationships and bring together diverse and quite incom-
patible interests around a causal model that justied the allocation of resources. I
explain how the conceptual and linguistic devices that enrolled support also built
contradictions into the design that made its straightforward execution in practice
impossible.
The second proposition is that development interventions themselves are driven by
the exigencies of organizations and the need to maintain relationships rather than by
policy. My book (Mosse :oo,a) describes in some detail how the informal everyday
practices of project workers constituted a system of relationships shaped by the politi-
cal logic and culture of the project agency, and by its demand for administrative order
(i.e. by what may be called system goals; Quarles van Ufford I,88), and routinely
contradicted the prescriptions of ofcial policy, substituting bureaucratic rules, targets,
and controls, or relations of patronage, for anticipated community self-reliance (see
Mosse :oo,a: Io,-,I). In fact, the project was a world comprised of different autono-
mous spheres (of the village, eldworkers, ofce, managers, consultants, donor advis-
ers), mediated by institutional brokers, in which policy models could not organize
practice. Which is not to say that policy was irrelevant, but that as a kind of mythology
it was only partly a charter for action, since it had symbolic functions accounting
upwards, legitimizing expertise, signifying alliances, or concealing differences that
were at least as relevant as pragmatic ones (Leach :oooa [I,,,]: ,,).
David Mosse 939
Although their practice often contradicted the prescriptions of participatory project
design, staff (including consultants, and Bhil villagers too) none the less worked hardest
of all to sustain and protect ofcial interpretations of actions, so that they articulated
with higher DFID policy, matching events to theory in many and sophisticated (though
not always conscious) ways because, thereby, success and their (our) own interests
were secured. Here were several project rituals in Leachs sense of procedures to
overcome the anxieties which are generated by [the] lack of t between how things are
and how we would like to think about them (:ooob [I,,o]: 8,).
6
My third proposition,
then, is that development projects work to maintain themselves as systems of repre-
sentations as much as operational systems.
7
The work of international development
consultants such as myself, while appearing to assist development operations, in fact
mostly serves to produce/sustain policy models offering an authoritative interpretation
of events that result from quite different logics; not (as we imagined) preceding or
directing action but following it. Through such expert discourse unruly practice is
stabilized, and the gap between policy and practice constantly negotiated away. Again,
I have to refer the reader to the ethnography which shows how such interpretations
have, further, to be sustained socially by enrolling supporters and building an inter-
pretative community that made the project successful; and how policy designs also
provide the framework of self-objectication for project actors accounting for them-
selves to each other and to outsiders.
The fourth proposition is that, correspondingly, project failure is not the failure to
turn designs into reality, but a certain disarticulation between practices and their
rationalizing models. Failure, as I discovered, is a failure of interpretation. And this I
demonstrate ethnographically through the exploration of a project crisis in the context
of evaluation and the rapid shift in DFIDaid policy after I,,,. Such policy changes have
the effect of making the chains of translation in development more complex and
harder to negotiate (Mosse :oo,a: :Io). The nal proposition concerns the way in
which policy discourses of success and failure conceal the local social effects of
development interventions, not only perpetuating misleading explanations, but also
concealing valued outcomes, which, in the case of Bhil villages, include new forms of
patronage, access to resources (subsidies, agro-inputs), and the means to articulate new
aspirations of progress and cultural re-valuation (Mosse :oo,a: :o,-:,).
The ethnography explains all these as general and inherent features of the system of
international aid not as the failings of one particular project. It is not an evaluation.
Indeed its central concern is not whether development projects are successful, but how
success is socially produced or constructed. Its supposition is that in the hugely
complex cross-cultural world of development, most actors (including apparently pow-
erful ones) have very little control over events.
8
What is usually more urgent and more
practical is control over the interpretation of events; and as Bruno Latour reminds us,
the success of policy ideas or project designs is not inherent (not given at the outset) but
arises from their ability to continue recruiting support and so impose their growing
coherence on those who argue about them or oppose them (I,,o: ,8; Mosse :oo,a: 8).
Since it is prone to being misread, let me stress that my argument is not intended as a
criticism of aid and development. Rather, my analysis of the practices of a participatory
development project, including my own role within it, aims to understand the micro-
social processes of policy. So, rst, it does not imply a rejection of this enterprise (or of
aid projects more generally). My book makes clear that the project did have a signicant
positive effect on the lives of many thousands of disadvantaged Bhil adivasi people; but
David Mosse 940
often in unscripted and unintended ways (Mosse :oo,a: ::,). However, the point is that
the project did not work because it was well designed. Stability in the world of action
did not come from coherent policy; good policy was not implementable. None the less,
policy is absolutely central to what happens in development arenas (Mosse :oo,a: :o).
Second, this is not a case of cynical disengagement. My argument does not express
regret at my own involvement in this aid project, nor warn other anthropologists
against working in development. On the contrary, the implication is that that there is
greater need than ever for anthropologist involvement in the complex and contradic-
tory arenas of international aid. But in order to grasp the social and political processes
through which aid policy is made and transformed in practice, and which have a major
bearing on outcomes, anthropologists have to negotiate space for their involvement to
be more ethnographic and resist institutional pressure to conform to dominant policy-
driven or economics-based knowledge systems (see Mosse :oo). Anthropologists do
not have to choose between optimism and scepticism. In relation to development, as
Quarles van Ufford puts it, they participate in three separate domains the domain of
hope, of politics/administration, and of critical understanding (Quarles van Ufford,
Kumar & Mosse :oo,). That is to say, our engagement is with the discourses of moral
responsibility and policy vision (about what is to be done), with the strategic politics of
programme action, and with critical reection. These are not exclusive modes of
thought and action, but neither are they entirely reconcilable. The danger is in asserting
any one over the others (Mosse :oo,a: :o-,; Quarles van Ufford et al. :oo,).
Of course my ethnographic analysis is a positioned interpretation, which does not
preclude other accounts. It is also one in which I place myself alongside others who
worked on the project, taking my responsibility for shaping the projects design, for its
construction of success, for the navety, over-ambition, and wrong-headedness of my
own contributions. I can admit these as personal failings, but also see them as preg-
ured by the structural and discursive conditions of a development project without
doubting my own or others sincerity or commitment.
The nature of the objections
As I said, I shared my writing with my colleagues, collaborators, andinformants. Now,
most who responded to the drafts over eighteen months especially my social devel-
opment and eld staff co-workers in fact gave strong endorsement to my analysis,
describing it as balanced, truthful, insightful. However, my attention became preoc-
cupied with those key actors (including UK technical consultants and those in mana-
gerial positions) who took strong exception to my too negative and unbalanced
account, which was unfair and disrespectful, out of date, and evendamning of all our
work. This group, represented by a UK consultant and DFID Project Adviser,
9
dis-
agreed fundamentally with my conclusions and wanted the book re-written. Such a
reaction should disturb any ethnographer; the more so for me because these were my
close colleagues, co-workers over thirteen years.
The objection unfolded in stages: personal disagreement and friendly editorial
advice gave way to questions of abuse of contract (was I entitled to use information
acquired as a consultant) and appeal to the principle of participation in research the
notion that collective experience has to be collectively analysed. Here there was an issue
both of method by failing to make the research fully participatory I reduced its
validity and laid myself open to criticism for being unbalanced; and of morality I was
at fault for the individual appropriation of shared experience, substituting stand-alone
David Mosse 941
arguments for collective experience. How could I, one exasperated colleague asked,
the participation consultant, not want to review the project experience in a fully
participatory manner? But worse, by making myself an outsider, I was disinvesting
from categories of explanation that I had myself promoted, fed back, and routinely
used. Finally, in graver and angrier tone, the ethnography was challenged as potentially
seriously damaging to the professional reputations of individuals and institutions (the
Indian agency, a UK agricultural research institute, the DFID); it would adversely affect
the ability to attract nance and undo the work of those who had given their best years.
I was astonished by the strength of emotion conveyed through e-mail, telephone
conversations, and eventually face to face about a book which to all its independent
readers did not appear to defame or malign the reputation of anyone. Nor was there any
plausible explanation of how the book would damage organizations (DFID or the
project agency), destroy the programme, or its routes to funding.
10
Of course, I rose to defend the truthfulness of my account and the soundness of my
research methodology; but it also occurred to me that my critics were themselves
enacting the very argument they objected to, offering extraordinary conrmation of
the key point that authoritative actors work hardest to defend projects as systems of
representations, not only against the destabilizing contingencies of practice, but also
now against competing (ethnographic) representations existing potentially within the
same public space. Was there in the angry accusation that the book questioned our
professionalism an implicit recognition of the truth that managers and experts are
involved in the organization of interpretations as much as planned outcomes? More
generally, what reading of the ethnography did these emotive encounters reveal? And
what could they add to my understanding of the culture of aid projects? Let me take up
some key points.
First, my colleagues did not read my ethnography as an exploration of a general
theme (perhaps a theory of policy) through the particular. Of course, this was a story
about them. Theory and citation became a duplicitous hiding behind others. Second,
my colleagues did not share the ethnographys interpretist view of project reality as a
multiplicity of truth composed from different points of view. It would be read as a
singular statement about the project as it is, taken as objective (the team leader added)
because written by someone from a world-renowned institution. From their positivist
perspective, talk of alternative points of view simply dealt in the currency of spurious
facts or biased interpretations which, as I was told, fail to meet the normal standards
of social science research.
Third, the ethnography was read as an evaluation. My colleagues felt judged. They
did not, as others had, read a description that exonerated their struggle with the real
contradictions of development, but a commentary on the gap between the actual and
the ideal; a judgement against norms or best practice that critically assessed their
professional competence, and dismissed their effort and enthusiasm.
But the book was unfair not just because it was evaluation, but because it was bad
evaluation. For one thing, the ethnographic genre, unclothed in the ofcial etiquette of
praise and indirect comment, appeared unacceptably critical. For another, it did not
judge the project in its own terms, but added complexity, clouded issues, and intro-
duced diversions and irrelevant details. At a meeting, one of the objectors referred to
the book as a eld of mixed crops when everything is sown higgledy-piggledy its all
a confusing muddle .
11
Moreover, while judging performance as falling short, I offered
no scale. Show us the model of true participation; what did we do wrong? But most
David Mosse 942
important, the book was bad evaluation because it had not involved the usual nego-
tiation between evaluators and project actors over who is qualied to construct
knowledge about a project, how it to be done (methodology), and what is to be said.
The ethnography failed to be what all good evaluations are, namely an acceptable story
that mediates interpretative differences in order to sustain relationships and the ow of
resources (Phillips & Edwards :ooo).
12
Ultimately it was from this expectation of a
shared denition of the truth about the project that the objectors took their right to
propose changes to the text.
13
The next point is that the ethnography was read from a managerial perspective. It
was unnecessary and embarrassing because it refused to explain outcomes in terms of
design, and evaded the expectation that problems should really be analysed only in
relation to solutions. It did not provide a proper project history of implementation,
learning, and improvement, which should reveal a progressive narrowing of the gap
between intention, action, and outcomes.
14
I had not explained the steps taken, how
the project [had] responded to particular events and problems. In my book, difculties
and contradictions were not, as they should be, dissolved by the unfolding project,
which makes history a dustbin of irrelevant errors and solved problems. It interrupted
a managerial view that accelerates history so that the aspirations of the present con-
stantly erase the experiences of the past, where it is always too early to judge the success
of new technology, whose disappointments are contingent (drought or monsoon
excess), whose latest results are always the most promising, and whose full advantage
(upon which justifying economic analysis is based) lies in the future. The ethnography
was not just out of date, dealing in moribund problems and ignoring the unfolding
success, it was also out of time, in Hastrups sense of giving attention to the routine and
ordinary, the out of sight, so avoiding the normal historical narrative of temporal
causation which accounts for events in terms of the most recent and most extraordi-
nary precedent that is, the project and its technical innovations (:oo: o:).
Next, there was a problem with the ethnographic treatment of my colleagues data.
On the one hand, the ethnographic account denied interpretative power to scientic
data: for example, to data on the genetic and generalized advantages of improved seed
technology, derived from context-free models, which were unravelled in the relational
world of debt-bound tribal livelihoods. On the other hand and this is something
which recalls a division in the earlier years of our own discipline recorded in Edwin
Ardeners I,,I lecture in this series in my analysis, project success was anyway not just
a matter of the measurement of achievement and empirical or statistical evidence (of
yield increases, trees planted, functioning groups). Success was a matter of denition, a
question of meaning, of sustaining a particular interpretation of events through the
categorizations and causal connections established by the policy model. And this model
was not itself empirically falsiable. But it was replaceable. So when DFID policy
changed in I,,,, the project became, by denition, a failure (Mosse :oo,a: I8-:o). The
socio-economic studies we undertook then to demonstrate impact were in fact ori-
entated to re-model tribal livelihoods so as to show how the project, re-aligned to a new
policy framework, would improve them. The research served more to clarify and justify
a new development model than to demonstrate its effects. After all it was on this that
project survival urgently depended. Showing the need for further action is always
politically more important than demonstrating results (Quarles van Ufford I,88: :,).
So, the ethnography dismissed empirical evidence and implied that the project was a
self-verifying system in a way that was considered damaging to scientic reputations.
David Mosse 943
A further point is that the objections to the book reveal a particular representation
of agency in the project world. The multiplicity of actors and intentions is concealed as
agency becomes deected as the property of ideas or expertise, design, technique, good
policy, or the problem-solving project. Project workers (including consultants) hide
their own contingent actions and the wider politics of aid so as to, as Timothy Mitchell
puts it, allow reason to rule, and allow history to be arranged as the unfolding of a
locationless [policy] logic to which expertise is attached (:oo:: I,). Managers insisted
upon precisely the kind of essentialized abstractions and transcendent agents that
anthropologists have come to berate themselves for constructing (Hobart I,,o: ,). But
while success sustains the project as a unied source of intention and power (successful
projects are well designed), failure fragments into the dynamics of blame (Latour I,,o:
,o). As a project worker put it to me, We always appreciate our successes, but failure is
always seen as the failure of an individual.
An ethnography which draws attention to diverse intentions and motivations by this
detail involves unethical disclosure; and even where (as with mine) it eschews
the individual personality, action, or event in order to reect on systematic effects or
outcomes beyond intention even though it distributes agency, it becomes a source of
personal offence. My colleagues unpacked structure into their agency, claiming per-
sonal damage to professional reputations. Now, the tactical need to discredit the
account in terms that would register as defamation or a contravention of research
ethics (harm to informants) had some part in this, but it seems to me also to derive
from this paradoxical way in which agency is framed in project arenas the collectiv-
izing of success and the individualizing of failure
15
when confronting an ethnography
that by contrast aims for a symmetrical treatment of success and failure.
As I mentioned, my colleagues insisted that objective truth has to be collectively
dened; to become facts, interpretations had to be subject to group appraisal and
agreement. An ethnographic approach which interviewed people individually or in
groups and then collated and compared diverse opinions, events, and experience in an
independent interpretative analysis simply did not qualify as proper social science. I
was even reprimanded (in a meeting) by one junior manager for using unreliable
private conversations instead of statements made in public about events, on the
grounds that informally people will invent stories, confuse, and conceal, but publicly
they will speak the truth. Team discussion would compensate for the failures of indi-
vidual self-censoring, and this, of course, is why it was not a good means to research
such complex and contested social processes.
But there is a broader point here, namely that for these project actors social research
has to preserve and honour its social context. At one level this means simply that there
is an ethical obligation to those who helped you and gave their time and materials. At
another, it is an epistemological position that implies that the limits of what can be
known, revealed, or written about are determined by social relationships. Perceived
harm, risk of damage to reputations, or embarrassment to institutions
16
invalidates the
analysis. Fairness in research is a question of respect (and unfairness, disrespect) rather
than veriability. Research data and analysis are correct (and mine was incorrect) in
the normative sense of socially appropriate (as incorrect behaviour) as well as factual:
I am sorry, but we are not talking about simple factual errors ... what we are talking
about is incorrect statements about events or decisions made on the project ... we are
very disturbed by your draft. Concerned with correctness rather than fact, the
objections were epistemological not ontological. Indeed, my colleagues positivism
David Mosse 944
concealed an essentially relational epistemology which rejected the notion of evidence
as external to the situation.
As a nal point, they felt that the moral nature of their actions should shape the way
they were described. It was wrong to appear to criticize those who work selessly,
enduring hardship for the poor. Moreover, it was construed as irresponsible to question
participatory methodologies (e.g. in crop research) lest the state take against them. The
projects strategic truth had to be preserved.
Now I see a triple signicance to these objections. First, the reaction to my ethnog-
raphy reveals, indeed elaborates, the same framework of expectations that organized
the project system, further clarifying its conceptions of time, agency, or evidence.
Second, the objections were not about what was to be known but about how knowledge
was to be arrived at. They revealed an apparent divergence of epistemology between an
ethnographer and his interlocutors or rather one dominant section of them that
suggests limits to any collaborative ethnography.
17
Third, and more importantly, the
process of objection showed how ethnographic writing threatened the project as an
epistemic community a set of relationships around shared meanings drawing
anger from those dominant gures whose prestige was most closely tied to authorized
representations.
Let me consider the latter point further. The ofcial-textual view of the aid project
as an explicit system of rules and procedures (regularized in brochures and training
manuals), a scientic order, a replicable model, and a history of signicant actions/
events had a certain necessity. It was necessary (a) given the high degree of uncertainty
in international development involving people who do not know or understand each
other; (b) in order to put back together the worldview of project staff that was con-
stantly fragmented by the everyday contradictions of practice; and (c) because actors
invested in these objectications, through which systems of expertise, status, esteem,
and reward operated. Ofcial views and habitual objectications were (to different
degrees) part of peoples assertions of power, self-denition, and representation to each
other and to outsiders government, donor, other experts and researchers who would
read this version of project reality (cf. Geertz I,,,: ,,).
Such self-objectication as structure, rule, or replicable model is not unlike the
native offerings re-inscribed by anthropologists and so keenly criticized by Bourdieu
(among others) in his call to penetrate the strategies of practice, the temporality and
indeterminacy of social life (I,,,; Jenkins I,,: ,). But what objection reveals is the
social (and emotional) effects of such acts of ethnographic description that pull apart
socially constitutive knowledge,
18
particularly when they take similar (here, textual)
form and potentially exist within the same public space. We may not realize it, but our
analyses can be experienced as profoundly disempowering; they may provoke claims of
serious damage to professional reputations.
Then, there is the general problem that knowledge born of inter-subjective experi-
ence, when re-contextualized for a different audience within a broad analytical schema
(Descola :oo,), can produce a disconcerting misrecognition among those who shared
the experience. In Davids research we, his colleagues, have become objects of study,
they complained. As Hastrup puts it, our ethnographic work involves objectication, as
the gradual transformation of eldwork relations into object-knowledge apart from
relations (:oo: ,o).
My writing, then, ruptured relations and broke the rules of fair play within a
professional team of which I was a member. Undoubtedly this is partly what lay behind
David Mosse 945
the strong expressions from project managers and consultants who wrote (and at our
last meeting spoke) of the loss of trust, of being hurt by a valued friend and respected
colleague of long association. Of course, I did not intend this it upset me too; but
perhaps it is inevitable to what anthropologists do. Like others, in producing the
ethnographic account I refused the roles allocated to me, cut the networks of eldwork
(Strathern I,,o).
19
I had to disembed myself, erect boundaries, or put distance between
myself and the social worlds I described such that the academic individual was seen to
deny the moral person of eldwork.
20
Moreover, I now made inferences which my
subjects could not share while holding their own; based on evidence which their
schemes of understanding would never generate (Hastrup :oo: o,). I made connec-
tions fromaspects of experience not captured in current categories (Hastrup :oo: o,)
that were damaging. Meanwhile the disagreement that I had brought within the com-
munity itself weakened the hardness of project facts (cf. Rorty I,,I). And the further
my ethnographic representations travelled from arenas of social negotiation from the
consultant report, the conference paper, the journal article, to the book and the wider
and more public their consumption, the more strands were broken and the greater the
anger and anxiety.
21
Unfolding controversy
Let me now turn from the question of what my colleagues objected to, to that of how
they did so between January and May :oo. My principal observation is that my critics
reacted not by engaging with my text, but by challenging the boundaries that my
ethnographic writing introduced. They refused a textually mediated process in favour
of a socially mediated one that would, in some way, re-embed the production of
representations (research outputs) into the elds of power, the moral community, or
the family of the project.
This meant, rstly, that for over fteen months my colleagues failed or refused to
send written comments on the text. Many deadlines passed. They could not give written
comments, they said, because we disagree so fundamentally with your version of events
and the conclusions you draw ... it would take weeks and [would] not be an effective
way of communicating. In fact the text itself was dismissed as an independent object
:,o pages of difcult academic writing that development professionals and practitio-
ners cannot nd time to read, noted the team leader, adding that it is only by me
writing toX [a manager] and asking him to read specic paragraphs that I got him to
respond. Nor were they interested in making a substantive response, putting on record
reactions and alternative points of view; something which fully accepting their right
to reply I offered to do by means of a postscript, or opening a web-site.
What my colleagues repeatedly demanded was that we meet as a team to discuss my
draft and how it should be re-written, section by section over a period of three or so
days, and that, before we meet, I declare my preparedness to make changes. As the
dispute unfolded, they were careful to say that they did not insist on particular changes,
but that they were condent that I would be persuaded by their point of view and
would want to re-write many sections of the book. What they insisted on was not
textual change per se, but the social-emotional process a persuasive team workshop
that would produce such change through my re-inclusion in the moral community and
history of the project. Theirs was a moral critique of, and practical challenge to,
ethnographic exit. Only through being socially re-embedded could the text becomefair
and balanced, and the project as an interpretive community be re-constituted; if not
David Mosse 946
by appeal to friendship, loyalty, or obligation, then by implicit threat (damage
and defamation). Correspondingly, for my manager-critics to concede to a socially
disembedded academic process (that disallowed moral pressure over the author)
through textual engagement, written responses, and the debating of alternative points
of view would be profoundly disempowering.
22
And I could not have re-subjected
myself to the rules and power of eld relationships, or dismantled the boundary
between the dynamics of project life and writing, without abandoning my ethno-
graphic project or the integrity of my analysis.
So, I resisted the fallacy that the social has to be analysed socially and that evidence
is a matter of consensus (at least in this context), and did not accept that after one-and-
a-half years the objectors had had insufcient opportunity to share, discuss, and
respond to my draft. It was when I made it clear that I was unwilling to subject my
account to the adjudication of a select group of informants (or to meet and suspend
publication with that objective) that they protested to my publisher and academic
managers, now adding that the team considered that my study breached the ASA
Ethical Guidelines on at least four points concerning both the basis (negotiated
consent) and the outcome (harmful effects) of the research.
23
But this remained a second-order negotiation: not about facts or interpretations, but
about the terms of exchange textual or social, about whose rules would apply. My
senior academic colleagues who were approached were mystied by the objectors
intensely personal efforts (through their leader) by the repeated phone calls, emails,
offers to travel to London personally to show them sections of the book that disturbed,
to y team members from India and yet their persistent refusal, as one put it, to
comply with our request that you supply a written list of what you assure us are
substantial objections regarding the text. He continued: You have referred to these
objections repeatedly in written communication. Yet you remain unable to specify to
those ... you have involved in the procedure (pro-Director, Ethics Committee, Depart-
mental Convenors, not to mention the ASA) the precise substance of your objections.
This is a most serious matter.
But, of course, the objectors appeal to academic authorities was not intended to
open up the text or concede to conventions of scholarship, but rather to augment the
social persuasions of the moral community by bringing a disciplinary power to bear
on me, the author. The appeal was for adjudicators of disagreement in an editorial
process that should be collective. But to expect senior professors to play roles which
erase the boundaries that preserve academic independence rather than defend them
was a misjudgement. Which is not to say that there was no equivocation, for example,
on whether the university had a primary duty [to] respond to [an] accusation that a
publication will seriously damage third parties ... or to support academic freedom
(senior academic, internal correspondence on the case). However, those to whom my
critics appealed including the ASA Committee that met in March :oo concluded
that they had no remit to adjudicate, or act as a court, for what were only ethical
guidelines. And there was a private view that neither the basis nor the outcome of my
research contravened them. Consent to research had been given, and as a senior
colleague put it, the absence of attery is not harm.
The university authorities did agree that there should be a one-day meeting at which
the university, the ASA, and the two parties would meet, but that this would not be any
kind of court of arbitration, that it did not have to resolve differences, that it could not
impose on me any obligation to make particular changes to the book. They insisted that
David Mosse 947
this meeting would take place only on the basis of a full and detailed list of errors and
objections provided by the books critics showing evidence of inaccuracy or unargued
bias that would support allegations [made] against the professional standing of a [an
academic] colleague.
24
Academic rules of procedure had been enforced and with them
a quite different sense of open and fair which re-constituted the boundary between
social life and ethnographic analysis (eld and desk). The meeting did go ahead on that
basis, and it was preceded by a detailed list of comments. The shift to formal academic
process was irreversible. Very sorry it has come to this, the team leader said to me on
the day. In truth, he was.
The written comments for the meeting were in their way extraordinary. They lled
fty-six pages, and were categorized into a scale of I to in decreasing seriousness,
starting with statements regarded as defamatory or potentially damaging to profes-
sional reputations of identiable individuals, most of which concerned allusion in the
text to motives, compromises, or conicts that departed from ofcial representations of
roles and processes. While the language of defamation was sufciently serious for me to
take legal advice, and while the objections had been helpful in indicating alterative
points of view or correcting certain factual errors, the more I examined these com-
ments, the less they seemed to be matters of substance (evidence or argument) and the
more they invoked the moral community. That is, their concerns were matters not of
ontology but of relational epistemology. The personal and accusatory tone was unmis-
takable. They were about a person not a text: David should know this ... David, quite
frankly this comment is not worthy of you ..., etc. One independent participant in the
dispute commented, after the event, They had one, big fundamental objection [an
epistemological one] but they did not know how to write it so they provided fty-six
pages of minor objections in its place.
25
Certainly, assertions of the moral community, anger at the breach of its codes, the
hurt of being judged, the bafement of diverging epistemologies, were palpable in the
highly charged day-long ritual of objection chaired on : April by a former DFID Chief
Adviser, an anthropologist well placed to look both ways in the divide between
academics and development managers. She allocated a balance of time for the com-
plainants (now including four who had own in from India), for myself, and for
independent comment from representatives of the university and the ASA, among
others. Since the format presentation and response was academic and did not
require or allow a drive to resolution, it restrained the social control of an anthropo-
logical text. I listened carefully to the objections raised, responded, and undertook to
review my text in the light of the proceedings. In the end, I did not change my analysis,
although I claried its purpose, and modied phrasings that offended, where I judged
this appropriate. Thus concluded an improvised procedure that gave a green signal for
the publication of the ethnography. It did so ultimately through re-afrmation of the
Malinowskian boundary between eld and desk. Those who travelled to London
expecting that moral and persuasive pressure would result in substantive changes to my
ethnographic text were deeply disappointed.
Implications
For me, this was a signicant series of events raising important questions for ethno-
graphic practice and the status of anthropological knowledge, and I would like nally
to come to some of these.
David Mosse 948
The rst concerns the truth of an ethnographic account. The re-working of eld
relations and the contest over boundaries that I believe this dispute to have been about
were none the less expressed in terms of a challenge to the veracity of my evidence and
interpretation. And there is no question that objection disturbed the analysis. Ques-
tions were raised in my own mind: had I got this or that wrong? Another round of
research and reection began, taking me back from the text to the interview tapes, my
eld notes, the source materials, the studies, and the shared experiences that under-
pinned it, to verify and clarify. I emailed and phoned those who I had worked with on
the project and whom I had interviewed. An analysis that exists within a eld of
objections has to be sure of itself. Moreover, to borrow Albert Hirschmans (I,,o)
terms, if anthropologists of development are going to turn ethnographic exit from the
self-deceiving loyalties of policy and projects into voice in public, they have to be able to
defend their accounts.
26
But howwas I todefendtherightness of my ethnography against those whocouldsay,
youare wrong, I was there, orwhat evidence doyouhave tobackthis statement? or even,
come on, we knowyou! Onthe one hand, I didnot feel it was sufcient tosay,I was there
too, and this is my subjective interpretation; take it or leave it. The matter of factual
accuracy remained. Surely, as Malinowski taught us, it was necessary to allow some
separationof facts frominterpretation, tobeexplicit about theactual workof inquiryand
the material and experiences upon which my generalizations were based; to unpick that
seamless Geertzian interpretative web spun from eld notes to ethnography (Spencer
I,8,: I,o). But onthe other hand, shouldI be drawnintodefending this or that statement
withreferencetothis or that specicpieceof supportingevidence(event or conversation)
on its own? It is not just that this might compromise the condentiality of informants
including those in the meeting-room contradicting themselves but that, as in any
ethnography, the case anddiscursive material was illustrative, chosenfor its succinctness,
but drawn from a long interactive experience over years the many encounters, events,
memories, notes, reports, conversations that make up eldwork. After all, this is where
ethnography differs from investigative journalism.
In any case, as Kirsten Hastrup points out, anthropologists can never actually prove
the rightness of their generalizations with reference to evidence or experience (as an
independent measure of validity), since these are neither separate from, nor prior to,
the anthropologists own frame of interpretation, the pre-existing scheme of objecti-
cation that transforms facts into evidence or imputes causation (:oo: ,o, oI).
27
At
the very least, anthropologists need to examine the social basis of their own evidence-
making. They need to examine their ownpoint of view their personal and academic
predilections, judgements, and aesthetics as the product of social conditions (and
professional location), something that Bourdieu (:oo,) referred to as participant
objectication.
Since it cannot prove through evidence (the more so in relation to institutions which
in-build deniability), what ethnography aims at, Hastrup suggests, is a kind of expla-
nation beyond the truth of events themselves; it is not simply knowledge about
particular events, practices and ideas, but about the processes by which these come to
appear meaningful, perhaps inevitable or mandatory, possibly contestable or even mad
(:oo: o8). The kind of connections I made between individual statements, actions,
events, and larger schemes of policy in a development project (and so the kind of
explanations I offered) came, moreover, from my being implicated in its processes
(Hastrup :oo: oo). As an anthropologist I do not have knowledge or experience of
David Mosse 949
culture, but experience contingent events like everyone else and make sense of them
(Hastrup :oo: o8). Rightness, Hastrup insists, is an epistemological awareness not an
ontological certainty (:oo: oo).
28
Others, too, have concluded that ultimately eldwork is a kind of social apprentice-
ship, often beyond language, through which anthropologists negotiate the opacity of
social life its mutual interpretations and concealments along with everyone else
(Bloch I,,I; Jenkins I,,: I). For this reason, there is no neutral or uninvolved
knowledge (an understanding that everyone might share [Jenkins I,,: ,]), no
sharp divide between anthropologist and subject, eldwork and the processes of every-
day social life. Insider ethnography is only a case of the general situation. It is a matter
of Latours relativist sociology in which I am a project actor along with others; my
policies and points of view stand with theirs; as does my analysis (I,,o: I,,). My
colleagues and associates are also sociologists offering theory, explanations, trying to
stabilize the project world from their varied points of view. But then, potentially, there
are as many stories and authors as actors.
Gupta and Ferguson only re-state the issue when they comment, sensibly enough,
that the interpretative account that is anthropological always coexists with other
forms of knowledge, and see the political task not as sharing knowledge with those
who lack it, but as forging links between different knowledges that are possible from
different locations (I,,,: ,,, my emphasis). But it is precisely the nature of these links
that is at issue, especially where the knowledge in question is that of the ethnographers
subjects. My ethnographic account does not just stand alongside or compete equally
with other or preceding ones, it attempts to encompass them in the guise of subject
matter.
29
My narrative adds interpretations to those of actors whose experience I share;
it explicates different points of view, it tries to become the meta-narrative; and it is the
one whose divergent evidence and inferences will be challenged as public misrepresen-
tation, not least because the interpretations that ethnography adds come from reec-
tion on the experience of dislocation and alterity.
30
If the autonomous production of anunderstanding or representation of others has
lost meaning as a goal for ethnography, then perhaps it can partly be re-instated
through efforts that risk placing ethnographies back within the eld of relations that
they describe. As Latour puts it, what the social sciences can do is to re-present the
social to itself: That is, not to dene the unknown structure of our actions ... [but to be]
able to modify the representation the public has of itself fast enough so that we can be
sure that the greatest number of objections have been made to this representation
(:ooo: I:o, emphasis in original).
31
This view may be suited to an anthropology that wants to sustain claims to right-
ness in public without either assuming epistemological privilege or retreating into
narcissistic reexivity (Bourdieu :oo,: :8I). The process of objection an emotion-
lled relationship reminds us that ethnographic knowledge is at root a social
phenomenon, and persistently so (Hastrup :oo: ,o); and that anthropological
right-ness in the sense both of veracity and of entitlement to represent is also
social. Rightness is not a matter of the ontological status of our evidence (and this
my critics grasped), and not just an individual epistemological awareness, but the
outcome of social contests over boundaries and the location of knowing. And this
suggests that ethnography may sometimes require institutional experiments beyond
eldwork, bringing ethnographers and their subjects together around its written
outputs, experiments in objection, and the defence of ethnographic rightness such as
David Mosse 950
the one I have described. Of course these exchanges cannot seek resolution. After all,
following two high-prole UK government inquiries in :oo it was clear that
weapons inspector David Kelly and BBC reporter Andrew Gilligan together had had
a kind of ethnographic insight or inference beyond the truth of events which was
simultaneously entirely right and fully deniable.
32
To conclude: contests after the eld may be crucial to the recovery of anthropo-
logical claims, not least because they do not reveal any stable constellation of power in
social research. Anthropologists have the power to represent; and their informants have
different capacities to object. Objections challenge anthropological authority, and
should be welcomed for disallowing analytic closure, for sharpening our historical
sense by refusing the ethnographic present, and making us clarify our generalizations
(although, as I have shown, in their own structure, objections may also conrm the
essentials of our interpretations anything you say may be taken as evidence ...!).
Objections also remind us that however we may try to convince people that we are
right, ultimately ethnographers have to concede that what we constitute as evidence is
not separable from our relationship with our informants (Hastrup :oo,). Conicts
arise when, as Hastrup puts it, knowing a subtle (epistemological) relationship
between subject and object, becomes knowledge a (near-ontological) certainty
(:oo,), separate and in public. As ethnographic inquiry proceeds, relationships become
evidence, which complicates the use of evidence as an independent measure of valid-
ity (:oo,). At the same time, ethnographic representations have the potential to
unravel when our informants (as did mine) attempt to unpack our evidence back into
relationships with them.
When anthropologists resist this, contests may unfold in which it becomes clear that
anthropological writing is not, after all, an individual effort. Ethnographys objecti-
cations (or its style of reasoning) are themselves (no less than policy models) stabilized
socially and depend upon authoritative supporters, specic institutional discourses,
and (as I discovered) processes of endorsement. We should not forget that in the end
anthropological knowledge is a social achievement (Crick I,8:: :o, in Hastrup :oo:
,o), one that, as Bourdieus notion of participant objectication suggests, can be
subject to the same sociological re-conguration as policy discourse, and which reveals
similar contradictions between individual practice and the professional models
(including Malinowskian eldwork) that reproduce the discipline (Grimshaw & Hart
I,,,: ,,).
Perhaps the outcome could have been other, but in this case academic discourse itself
demonstrated considerable power, even against quite determined objectors. When
studying up, we may regard those who try to censure independent research as powerful
we may imagine that we speak truth to power. But the outrage at rupture can just as
easily reveal the fragile hold that those who appear to be in power in political,
administrative, or policy systems have over their legitimizing representations (or the
enormous effort needed to sustain them).
It is this threat to authorized representations that makes the study of public policy
and institutions so challenging, especially as insiders. On the one hand, at the extreme,
defamation proceedings might rule out the possibility of such ethnography. On the
other, where harm is perceived to have been done, anthropologists have to engage with
that perception. Either way, anthropology does not have the option (moral or episte-
mological) of a devotion to science that disregards social relations that are the basis of
its knowledge.
33
The right to academic knowledge has to be negotiated among other
David Mosse 951
legitimate claims. And the negotiation of ethnography as asituated intervention rather
than a disinterested observation (Gupta & Ferguson I,,,: ,8) requires that its practi-
tioners are clear on their position, perspective, and purpose.
Perhaps above all my case shows that the power of ethnography lies in the fact that
not only anthropologists but also their texts are active agents in the worlds they
describe, enlivening action in particular ways. Thus the social mobilization of objec-
tions to my book is as integral to project action as other events it described. And
when writing becomes as much a part of our engagement with our subjects as eld-
work, we may be forced to address the Malinowskian division that has allowed an
individualist free-ow rulelessness to writing in contrast to the rule-governed expec-
tations of eldwork, to borrow a useful distinction from Parkin (:ooob: :oo). Col-
laborative ethnography is not a solution. But maybe anthropologists should at least
anticipate the continuation of eldwork relations into writing and publication, and
the ruptures that may arise. Can we really blame our informants for their misunder-
standing of our intellectual goals? Can we prepare informants for ethnographic
outputs? Does the ethnographic account have to be in the nature of an ambush
on social life? These are not issues that current ethical guidelines, framed on the
basis of a narrower conception of power relations in research, are well placed to
clarify.
Inviting objections will not erase the questions of power that surround research, but
it may at least bring an ethnographic awareness to our writing. Writing itself can then
be viewed within an anthropological frame that accounts for the relationship between
ideas and social relations, and as party to transactions between moral persons, in my
case within the ambiguous gift-world of an aid project.
Perhaps ethnographers of policy, professionalism, or international development
domains in opposition to which Malinowski dened the ethnographic eld may have
a key role in re-examining the methodological and institutional foundations of social
anthropology, a task that was at the centre of Malinowkis own work.
NOTES
This a slightly modied version of the lecture delivered on : June :oo, at the London School of Economics
and Political Science. I am grateful to the LSE Department of Anthropology for the invitation to give that
lecture. Thanks for helpful comments and suggestions are due to Rosalind Eyben, Richard Fardon, and Ingie
Hovland, to those who listened to a preliminary version given in Edinburgh, and above all to my objectors
for their engagement with my ethnographic writing. I am also grateful to Glenn Bowman and to my
anonymous reviewers for helpful suggestions in nalizing the article.
1
Writing, Hobart suggests, is the antithesis of dialogue, which academics reattach to textualizing (I,,o: :,
n. :,).
2
As Ingie Hovland reminded me (pers. comm., , April :oo,); see also Brettell (I,,,a).
3
The more celebrated cases of negative reactions to anthropologists work from those who have been the
objects of study, include outrage among the Ik when they were informed of the content of Turnbulls
derogatory The mountain people (I,,:; Heine I,8,); the community and press reaction to Vidich and
Bensmans ethnography of a village in upstate New York (Small town in mass society, I,,8); the public
controversy in Mexico following the Spanish translation of Lewiss Children of Snchez (I,oI); or the upset
from the people of Ballybran (Ireland) described in Scheper-Hughess Saints, scholars and schizophrenics
(I,8:). Brettells collection When they read what we write (I,,,a) provides an overview of these cases and
brings together a range of more recent experiences in ethnographic writing. In what Brettell describes as a
revolution in readership (I,,,b: ,), informants confronting public representations of their lives and words
accuse anthropologists of betrayal and broken condences, public shame, and damage to reputations or
self-images, of violations of condentiality or of the etiquette of gossip. They take exception to too little or
too much anonymity, to the fragmentation of lives in text, and they are disturbed by the idiom and
David Mosse 952
terminology of social science, in which they do not recognize themselves. Fuelled more by accounts in the
press and rumours of the book than by actually reading it, they respond by rejecting the anthropologist as
an outsider/stranger who can never understand (Brettell I,,,a).
4
Relational in the sense both that knowledge is collaborative, dialogical, gained by way of relations; and
that (in consequence) the relationships between researcher and object of inquiry become a property of the
object itself (Hastrup :oo: ,,).
5
There is a parallel here with situations in which ethnographic subjects have strongly reied cultural
self-representations e.g. cultural or ethnic nationalist groups where anthropological re-interpertation
may even carry mortal risk (Whitaker I,,o: ,).
6
Imagine, for example, a visit by the ofcial DFID monitoring mission when interpretative possibilities
are constrained through the organization of space and time in visited villages so as to resemble the policy text
read by outsiders; the co-existence of completed PRA (Participatory Rural Appraisal) maps on the ground,
the smart treated landscape, new woodlots, and colourful groups of women provide the simultaneous
presence of the village plan and its execution contingency and time are suppressed; individuals and events
subordinated to policy ideas.
7
A distinction borrowed from Baudrillard (I,88: :o,) via Hobart, who explains the work of development
organizations as asserting determinate representations (I,,,: o, Io).
8
And this of course is no less true of British domestic policy. As his former chief speech-writer notes, Blair
promises to deliver on things he has little power over: exam results, crime levels, cancer mortality rates
(Hyman :oo,: ,8o).
9
Expressing an individual rather than an institutional view.
10
Although (I warned), such assertions might become self-fullling by giving the damaging impression
that the project was a weak organization able to survive and attract funds only on the basis of fragile
representations.
11
As recalled by Rosalind Eyben (pers. comm., June :oo).
12
As I note in the book itself, In spite of formal demands for objectivity and independence, experts are
charged with producing, and themselves intentionally construct, the evaluation story as a shared commod-
ity (Phillips & Edwards :ooo: ,,) (Mosse :oo,a: I,8).
13
A point made by Rosalind Eyben (pers. comm., June :oo).
14
A proper project history would begin with policy intention and design and explain how this was
implemented. My account reversed this, starting with events and relationships before turning to rationalizing
representations.
15
One can also say that while narratives of success emphasizing expert ideas are theory-rich, those of
failure are by contrast event-rich (Mosse :ooo).
16
The ambiguous concept of embarrassment (to DFID or its partners) sets the criteria for refusing use of
data from consultancy in DFID contracts.
17
Presumptions about the possibility of consensual narratives, or the unity of native voice (Lassiter :oo,),
suggest that critical anthropology could learn from the critiques of participation in development (e.g. Cook
& Kothari :ooI).
18
Making social maps from gaps and spaces, listening for the unsaid rather than the guarded statement
(Dresch :ooo: I:,).
19
This is to say that ethnography itself (of a certain kind) is among other mechanisms Strathern discusses
proprietorship and certain kinship arrangements that cut the self-enlarging social networks that actor
network theorists describe; and that it can do so not just conceptually, but also socially.
20
This is not to say that writing need always have this effect or that it cannot in some instances work as a
bridge (Ingie Hovland, pers. comm., , April :oo,); indeed some of my relationships from the eld were
strengthened through this writing.
21
Comments that were incorrect in the book were previously acceptable in my consultancy reports, which
remained internal and had a restricted readership. It is signicant also that I (or my colleagues) could publish
on crops or tribal communities without offence, since these realms were distant from the social world of my
critics and subject to a technical discourse.
22
It is for this reason that dialogical or reciprocal ethnography, or the inclusion of (negative) native
responses (Lassiter :oo,), ultimately fails to address questions of power in ethnography.
23
The latter revolved around the serious concerns that the book will harm professional reputations,
especially of Indian mangers, uncomfortable with academic prose and ill equipped to defend themselves, who
had not been given adequate chance to reply, and who, while not named, could be identied.
24
Correspondence from the university to the objectors, March :oo.
David Mosse 953
25
Rosalind Eyben (pers comm., June :oo).
26
In this controversy (and in the book itself), when defending an interpretation that I claimed to be true,
I revealed a representationalist view of truth (that mind or language allows for the representation of reality,
and that some representations are more accurate than others). I insisted upon evidence independent of
relationships and resisted assertions that truth was a matter of agreement and consensus. An anonymous
reviewer cogently summed up the irony in the stand-off between Mosse and his objectors: their positivist
ontology was protected by a relational epistemology; his relationalist ontology is defended by recourse to
what is ultimately (at second order) a positivist theory of truth. (JRAI Reviewer comments, September :oo,).
27
Ethnographic comprehension is necessarily deductive (Descola :oo,: o,).
28
Anthropological understanding, Descola notes, comes from confronting acts/utterances with our own
responses to the same circumstances, and from identication with the motivations that may lie behind the
actions of others, rather than with the culturally codied responses that these motives generate (:oo,: ,o).
Similarly, Bourdieu comments on the importance of anthropologists using their own native experience in
order to understand and analyse other peoples (:oo,: :8,).
29
I am grateful to Richard Fardon for clarifying this point (pers. comm., I, October :oo).
30
In the case of insider ethnography, social exit rather than entry is the critical shift for defamiliarization.
If we agree with Bowman (I,,8) that ethnographers abroad cannot communicate another culture to their
readers, but only the dislocation of their experience of alterity, then ethnographers at home convey to their
readers and to their colleagues or co-residents the equivalent conceptual distortion of exit from familiar
shared assumptions. Considered through the readers own experiences, the fractures of exit, like the fractures
of eldwork, can be embraced as insightful or rejected as disturbing (or wrong).
31
Others inuenced by Wittgensteins later philosophy point in a similar direction. Whitaker suggests
treating ethnography as a set of pedagogic experiments, descriptive tries, or exercises in publicly displayed
learning, judged not by how well they construct their object, but by how well they bring the parties involved
into some kind of lucid contact (I,,o: 8), and Hobart suggests reframing the problematic anthropological
goal of understanding in terms of mutual recognition (I,,o: ,I). Whether objection or the more bloodless
lucid contact (or mutual recognition) is the better metaphor will depend on the degree of contestation and
what is at stake.
32
The reference being, of course, to the controversy over the representation of the threat of weapons of
mass destruction that justied the decision to go to war in Iraq in :oo,, and eventually led to Kellys suicide;
and to the Hutton Inquiry (:oo) into the circumstances surrounding the death of Dr Kelly, and the Butler
Inquiry (:oo), which reviewed intelligence on weapons of mass destruction.
33
Thanks to Ingie Hovland for raising this point and some of the questions in the next paragraph in
response to an earlier draft.
REFERENCES
Avuvxvv, E. I,,I. The new anthropology and its critics. Man (N.S.) o, ,-o,.
B.uuviii.vu, J. I,88. The masses: the implosion of the social in the media. In Jean Baudrillard: selected
writings (ed.) M. Poster (trans. M. Maclean), :o,-I,. Cambridge: Polity.
Biocu, M. I,,I. Language, anthropology and cognitive science. Man (N.S.) io, I8,-,8.
Bouvuivu, P. I,,,. Outline of a theory of practice (trans. R. Nice). Cambridge: University Press.
:oo,. Participant objectication. Journal of the Royal Anthropological Institute (N.S.) p, ::,-,,.
Bowx.x, G. I,,8. Radical empiricism: anthropological eldwork after psychoanalysis and the Anne
Sociologique. Anthropological Journal on European Cultures o: i, ,,-Io,.
Bvv11vii, C.B. (ed.) I,,,a. When they read what we write: the politics of ethnography. London: Bergin &
Garvey.
I,,,b. Introduction: eldwork, text, and audience. In When they read what we write: the politics of
ethnography (ed.) C. Brettell, I-:. London: Bergin & Garvey.
Buv.wos, M. I,,8. The extended case method. Sociological Theory :o, -,,.
C.s1viis, M. I,,o. The rise of the network society. Oxford: Blackwell.
Coox, B. & U. Ko1u.vi :ooI. Participation, the new tyranny? London: Zed Books.
Cvicx, M. I,8:. Anthropological eld research, meaning creation and knowledge construction. In Semantic
anthropology (ed.) D. Parkin, I,-,,. London: Academic Press.
Dvscoi., P. :oo,. On anthropological knowledge. Social Anthropology :, o,-,,.
Dvvscu, P. :ooo. Wilderness of mirrors: truth and vunerability in Middle Eastern eldwork. In Anthropolo-
gists in a wider world: essays on eld research (eds) P. Dresch, W. James & D. Parkin, Io,-:8. Oxford:
Berghahn Books.
David Mosse 954
Fvivux.x, J. :oo. The relocation of the social and the retrenchment of the elites. In Forum on The retreat
of the social: the rise and rise of reductionism (ed.) B. Kapferer. Social Analysis {8: , Io:-8.
Gvvv1z, C. I,,,. From the natives point of view: on the nature of anthropological understanding. In
The insider/outsider problem in the study of religion: a reader (ed.) R.T. McCutcheon, ,o-,,. London:
Cassell.
Gvixsu.w, A. & K. H.v1 I,,,. The rise and fall of scientic ethnography. In The future of anthropology: its
relevance to the contemporary world (eds) A. Ahmed & C. Shore, o-o. London: Athlone.
Guv1., A. & J. Fvvcusox I,,,. Discipline and practice: the eld as site, method and location in anthro-
pology. In Anthropological locations: boundaries and grounds of a eld science (eds) A. Gupta & J. Ferguson,
I-o. Berkeley: University of California Press.
H.s1vuv, K. :oo,. Convincing people: knowledge and evidence in anthropology (Abstract). In Abstracts:
American Anthropological Association, :o:nd Annual Meeting, Chicago, :,-:, November, :o. Arlington, Va.:
American Anthropological Association.
:oo. Getting it right: knowledge and evidence in anthropology. Anthropological Theory {, ,,-,:.
Hvixv, B. I,8,. The mountain people: some notes on the Ik of north-eastern Uganda. Africa ,,, ,-Io.
Hon.v1, M. I,,,. Black umbrellas: the implication of mass media in development. Unpublished paper
presented at EIDOS Workshop on Globalization and Decivilization, Agricultural University of Wagenin-
gen, November.
I,,o. Ethnography as a practice, or the unimportance of penguins. Europa i, ,-,o.
Hivscux.x, A.O. I,,o. Exit, voice, and loyalty: responses to decline in rms, organizations, and states.
Cambridge, Mass.: Harvard University Press.
Hsx.x, P. :oo,. : out of :o: from Downing Street vision to classroom reality. London: Vintage.
Jvxxixs, T. I,,. Fieldwork and the perception of everyday life. Man (N.S.) ip, ,,-,o.
K.vvvvvv, B. :oo. Introduction: the social construction of reductionist thought and practice. In Forum on
The retreat of the social: the rise and rise of reductionism (ed.) B. Kapferer. Social Analysis {8: , I,I-oI.
Kuvvv, A. :oo,. Alternative histories of British social anthropology. Social Anthropology :: :, ,-o
L.ssi1vv, L.E. :oo,. Collaborative ethnography and public anthropology. Current Anthropology {o, 8,-Ioo.
L.1ouv, B. I,,o. Aramis, or the love of technology (trans. C. Porter). Cambridge, Mass.: Harvard University
Press.
:ooo. When things strike back: a possible contribution of science studies. British Journal of Sociology
,, Io,-:,.
Lv.cu, E.R. I,oI. Rethinking anthropology. London: Athlone Press.
:oooa [I,,,]. Malinowskis empiricism. In The essential Edumund Leach, vol. I, Anthropology and
society (eds) S. Hughes-Jones & J. Laidlaw, -o:. New Haven: Yale University Press.
:ooob [I,,o]. Social anthropology a natural science of society? In The essential Edumund Leach, vol.
I, Anthropology and society (eds) S. Hughes-Jones & J. Laidlaw, 8o-,,. New Haven: Yale University Press.
Lvwis, O. I,oI. The children of Snchez: autobiography of a Mexican family. New York: Random House.
M.iixowsxi, B. I,::. Argonauts of the Western Pacic. London: Routledge & Kegan Paul.
I,o,. A diary in the strict sense of the term (trans. N. Guterman). London: Routledge.
M.vcus, G. I,,8. Ethnography through thick and thin. Princeton: University Press.
Mi1cuvii, T. :oo:. Rule of experts: Egypt, techno-politics, modernity. Berkeley: University of California Press.
Mossv, D. :oo. Social analysis as product development: anthropologists at work in the World Bank. In The
development of religion/the religion of development (eds) A.K. Giri, A. van Harskamp & O. Salemink, ,,-88.
Delft: Eburon.
:oo,a. Cultivating development: an ethnography of aid policy and practice. London: Pluto Press.
:oo,b. Global governance and the ethnography of international aid. In The aid effect: giving and
governing in international development (eds) D. Mosse & D. Lewis, I-,o. London: Pluto Press.
:ooo. The ethnography of failure and the failure of ethnography. Unpublished paper for the Work-
shop Beyond Result: Failure as a Moral, Economic, and Socio-political Discourse. Cambridge, :o March.
P.x.sio1ovouiis, P. :oo:. Anthropology consultancy in the UK and community development in the Third
World: a difcult dialogue. Development in Practice :i, ,-,8.
P.vxix, D. :oooa. Templates, evocations and the long-term eldworker. In Anthropologists in a wider world:
essays on eld research (eds) P. Dresch, W. James & D. Parkin, ,IIo,. Oxford: Berghahn Books.
:ooob. Epilogue: eldwork unfolding. In Anthropologists in a wider world: essays on eld research (eds)
P. Dresch, W. James & D. Parkin, :,,,,. Oxford: Berghahn Books.
Puiiiivs, S. & R. Euw.vus :ooo. Development, impact assessment and the praise culture. Critique of
Anthropology io, ,-oo.
David Mosse 955
Qu.vivs v.x Uvvovu, P. I,88. The hidden crisis in development: development bureaucracies in between
intentions and outcomes. In The hidden crisis in development: development bureaucracies (eds) P. Quarles
van Ufford, D. Kruijt & T. Downing, ,-,8. Tokyo and Amsterdam: United Nations and Free University
Press.
A. Kux.v & D. Mossv :oo,. Interventions in development: towards a new moral understanding of
our experiences and an agenda for the future. In A moral critique of development: in search of global
responsibilities (eds) P. Quarles van Ufford & A. Giri, ,-o. London and New York: Routledge.
Rov1s, R. I,,I. Objectivity, relativism, and truth: philosophical papers, vol. I. Cambridge: University Press.
Scuvvvv-Hucuvs, N. I,8:. Saints, scholars, and schizophrenics: mental illness in rural Ireland. Berkeley:
University of California Press.
Svvxcvv, J. I,8,. Anthropology as a kind of writing. Man (N.S.) i{, I,-o.
:ooo. British social anthropology: a retrospective. Annual Review of Anthropology ip, I-:.
S1v.1uvvx, M. I,,o. Cutting the networks. Journal of the Royal Anthropological Society (N.S.) i, ,I,-,,.
Tuvxnuii, C.M. I,,:. The mountain people. New York: Simon & Schuster.
Viuicu, A.J. & J. Bvxsx.x I,,8. Small town in mass society: Class, power, and religion in a rural community.
Princeton: University Press.
Wui1.xvv, M. I,,o. Ethnography as learning: a Wittgensteinian approach to writing ethnographic accounts.
Anthropological Quarterly op, I-I,.
Youxc, M. :oo. Malinowski: Odyssey of an anthropologist, :88,:,:o. New Haven: Yale University Press.
Une anthropologie antisociale ? Objectivit, objection et ethnographie des
politiques publiques et des communauts professionnelles
Rsum
La mthode ethnographique de Malinowski nous a laiss en hritage la sparation du terrain et du
bureau . Le savoir des anthropologues est indissociable de leur relation avec ce quils tudient
(lpistmologie est relationnelle), mais lcriture ethnographique coupe les liens du travail de terrain,
disperse le rseau et dresse des frontires : elle est ncessairement antisociale. Lorsque les anthropologues,
dans leur tude de ce que les gens croient, disent et font (et les incohrences entre les trois), sintressent
aux institutions interconnectes qui composent le monde moderne, la politique et aux communauts
professionnelles dont ils peuvent galement tre membres, il leur devient plus difcile daller et venir entre
les mondes sociaux. Argumentant en faveur de limportance dune approche ethnographique des pratiques
des institutions de pouvoir, lauteur utilise des recherches rcentes sur laide internationale et le dvel-
oppement pour montrer comment les informateurs inuents sopposent aux comptes-rendus eth-
nographiques, rsistent ltablissement des frontires anthropologiques et tentent de dtricoter le
savoir acadmique pour le rinsrer dans les relations.
David Mosse is Reader in Social Anthropology at the School of Oriental and African Studies. He has worked
on the anthropology of international development and natural resources management, and is currently
writing on mission history, popular religion, and Dalit politics in south India.
Department of Anthropology and Sociology, School of Oriental and African Studies, University of London,
Thornhaugh Street, Russell Square, London WC:H oXG, UK. dm::@soas.ac.uk
David Mosse 956

Mosse 2006 AntiSocial

Uploaded by

Copyright:

Available Formats

Mosse 2006 AntiSocial

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Mosse 2006 AntiSocial

Uploaded by

Copyright:

Available Formats

Anti-social anthropology?

Objectivity, objection, and the

You might also like