Professional Documents
Culture Documents
WORKING PAPER
1
DIIS WORKING PAPER 2015:01
FINN STEPPUTAT
Senior Researcher, Peace, Risk and Violence, DIIS
FST@diis.dk
JESSICA LARSEN
PhD Candidate, Peace, Risk and Violence, DIIS
JELA@diis.dk
2
DIIS WORKING PAPER 2015:01
CONTENTS
Introduction 4
Ethnography: participant observation and beyond 5
Ethnography and the political 8
Political anthropology 8
Political ethnography 10
Interpretative Policy Analysis (IPA) 12
Conclusions 22
Literature 23
3
DIIS WORKING PAPER 2015:01
“Producing ethnography is ongoing, and so are relationships that make up the ‘lives’ of pol-
its effects” (Rancatore 2010: 75). icies. Interpretative policy and organization-
al studies have shown that the lives of poli-
cies are formed by dynamics and rationalities
that have less to do with their explicit aims
INTRODUCTION than with their capacity to bring together pol-
icy actors or to depoliticize highly political is-
The aim of this working paper is to sketch out sues, and with contingencies, turf battles, and
the contours of an emerging methodology institutional interests. We suggest that eth-
for studying global policy regimes at a time nography provides approaches that are well
when changing world orders, new actors, and suited to studying these kinds of dynamics
new technologies challenge and transform and contingencies which seem to be even
these regimes.1 In fields such as environment more important with the degrees of com-
and climate change, migration, development, plexity and encompassment that characterize
natural resources, trade, and state & peace- current global policies.
building, comprehensive policy regimes As Shore and Wright (2011) suggest, pol-
have developed to deal with problems at a icy studies can provide windows onto ‘larg-
global scale. The policy processes involve a er processes of governance, power and so-
wide range of international, national, trans- cial change that are shaping the world today’
and subnational, governmental and non- (Shore and Wright 2001: 1), which is a very
governmental actors; they blur or redefine good reason in itself for developing this kind
boundaries of private and public domains, of of studies. But they might also point to rea-
the state and the ‘non-state’; and they produce sons why policies often fail to reach their
effects (and affect) at many different levels. aims, why they can have more unintended
The paper focuses on the role of ethno- than intended effects, or why they continue
graphy in global policy studies. Ethnography to be reproduced despite poor ‘results’. One
is apt for studying meanings and effects of example would be state- and peace-building
policies among people ‘on the ground’ but policies that emphasize the importance of
may also be used to ‘study up’ (Nader 1972); context, local ownership and non-state ac-
that is to study policies among their makers, tors while consistently producing abstract,
translators (Latour 1986), and unmakers, and state-centered blueprints and marginalizing
to understand how assemblages of knowl- local actors (Baranyi and Desrosiers 2012).
edge, policy, practice and technology come Linked to the surging popularity of
together, transform, and fall apart. Closely ANT, as well as practice-oriented and prag-
associated with the study of context and con- matic approaches in social and political stud-
textualization, ethnography can help to show ies, the interest in ethnography seems to be
how the encompassing, abstract and mobile increasing in various disciplines and fields of
templates of global policies are articulated policy studies. Thus, we see how political eth-
in a contingent, unstable, and messy inter- nography, political anthropology, global eth-
nography, and interpretative policy analysis,
1
This working paper was written to frame the discussions are converging across disciplines to form a
at the international workshop, ’Global Political Ethnography:
Studying Policy Regimes in Flux’, at the Danish Institute for
dynamic field worth exploring with a focus
International Studies (DIIS), November 25-26, 2014. on global policy analysis. At the same time,
4
DIIS WORKING PAPER 2015:01
the use of ethnography in the analysis of dif- icy analysis, we need conceptually and theo-
ferent policy fields is uneven, and disciplines retically informed reflections. We will con-
that have incorporated ethnography vary in sider various contributions, including global
their understanding and use of the method- ethnography and others that provide valuable
ology. suggestions and building blocks. But there is
On this background, the purpose of this a lot more to be done conceptually and theo-
working paper is twofold: Firstly, we seek to retically to develop what we will call a global
contribute to the emerging cross-disciplinary political ethnography as such.
dialogue and provide some common ground The paper is organized in three parts: the
for discussions by introducing different aca- first looks into ideas and practices of ethno-
demic fields in which ethnographic studies of graphy; the second part focuses on approach-
politics and (global) policies are developing. es to studying policy and the political, includ-
We do not pretend to provide a full overview ing political ethnography, political anthropol-
of these extensive fields, but since it is our ogy and interpretive policy analysis; and the
impression that disciplinary developments third part discusses different contributions
to some degree have taken place in isolation that have taken on the challenge of study-
from one another – political ethnography ing ethnographically the lives of global pol-
separately from political anthropology,2 and icies in a context of accelerated processes of
analysis of international policies separately globalization through concepts of scale, net-
from domestic policy analysis3 – we will pres- work, field, apparatus and assemblages. To-
ent the different strands of ethnographic en- gether, these contributions draw the contours
gagement with the political as building blocks of an emerging field of global political eth-
to consider the development of a global nography.
political ethnography.
Secondly, the paper focuses on and seeks
to push the methodological discussions con-
cerning how to approach ethnographical- ETHNOGRAPHY: PARTICIPANT
ly the highly complex global policy process- OBSERVATION AND BEYOND
es that are currently developing. What are
the appropriate empirical scale(s) and units In this paper, ethnography is taken as a
of analysis when studying global policy net- collective term for method and methodology,
works? How are the voices and practices of that is, data collection techniques on
actors operating at different scales and in dif- the one hand and the consideration of
ferent sites balanced, weighed, and connect- epistemological and ontological issues in the
ed in policy narratives informed by ethno- research process on the other (Schwartz-Shea
graphic analysis? In order to identify sites, and Yanow 2002). What concerns us most
encounters, situations and materials where here is methodology. But since ethnography
ethnographic approaches can generate dif- has long been associated with the method
ferent and maybe more critical insights than of participant observation in particular, we
more conventional approaches to global pol- will start by discussing issues of method in
the scholarly literature that deals with the
2
See for example Joseph and Auyero 2007 ethnographic study of global governance and
3
See Yanow 2015 policy.
5
DIIS WORKING PAPER 2015:01
A great strength of ethnography is its ability scholars who have been ‘studying up’ (Nader
to generate empirical data, which otherwise 1972), looking into elites, security commu-
would not feed into scientific thinking and nities, criminal networks or similar, have ex-
policy analysis. Ethnography enables ground- perienced problems of ‘being there’. In such
ed knowledge by way of the researcher im- studies, access to fieldwork sites, informants
mersing herself in the setting(s) relevant to and knowledge can be particularly challeng-
the object of study and connecting the units ing, and the method of participant obser-
of data to an analytical whole that generates vation thus less applicable. Gusterson, who
intimate understandings of local practices studied nuclear power plant laboratories with
and social relations. Through critical reflec- no access granted, developed instead of par-
tion and theoretical consultation, the empir- ticipant observation an approach that he calls
ical experiences generated through fieldwork ‘polymorphous engagement’. Gusterson in-
can be translated into scientific discourses teracted with informants across sites, some-
that deepen, or even change, our understand- times in virtual form, and used a mix of in-
ing of social life and, perhaps, to the refor- terviews, newspaper and document reading
mulation of theory. (Gusterson 1997: 116). He even conduct-
Broadly speaking, we may character- ed ‘journalistic interviews’ and talked on the
ize ethnography as ‘the science of contex- phone to the same extent as participant ob-
tualization’ in which context is ‘construed servation in traditional fieldwork. Indeed, as
through personal relationships bound by self- Forsey (2010) has suggested, there are many
knowledge, expectation, and commitment, situations in which the ethnographer is par-
and by language, memory, and imagination ticipant listening, rather than participant observ-
to registers and relativities of experience be- ing.
yond the present here and now’ (Greenhouse Obviously, the intensified processes of
2010: 2). Thus, ethnography always involves globalization in the 1990s presented a differ-
‘experience-based inquiry into the interpre- ent kind of challenge to the traditional ethno-
tive, institutional, and relational makings of graphic approach in anthropology. In re-
the present’ (ibid.). sponse, Marcus (1995) introduced the idea of
Within anthropology, participant observa- multi-sited fieldwork. This approach prom-
tion has been the ethnographic method par ised a way of updating ethnography as a rel-
excellence since Malinowski. In the words evant methodology to study phenomena
of the Comaroffs, any first-hand knowledge dispersed across borders and articulated in
generated by participant observation has en- flexible networks. Multi-sited ethnography
joyed an ‘a priori privilege’ (Comaroff and broke with static and mono-locational con-
Comaroff 2003: 153). Thus, for anthropolo- ceptualizations of community, identity and
gists more than for ethnographically inclined culture. It had as its object(s) of study the
scholars from other disciplines, the dominant connections and relations within a system by,
model of fieldwork has been a ‘celebrated literally, following the people, the thing, the
and mystified notion of “being there”’ (Han- symbol, and other conceptual entities.
nerz 2003: 202). However, as Hannerz (2003) has observed,
Several voices from within the discipline even multi-sited studies are not necessarily
have challenged the hegemony of ethno- served by repeating the participant observa-
graphy as participant observation. Notably, tion mantra of ‘being there … and there …
6
DIIS WORKING PAPER 2015:01
and there’ for an extended period of time. He ity of being an insider and outsider at the
argues that since multi-local studies typically same time.
focus on topics that are trans-local – in Han- Despite these and many other critical
nerz’ case foreign news correspondents – it is voices, anthropologists have tended to hold
less a question for the ethnographer of con- on to the primacy of participant observation
fining herself to singular localities where she and hence of ‘being there’. This, according to
should know everybody. Rather the task is to Feldman (2011a) and others, is an expression
get into the dispersed networks; being pres- of the empiricism that permeates the disci-
ent in particular sites at particular moments, pline. He argues that ‘anthropology has not
such as conferences, festivals or meetings; made a decisive epistemological distinction
and making extensive use of a range of doc- between ethnography and participant obser-
uments. The trans-local object under study vation because it has not clearly distinguished
could also be an abstraction, such as the state, between ‘connections’ and ‘relations’ as meth-
in which case the ethnographer cannot re- odological constructs’ (Feldman 2011a: 378).
duce the object of study to the object of ob- In Feldman’s interpretation, connections and
servation (Trouillot 2001). relations differ qualitatively: whereas connec-
Another difference from the ‘being there’ tions implies the anthropologist mapping the
model of fieldwork regards the perception links between geographically separated peo-
of ‘local’ interlocutors as specialists in their ple and sites in the way suggested by Marcus,
own culture. The global condition means relations imply the study of abstract, arbitrary
that there are no true natives; as Ferguson and mediated contact (Feldman 2011a). Here,
(1999) notes, nobody ‘can claim to under- the mediating agents can be money/exchange
stand it all or even take it all in’ (Ferguson value, policy discourses, social norms, objec-
1999: 208). Since also the informants find tifying techniques of governance etc.
the world opaque, we have to understand As we will return to below, Feldman uses
both the ’structure of knowledge’ and the the notion of ‘apparatus’ to conceptualize
’nature and social organization of ignorance this complex of mediating agents. He em-
and misunderstandings’ (Hannerz 2003: phasizes the abstract, non-local nature of
210). Further, the intensified processes of the apparatus and hence the inadequacy of
globalization have also challenged ‘place’ as an empiricist ethnography based on direct,
a bounded methodological construct (Appa- face-to-face engagement. Instead, he sug-
durai 1995), and hence the nature of the par- gests a ‘non-local ethnography’ – performed
adigmatic anthropological ‘field’. Grappling by means of a pragmatic mixture of methods
with the changing field of anthropological such as archival studies, media analysis, ob-
fieldwork, Gupta and Ferguson (1997: 26) servation and interviews – claiming to main-
suggested that ethnographers employ ‘[c] tain the essence of participant observation in
reatively eclectic methodological strategies’, terms of its two key qualities: first, the ‘value
which include close observation of everyday of displacement’, the removal from the famil-
life, as well as more macro-level textual ap- iar, as a way of interrogating or problematiz-
proaches (see Li 2007 for an example). This, ing the hegemony of the taken-for-granted
again, releases the ethnographer from ‘being of the researcher; and second, the possibility
there,’ while activating through critical re- of knowing the role of contingency by seeing
flection the anthropologist’s special sensibil- how chance, contestation, alliance, petty mal-
7
DIIS WORKING PAPER 2015:01
ice, etc. may produce change or continuity at oriented approaches, as well as the Actor-
critical moments (Feldman 2011: 46). Network Theory (ANT) – not least via the
work of Latour, who is greatly inspired by
anthropology. These approaches have fore-
grounded ethnographic approaches as a way
ETHNOGRAPHY AND of getting closer to practice and the ‘hetero-
THE POLITICAL geneous realities that enter into the fabrica-
tion of some state of affairs’ (Latour 2005:
Roughly speaking, ethnographic approaches 142; Bueger 2013). In this section we will
to the study of the political have been limited briefly point to the somewhat parallel devel-
and marginal in social and political science opments of political anthropology and politi-
for decades. On the one hand, scholars in cal ethnography, as well as interpretative poli-
international relations and political science cy analysis. These approaches focus on policy,
– the staatswissenschaft so enmeshed with the rather than politics in general. They are en-
development of the modern state – have ergized by their opposition to mainstream
focused on formal political institutions, even positivist policy analysis and have gathered
though there has been some recent interest momentum in the 2000s as a forum for eth-
in the use of ethnographic approaches nographic and other forms of interpretative
in these disciplines.4 On the other hand, approaches. The three fields all contribute to
anthropology became intimately identified what we see as the emerging field of global
with ethnography in the 20th century – but political ethnography.
the discipline’s relationship to the political has
for decades been limited by various impasses,
as we will show below. In parts of sociology,
in particular due to the legacy of the Chicago Political anthropology
School’s use of participant observation, Whereas anthropology was always about
scholars have held on to ethnographic politics in some sense, the discipline has
approaches. But the sociological engagement persistently had difficulties in locating and
with ethnography in the study of politics and bounding the political (Spencer 2007).
the political has been minimal (Joseph and Indeed, when the sub-discipline of political
Ayuero 2007). anthropology developed in the 1940s and
Since the 1990s, however, things have 1950s, one of the main insights was that
changed somewhat. Anthropology has seen the political cannot be neatly separated
an invigorated political anthropology devel- from cultural and social spheres. This was
op, whereas in political science, IR, sociology particularly so in the ‘state-less’ societies that
and other disciplines a growing, though still anthropologists studied at the time. But also in
marginal, interest in political ethnography contexts of ‘modern states’, anthropologists
has (re-)emerged. Contributing to this has have systematically pointed out how informal
been the influence of pragmatic and practice- networks and institutions permeate formal
‘political’ ones. The lack of definition and
boundaries of politics has been a point of
4
Early political science did exhibit an interest in the close
encounters with politicians’ daily life, as argued by Schatz
criticism from political science scholars (e.g.
2009 Easton 1959), but anthropologists have
8
DIIS WORKING PAPER 2015:01
emphasized sensitivity to the pervasiveness in the 1990s, the contours of a new politi-
of power and politics and the constant cal anthropology emerged. Thus, anthropol-
grappling with its location as a strength of ogist engaged participant observation inside
the discipline and a fertile point of reflection organizations and bureaucracies looking at
(Vincent 2002; Biehl and McKay 2012). rituals, symbols, identities and other issues
As Vincent (2002: 2) states, the anthropol- among employees in state bureaucracies, the
ogy of politics is intimately related to eth- EU, NGOs, and private companies (Geertz
nography, ‘the anthropologists’ pride and joy, 1980; Abélès 1990, 1992; Herzfeld 1992;
the discipline’s life-blood on which everything Wright 1994; Riles 2001, to mention a few).
else in their craft depends.’ Classical politi- A particular thriving field has been the, by
cal anthropology incorporated the colonial now institutionalized, ‘anthropology of the
administrative practice of ethnography (Sa- state’.6 Whereas Radcliffe-Brown [1940] fa-
lemink 2003), making ‘thick descriptions’ of mously discarded the state as an object of
types of political organization and leadership, anthropological study – considering it a fet-
ways of dealing with conflict in colonial so- ishized and mystified object, and as an ide-
cieties without formal institutions, brokerage ological construct without agency and hence
etc. But in general, anthropologist juxtaposed without existence in the phenomenological
these thick descriptions to thin, a-historical world – a new generation of anthropologists
and homogenizing representations of mod- asks how the state as an idea comes into being
ern and capitalist institutions, including (mar- and with what effects. Inspired by sociologist
kets, bureaucracies and) the state, which, by Abrams (1988) they see the state as both an
the mid-twentieth century, had become the idea and a loose, trans-local system of institu-
dominant horizon for political authority and tions, policies and practices.
imagination across the world (Hansen and Therefore, to study the state in its his-
Stepputat 2006). torical and geographical specificity, anthro-
Paradoxically, the established sub-discipline pologists seek to disaggregate the state into
of political anthropology faded away as an- practices and processes to identify sites for
thropology became ‘increasingly political, ethnographic exploration (for instance meet-
even politicized’ from the 1980s onwards ings, encounters with state employees, acts
when post-structuralist, feminist and post- of violence, everyday policing). The aim is
colonial approaches invaded anthropology to understand how ideas, imageries, effects,
(Thomassen 2012:263).5 An intense interest and affect of the state are forged through
in discourse and representation was comple- representation and performance (Trouillot
mented by Gramscian themes of hegemony 2001; Hansen and Stepputat 2001; Sharma
and resistance, not least inspired by the sub- and Gupta 2006; Stoler 2007). Importantly,
altern studies school. Western anthropology anthropologists explore the state in a trans-
moved closer to home, and as ethnograph- national perspective, analyzing for example
ic studies of modern institutions multiplied how the development of modern states in
5 6
Also important was the influence of world-systems theory E.g. Gupta 1995; Nugent 1997; Trouillot 2001; Hansen and
(Wolf 1982; Mintz 1985) and more historically informed anal- Stepputat 2001, 2005; Nuijten 2003; Das and Poole 2004, Co-
yses of power and politics, in particular in relation to colonial maroff and Comaroff 2006; Krohn-Hansen and Nustad 2005;
societies Sharma and Gupta 2006
9
DIIS WORKING PAPER 2015:01
Europe was related to the ways in which co- political scientists and political sociologists
lonial states developed, or how templates and – with few exceptions prefer surveys,
discourses of governance circulate globally formal modelling, statistical approaches and
(Sharma and Gupta 2006). secondary sources rather than ethnography
The latter has been a consistent object of for generating their data (Joseph and Auyero
study in another, overlapping subfield of an- 2007:1, 2).9
thropology, the anthropology of policy. As Ten years later, a still marginal but grow-
we will consider more in detail in the section ing community of scholars in political so-
on interpretative policy studies, anthropologi- ciology, political science, and international
cal policy studies have particularly focused on relations, grapples with ethnographic ap-
the complex of policies and politics associat- proaches in their research (e.g. Autesserre
ed with the spread of neo-liberalism, such as 2010; 2014; Vrasti 2008; Pouliot 2008; Weav-
good governance, empowerment and parti- er 2008; Epstein 2008; Schatz 2009; Ran-
cipation, democratic reform, and new public catore 2010; Pachirat 2011; Aronoff and
management.7 As in the case of the anthro- Kubik 2013). As Pachirat10 has described
pology of the state, these studies are based on it, ‘ethnography [in the big family of re-
what is usually considered the ‘main strength search methods in political science] is clearly
of anthropology: examining global processes the youngest, somewhat spoiled, attention-
by studying how these are manifest in every- seeking child, always poking fun at and an-
day practices’ (Krohn-Hansen and Nustad noying her more disciplined, goal-oriented
2005: 7). and outwardly-successful older siblings.’ In
particular in IR and political science, eth-
nographers have to confront the lack of un-
derstanding from scholars who are unfa-
Political ethnography miliar with ethnography, but also what they
Writing from the discipline of sociology, experience as anthropologists’ attempts at
Joseph and Auyero (2007) noted in the mid- ‘policing’ ethnography (Yanow 2009b) – de-
2000s that ethnography was experiencing a spite the fact that the origins of ethnogra-
revival within the discipline. However, they phy lie in administrative (colonial) practices,
also noted a ‘double absence’: on the one rather than in the academic discipline of an-
hand, ‘politics and its main protagonists (state thropology (Salemink 2003).
officials, politicians, and activists) remain In contrast to the traditional notion of an-
largely un(der)studied by ethnography’s thropological ethnography, Yanow (2009a)
mainstream’;8 on the other hand, those who describes political ethnography as 1) mul-
study politics as part of their profession – ti-sited, not bounded physically, but rather
7
E.g. Ferguson 1994; Yanow 1996; Shore and Wright 1997;
9
Keck and Sikkink 1998; Ferguson and Gupta 2002; Paley 2002; There are notable exceptions to this: Led by Apter (1973),
Mosse 2005; Tsing 2005; Ferguson 2006; Englund 2006; Li Scott (1987, 1990) and Fenno (1990), ethnographic methods
2007; Greenhouse 2011 of data gathering were incorporated in political science, not
8 least taken up in feminist and post-colonial writings of politics
The works that the authors cite as exceptions to this rule
(e.g. Cohn 1987). See Vrasti (2008) for a selective review of
suggest a certain disconnect between political ethnography
the ethnographic turn, and Aronoff and Kubik (2013) for ex-
and the reinvented political anthropology; there are surpris-
amples of its application.
ingly few overlaps of the works mentioned in the previous
10
section and the ones mentioned here. Quoted in Wedeen 2010: 256
10
DIIS WORKING PAPER 2015:01
letting the policy define the field, the delim- every-day political practices. Among other
itation of which you find by ’following the political situations, they draw on their eth-
policy’, and 2) using documents as an impor- nographic studies of the Oslo Peace pro-
tant source of data. Interviewing is not eth- cess, in which they conducted partici-
nography but can be, if the interviewer seek pant observation in the Israeli Knesset and
to gain ’conceptual access to the unwritten, American Camp David and held interviews
unspoken, common sense, every-day, tacit with prime-minister Ehud Barak. They also
knowledge of the ‘Prime Ministry operating draw on studies of the Israeli Labour Party
manual’ (Yanow 2009a: 34). through regular participation in Party meet-
For political science, Schatz (2009a) has sug- ings and Standing Committee gatherings.
gested an ethnography much like the above. Their ethnography was thus carried out
It consists of two legs: immersion through among the formal(ly recognized) power bro-
participation, and an ethnographic sensibili- kers, legislators and policy makers to pro-
ty going beyond participation to extract the duce new, grounded understandings of na-
‘meanings that the people under study attrib- tionalism, democracy and civil society – not
ute to their social and political reality’ (Schatz from the bottom up or among the policy tar-
2009a: 5). Like Gupta and Ferguson, he sug- gets, but from the top (down).
gests that ethnography’s main method, partic- Despite these contributions, the use of
ipant observation, is coupled with analysis of ethnography in political science has tended to
‘human artifacts (texts, cultural products, and stay at the margins (Schatz 2009a: 3). In IR,
so on)’ (Schatz 2009a: 6), which can reveal the critics have pointed to the mal-appropriation
meaning that people attribute to their world, of ethnographic methods as selective, instru-
i.e. the ethnographic sensibility. This sensibil- mental and timid, and as neglecting important
ity, Schatz argues, transcends the divide be- reflexivity on the inherent ethical dilemmas
tween fieldwork and deskwork, observer and posed in conducting ethnographic research
observed, in an ‘epistemological commitment’ (see Vrasti 2008). Others hold more pragmat-
that goes beyond particular methods. ic views. Rancatore (2010) contests Vrasti by
Returning to the issue of access to elite arguing that the combination of ethnography
sites, ‘studying up’, Schatz adopts a quite prag- and IR does not overlook ethical dilemmas.
matic approach: it is not about being inside or Rather, these pose potentially useful find-
outside, but rather about striving for ‘the near- ings in themselves. We would hastily add to
est possible vantage point to study a given prob- this the point that such ‘findings’ are of lit-
lem (…) To the extent that she falls short [of tle value, if they are not brought out reflex-
gaining access to the ideal location] she pro- ively, something Vrasti might have retorted,
vides reasons why’ (Schatz 2009b: 307, orig- too. Schatz suggests that ‘ethnography helps
inal emphasis). Whereas we would add that ensure an empirically sound, theoretically vi-
conducting ethnographic fieldwork close to brant, epistemologically innovative and nor-
the conventional center of power does not matively grounded study of politics’ (Schatz
provide better data for analyzing politics, it is 2009a: 5). However, he combines with this
still a valid, if under-studied, perspective of inductive and interpretative approach, the
its articulation. propensity (some would argue) in politi-
Bridging anthropology and political sci- cal science of adjudicating, arguably positiv-
ence, Aronoff and Kubik (2013) examine ist, truth claims (Schatz 2009a: passim). Such
11
DIIS WORKING PAPER 2015:01
truth claims, we would argue, is not an a priori application of Weberian theories of authori-
task of ethnography. ty and bureaucracy in many policy studies; of
In this sense, we believe that (the method- the instrumental-rational model of the pol-
ology of) anthropology and political science icy process; and in particular the dichotomy
still tend to differ; we follow Nader (2011), between the political agenda-setting of policy
who considers ethnography a theory of de- making, and the a-political, administrative pol-
scription, in that it is contemporary and sit- icy implementation (Yanow 2015).
uated. Thus, echoing Wolf (1969), ethnogra- Thus, policy is not just a legal instrument
phy is a subjective kind of knowledge pro- implemented through objectively defined
duction that reflects the society of which it solutions to govern individuals or groups;
is a part – in much the same way that glo- even if they are often conceived and repre-
balization is not just an object of study, but sented as something neutral and rational. In-
also a subjective experience for both observ- stead, policies are dynamic documents with
er and observed. The normative bias held in almost agentive power in that they produce
the undertaking of putting forth truth claims and alter the normative and political process-
through ethnographic enquiry is thus inevi- es around them. To be sure, policies, by vir-
table. Indeed, science can never be politically tue of formulation and implementation, pos-
neutral (Nader 2011:217). But empirical ob- sess the authority and influence to (re-)define
servation should still drive interpretation. ‘target groups’, ‘problems’ and ‘solutions’ and
thus actively create renditions of the world,
rather than merely reflecting it. This ‘classifi-
cation of people and problems’ through pol-
Interpretative Policy Analysis (IPA) icy formulation (Wedel et al. 2005: 37) not
Today, policies that seek to regulate (or only transcends straightforward implementa-
deregulate) society and economy are found tion of objective solutions; policies are also
everywhere. From the 1960s onwards, interpreted by the diverse receivers. Once for-
policies have spread as an instrument of mulated, policies gives rise to a wealth of in-
change and have now become ‘truly global’ terpretive meanings and effects – intended as
(Shore and Wright 2011: 3), both in terms well as unintended. In addition, policies can
of their multiplication and their scope. Far be transplanted into other contexts, where
from relating to governmental actors only, the original purpose and intended effects are
policy decisions and practices today are often once again up for renegotiation (Shore and
diffuse and involve a range of inter- and non- Wright 2011: 3).
governmental, as well as private actors. As Yanow (2015) observes, the tradition of
While an ever-expanding field of policy engaging in participant observation in stud-
studies has developed over the same period, ies of bureaucracy in the 1940s and ‘50s dis-
IPA and what Shore and Wright (1997) call appeared when policy studies emerged in
the ‘anthropology of policy’ has emerged as a the 1960s. Thus, it is only recently that criti-
critical current in opposition to mainstream, cal policy studies have adopted ethnographic
positivist policy analysis. Starting with Lip- approaches, not only for studying policy ef-
sky’s ethnographic analysis of ‘street level fects and involvement ‘on the ground’, but
bureaucracy’ (1978, 1980) this current has also for studying up; or for ‘studying across’
developed a sustained critique of the naïve policy networks as Shore (2011) suggests.
12
DIIS WORKING PAPER 2015:01
With often neo-liberal approaches to policy- global civil society, private companies,
making, such networks have become (even) transnational entities, and ‘non-state’ actors
more diffuse and extended, and ‘following the in their analysis.
policy,’ tracing sites of agenda-setting, deci- Thus, the systems of global policy and gov-
sion-making, translation, silencing, reframing ernance that we would like to study ethno-
etc. across space and time, is a challenge for graphically are dense and multilayered, char-
the practice of ethnography (Yanow 2015). acterized by very complex dynamics between
While IPA clearly has its roots in domestic a multitude of actors who are involved in the
policy analysis, trans-nationalization and glo- attempt to generate, institutionalize and co-
balization is leaving their marks on IPA; new ordinate global policies. In this section, we
fields of analysis, such as development coop- look at different ways in which scholars have
eration (e.g. Mosse 2005; Li 2007) are making coped with the challenge of studying globali-
their way into IPA debates. Nevertheless, the zation and global policies ethnographically.
currently emerging ethnographic policy ana- We will not add to the numerous attempts
lyses within IR tend not to take the previous at defining globalization and the global con-
theoretical and methodological developments dition – in random order of appearance,
in IPA and domestic policy analysis into ac- we could mention the triumph of a capital-
count. EU policies are obviously a field in ist world (Wallerstein 1974); (neo-)coloniza-
which the domestic-international disconnect tion (Khor 2000); the weakening of the state
is being overcome (Yanow 2015), but there is (Beck 2000); the compression of time and
still much scope for taking IPA developments space (Harvey 1998); a contest of cultural
into a global political ethnography. processes between sameness and difference
(Appadurai 1990); and the intensification of
social relations linking distant localities (Gid-
dens 1990) – but delimit ourselves to empha-
TOWARDS A GLOBAL POLITICAL sizing that definitions of the global condition
ETHNOGRAPHY are submerged in political and normative dis-
courses.
As mentioned above, globalization in the The following paragraphs will consider dif-
1990s caught the scholarly attention and ferent aspects of global political ethnography,
achieved a sort of ‘atmospheric hegemony’ such as the question of (global/local) scale,
(Cunningham 1999: 538). This represented a the construction of field and the definition
serious challenge to the conventional notions of sites for ethnographic engagement, and
of ethnography, the anthropological ‘field’, more conceptual approaches to global politi-
and to notions of culture and society as neatly cal ethnography.
bounded and separate entities. Anthropology
saw the decline of what Gupta and Ferguson
(1997) have called the ‘stereotypical “among-
the-so-and-so” mold’ of fieldwork (Gupta The global and the local
and Ferguson 1997: 2). Similarly, political Whereas globalization has often been
science and IR scholars who studied issues associated with de-territorialization (e.g.
of global governance had to look beyond Deleuze and Guatarri 1980), ethnographers
the usual state actors and increasingly include have insisted that globalization is grounded
13
DIIS WORKING PAPER 2015:01
and instantiated in locations. This means that jects and agents. As Sassen (2004) suggests,
spaces are constantly being re-territorialized concrete localities are ‘multiscalar’. Thus, we
(Gupta and Ferguson 1992). Thus Burawoy, have to move beyond simple assumptions
who has conceptualized a ‘global ethnography’ of local versus global in the study of global
(2000), states that ‘[o]nly in the locality – the processes and their relationships. Otherwise
ethnographer’s hearth – can one study [the] we risk failing to acknowledge not only how
concrete effects of globalization’ (2001: 149). ‘local’ processes effectively inform or shape
Tsuda et al (2014), taking issue with Marcus’ the ‘global’ framework, but also how the con-
idea of multi-sited fieldwork as a necessary ventional global of international corporate
approach to studying globalization, claim that offices, ministries and convention centers are
globalization can be captured through in- also local but on another scale. Sometimes the
depth fieldwork in a single location. Likewise, entities are fused into a whole, which might
Gusterson11 suggests ‘tilting the field’ by hide important specificities of both; some-
studying the totality from the vantage point times they are separated into oppositions ne-
of one particular site and to trace connections glecting their mutuality; yet other times, focus
from there to the wider field. is on their relational dynamics risking a priori
Nevertheless, as ethnography is ‘necessar- assumptions of connectedness.
ily situated experience,’ there is a risk of ob- Burawoy’s own technique to avoid objec-
jectifying the global as something ‘external tifying the global consists in a triple strate-
and inevitable’ (Burawoy 2000: 29). In this as- gy that employs three defining features of
pect, Burawoy joins many others who have globalization: ‘connection’, ‘imagination’ and
warned against constructing a global/local di- ‘force’.13 ‘Connections’ and ‘imaginations’
chotomy.12 As Cunningham (1999) and Turn- refer to the flow of people, things and ideas,
er (1999) have suggested, the ‘global’ is some- while ’forces’ refer to their socio-political or-
times constructed as a context against which ganization across time and space. Firstly, he
to forge a particular (threatened) local iden- considers global forces as constituted at a dis-
tity. Also Tsing (2002) has encouraged an- tance in order to study how they are ’resist-
thropologists to stop distinguishing between ed, avoided, and negotiated’ (Burawoy 2000:
the otherwise commonsense global forces 29). Secondly, he considers global forces as
and local places; they risk, she argues, draw- produced socially, through connections, in order
ing themselves into globalist fantasies ‘by ob- to enable the study of them in and between
scuring the ways that the cultural processes sites. Thirdly, he considers global forces as
of all “place” making and all “force” making something imaginatively constructed in order
are both local and global’ (Tsing 2002: 477). to counter the (post-modern) fragmentation
Like Turner (1999) and Tsing (2005) found in of the world (Burawoy 2000: 29).
the rainforest, the global condition of density Acknowledging the fragmentation of the
and layering conflates different scales of pro- world, the fragmented nature of knowledge,
and the ‘non-existing totalities’ (of society,
culture, etc.), Biehl and McKay, in an almost
11
2005, referenced in Shore and Wright 2011
12
Robertson (1995) sought to transcend the global/local di-
chotomy by coining the notion of ‘glocalization’ but, as always, 13
He borrows the notions of connection and imagination
such ideas of hybridity are prone to criticism qua the assump- from the Chicago School, while the notion of force comes
tion of a prior separation of the global and the local. from the Manchester School.
14
DIIS WORKING PAPER 2015:01
post-modern fashion, urge ethnographers, policy processes. They use the ‘social network
via Geertz, to ‘embrace the splinters’ (Biehl analysis’ which is both theoretical and meth-
and McKay 2012: 1223). Here, the splinters odological in privileging social relations over
constitute the ‘creative world-making’ that individual actors, and tracing sites of their
subjects engage in, and which the ethnogra- articulation and interaction. Mapping social
pher may study, by thinking through the networks provides a ‘snapshot of the work-
splinters, at once connecting them to social ings of transnational policy processes’ and
theory and abandoning totalities (ibid.: 1223- shows how different ‘levels’ (local, region-
24). For the authors, the goal is not to reveal al, national and international) are connected
critical power politics underneath the messi- and how such processes involve governmen-
ness of everyday life, but rather to give ‘crea- tal, non-governmental and corporate actors
tive form to people’s art of living’ (Biehl and (Wedel et al.: 40). Hence, taken-for-granted
McKay 2012: 1223); we believe, however, that entities and oppositions of the local and the
both are worthwhile ethnographic undertak- global, or private and state are believed a priori
ings. Nevertheless, we may regard the mak- to be mixed. But more than a snapshot, such
ing of global policies as attempts to embrace analyses can also offer ideas of how relations
the splinters and create connections in a frag- develop over time by, for example, giving at-
mented world. Likewise, the construction of tention to careers and trajectories of individ-
the ‘field’ that we will study is also a way of uals and organizations that move between
shaping or reinforcing connections across the different positions and roles in the policy net-
world. works. Here Wedel’s studies of Western aid to
In her book, ‘Friction’, Tsing (2005) devel- Eastern Europe (2001) and of the influence
ops an ethnographic approach to studying of a small group of neo-cons on US foreign
global connections as a way of studying ‘the policies (2004) are exemplary.
work of the universal’. She focuses on ‘zones In their overview of anthropological stud-
of cultural friction’, understood as intercon- ies of public policy, Wedel et al. use the no-
necting zones of awkward, unstable, unequal tion of level in a relatively un-problematized
and creative engagement across differences, way. However, spurred by the globaliza-
where new arrangements of power and cul- tion of the 1990s, level, as well as scale, has
ture may emerge (Tsing 2005: xi, 4-5). Look- been the subject of extended discussions in
ing at how neoliberal environmental policy is various disciplines. Not least in human ge-
enacted in different sites in the global South, ography, the ‘scalar ontology’ of a vertical
Tsing engages herself as a scholar at academ- local-subnational-national-regional-glob-
ic conferences, as a nature lover hiking with al hierarchy has been criticized, and a ‘flat’
fellow nature lovers, and as a passionate an- or ‘site’ ontology suggested to take its place
thropologist among Indonesian communities, (c.f. Marston et al 2005; Jonas 2006), since
activists and students. The result ‘may not be social and political processes ‘do not exist
a classical ethnography’, as she writes, ‘but it at specific spatial scales, levels or tiers’ (Hay
can be deeply ethnographic in the sense of 2014: 36). In regard to the analysis of global
drawing from the learning experiences of the policies, it is hard to see how scales or lev-
ethnographer’ (Tsing 2005: xi). els can be entirely avoided, however. Legisla-
Wedel et al. (2005) suggest a different way tion, jurisdictions, identities and many forms
of analyzing connections and relations of of material resources are bound to differ-
15
DIIS WORKING PAPER 2015:01
ent scales, and scales and levels are also in- tween different fields – providing the ‘last
herently part of the strategic and everyday judgement’ (Bourdieu 1999: 67) – we can-
engagements of policy makers (and -break- not assume that the state(s) have this role in
ers) (Jonas 2006; Hay 2014). Thus, instead contemporary global policy regimes. Further-
of taking as a point of departure the com- more, as we will see below, the formation of
mon-sense ideas of local, regional, nation- global policy fields may involve the merger or
al, and international arenas as hierarchical coordination of pre-existing fields, such as
and concentric arrangements of progres- security and development or trans-national-
sively larger scale, or alternatively to discard ized migration policies. These caveats aside,
scale altogether, we should rather the poli- the field makes good sense as a methodolog-
tics of scale in policy making and turn scale ical concept that can help develop the ethno-
and scale-making into an object of analysis graphic exploration of global policies.
(Tsing 2000; Gupta and Ferguson 2002). Shore and Wright suggest defining the field
as ‘not a particular people or organisation’ –
far less a reified policy itself – but as a ‘so-
cial and political space articulated through
Field, apparatus and assemblage relations of power and systems of govern-
Encompassing the social world of actors ance’ (Shore and Wright 1997: 14; 2011: 11).
operating at different scales, Bourdieu’s This creates ‘links between agents, institu-
concept of field seems relevant to use in tions, technologies and discourses and brings
an ethnographic analysis of global policies. all these diverse elements into alignment that
The concept encompasses Martin’s (2003) makes it analytically productive’ (Shore and
three senses of field – as a topological Wright 2011: 11). In furtherance of this sug-
space of positions, a field of relational gestion, Shore offers perspectives on how to
forces, and a battlefield of contestation – position the ethnographer, namely by choos-
with an emphasis on the latter. Empirically, ing within the field of all possible actors and
Bourdieu is interested in determining what activities and institutions of relevance, specif-
the field is about; where its limits lie; what ic sites in which to study policy-related prac-
forms of capital operate in the field; and, tices and processes (Shore 2011: 28).
most importantly, what legitimate principles, Defining the field, as well as the strategic
or rules of the game, define the social sites for studying global policies, cannot al-
world of the field and the participation and ways be done in advance of the study. In an
exchanges herein (Bourdieu 1991; Bourdieu extensive, collective research project, a team
and Wacquant 1992). Here, the rules of of researchers developed a multi-sited ethno-
the game could refer to the practices that graphic analysis of the global reinsurance in-
define policies as an overarching way of re- dustry by following traders in various com-
forming the social world, as well as to the panies and trading hubs across the world, as
ways in which a specific policy develops to these reacted simultaneously on certain disas-
frame and problematize particular issues, trous events (Jarzabkowski et al 2014). How-
thus contributing to the symbolic power of ever, the researchers spent a long time ob-
dominant groups (Bourdieu 1991). serving, analyzing and sharing interpretations
However, whereas Bourdieu regards the before they could define the ‘strategic sites’,
state as an instance of defining relations be- their ‘units of observation’ and their ‘eth-
16
DIIS WORKING PAPER 2015:01
nographic object’. As it were, they reached problem, and the desirability of public sector
a definition of their ethnographic object as reform is shared.
being a ‘global risk trading practice’, a prac- This kind of shared understanding and net-
tice which was both local and an intercon- worked governance which, in Neuman and
nected global practice (ibid.). Sending’s (2010) words, ‘adds up to govern-
The understanding and mapping of the ment over governments’ (Neuman and Send-
policy field can be a serious analytical chal- ing 2010:164) is a crucial subject in Foucault-
lenge in itself, as illustrated by Pal and Ireland ian analyses of neoliberal policy to which we
(2009) in their mapping of the global policy will now turn.
network around public sector reform. These While Pal and Ireland’s rather empiri-
policies are focused on the reform and (good) cal mapping of a policy network points to
governance of state administrations, rather a range of institutional actors, processes
than global issues as such. But the network’s and events where ethnographic explorations
reach and policy agendas are global in scope. would make sense, Feldman (2011; 2013) is
We may further argue that these policies, in interested in the kind of relations that can-
particular after 9/11, have become part of a not be traced and mapped directly as point-
nexus between international security and de- to-point connections spanning the world. He
velopment policies, which aim at improving looks at how disparate policy processes are
regional and global security by preventing coming together to regulate illegal migration
state failure. to Europe through the production of dis-
In distinction from ‘epistemic communi- courses that enable the emergence of an ‘ap-
ties’ (Haas 1992), which are more tightly fo- paratus’ through a ‘great conversation’ across
cused around professional-scientific research time and space (Feldman 2011: 45). Here
issues and credentials, the public policy net- Feldman borrows Foucault’s concept of ap-
work that Pal and Ireland delineate comprises paratus, namely the ’ensemble consisting of
a broader range of expertise and institutions, discourses, institutions, architectural forms,
including inter-governmental organizations, regulatory decisions, and philosophical,
bilateral donors, professional associations, moral and philanthropic propositions’ (Fou-
and non-governmental organizations – even cault 1980: 194). The apparatus works from
ones that do not see themselves as taking a certain political rationale and is strategic in
part in any alliance. Despite the limited use nature. It is bureaucracy decentralized, medi-
of directly coercive means or clearly delineat- ating between the policy subject and object
ed channels of ‘policy transfer’ (Stone 2004; through the ‘expert,’ here the technocrat who
2012), the accumulated changes in governance operationalizes connections between policies.
systems over the past 25 years are dramatic To study how this ‘nonlocalisable appara-
(Pal and Ireland 2009). The loose network is tus’ emerges, functions, and fabricates what
connected through different instruments and is presented as ‘an objective target of reg-
practices, such as conditionality, training, ad- ulation’, Feldman (2011: 45-46) suggests a
vocacy, networking, research, and systems of ‘nonlocal ethnography.’ As mentioned above,
indicators, standards and norms. The agenda this is a flexible methodology that retains
is not coherent, and ideas, objectives, orien- the assets of participant observation – criti-
tations and tools differ greatly. But there is a quing the ‘hegemony of “common knowl-
common understanding of governance as a edge”’ and tracing the ‘role of contingency in
17
DIIS WORKING PAPER 2015:01
human affairs’ – without being fixed to a few cupying a ‘common field in contingent, un-
places. It is a kind of concrete, empirical anal- easy and unstable interrelationships’, the
ysis that reaches through disparate policy do- ‘space of the assembly’ (ibid.: 12). In the case
mains and locations, using methods such as of the ISO standards, their introduction may
archival research, analysis of documents and impel people who are unable to manage the
speeches, as well as participant observation in standards to circumvent the discipline of the
offices, international meetings, conferences, standards. This circumvention – or ‘irregula-
and the like (Feldman 2011: 46-47). As one tion’ effect (Stepputat 2009) – is as much a
particular strategic site of ethnographic en- part of the assemblage as the standards them-
gagement, Feldman chose an office of coor- selves. Likewise, new forms of migration
dination between European migration policy, control procedures and technologies spur
criminal law enforcement, and border con- illicit migration agents to circumvent con-
trol, to see how a shared understanding of trols, as captured in Gammeltoft-Hansen and
crisis and (in-)security emerged. Sørensen’s (2013) idea of ‘migration indus-
Moving beyond Feldman’s ethnograph- tries’ that encompass public and private, as
ic discourse analysis, Collier and Ong (2005) well as legal and illegal entities. As such, glob-
suggest the notion of global assemblage to al assemblages hold an inherent tension. As
frame a practice-near analysis of how ‘glob- Collier and Ong write, ‘global implies broadly
al phenomena,’ such as techno-science, cir- encompassing, seamless, and mobile; assem-
cuits of exchange, systems of governance, blage implies heterogeneous, contingent, un-
and regimes of ethics, work to define materi- stable, partial, and situated’ (Collier and Ong
al, collective and discursive relationships, and 2005: 12).
to raise new problems of individual and col- To describe the place in which global as-
lective existence (Collier and Ong 2005: 4). semblages are articulated and, thus, can be
These phenomena, or ‘global forms’, have studied, Collier and Ong (2005: 4) use the
a global quality in that they are not depend- term ‘site’, which they describe as a ‘political-
ent upon the ‘props’ of a culture or a soci- ly and morally charged domain.’ Compared to
ety; they are universal in the sense of being the apparatus approach that focusses on the
abstractable and mobile. They have a capac- governmental side of affairs, an assemblage
ity for de-contextualization and re-contex- approach seems to expand considerably the
tualization, and being designed to produce number of potential sites for ethnographic
‘functionally comparable results in disparate engagement.
domains’ (ibid: 10-11).14 The cases of ISO Anyway, as Marcus (2010) has observed,
standards or biometric devices for migration the kind of sites where ethnographers engage
control are good examples. themselves in studying policy are increasing-
However, these global forms interact with ly sites of specialized knowledge produc-
‘ensembles of heterogeneous elements’, oc- tion. Here, the ethnographer works alongside
others who may be patrons, colleagues, part-
14
ners, and subjects of research at the same
The idea resembles the repertoires of technologies of
government that produce an abstract space (le Febvre 1991) time (Marcus 2010: 15). This situation is in-
by rendering populations legible and formatting contextual- herent in attempts at studying global poli-
ized information to be transmitted to administrative centres,
as developed in relation to historical analyses of state forma-
cies in the making, not least as ethnographers
tion (eg. Scott 1998). might be given access to otherwise restrict-
18
DIIS WORKING PAPER 2015:01
ed-access fora in their capacity of being con- eye. Thus as we saw above, Gusterson (1997)
sultants, experts, or policy analysts (see for conducted fieldwork on nuclear power plant
example Cohn 2006; Mosse 2005; Stepputat laboratories, rather than in them because he
2012). was not able to gain access. Such biases are
ubiquitous caveats to be addressed analyti-
cally, but in studying global policy processes
this presents a no less urgent ethical dilemma
Ethical considerations in terms of management of sensitive infor-
Ethnography is a relational, subjective mation, confidentiality and representation, to
endeavour in which the researcher engages which we return below.
with actors and practices in and of a field There seem to be two over-all approaches
through various methods. Whether by to dealing with ethical issues. Firstly there is
engaging directly through participation or what we may call an institutional approach,
more indirectly by way of gaining knowledge for instance adherence to procedures and
through interviews, observations or guidelines (e.g. AAA 2009 and Chatham
textual analysis, the ethnographer becomes House Rules), signing of informed consent
involved in, and is knowledgeable of, the and approval by research boards, whether for
social harmonies and tensions of the field. ‘real-life’ ethnography (Yanow and Schwartz-
This position grants attention to ethical Shea 2008) or online research (Warner 2009,
considerations. Whereas all ethnography Weeden 2012). Secondly, and more complex,
merits close consideration of ethics, the an approach to dealing with ethical issues
ethnography of global policy brings out this is ethics as an on-going exercise through-
circumstance more urgently, since power, out research – or in the distinction offered
politics and control are often of direct or by Fujii (2012): ‘procedural ethics’ and ‘eth-
indirect import in this context – whether ics in practice’ (Fujii 2012:717, see also Guil-
studying up among policy makers, moving lemin and Gillam 2004). However, as Fujii
among those subjected to the policies, or (2012) points out about procedural or in-
both. stitutional ethics, “consent forms are of lit-
Immersing oneself in one or more sites and tle value if participants do not understand
engaging with its actors’ practices, secrets and to what they were consenting” (Fujii 2012:
inclinations points to issues of confidentiality 718). Informants might be illiterate or living
(see Palys and Lowman 2012). Further, eth- in repressive or authoritarian societies (see
nography poses a potential threat to the sen- Goduka 1990), where individual choice and
sitive information of a field and the delicacy empowerment are scarce resources. In such
of its relations, not least in sites with ‘vested contexts, consent forms may set out a rela-
interests of powerful persons of the exercise tionship between researcher and informant,
of coercion or domination’ (Lee and Renzet- in which the latter has the authority to define
ti 1990: 512), which global political ethnogra- what sensitive information is; how to handle
phy often does. The reliance in ethnography it; and how to protect the informant from
on direct or indirect interaction may, then, adverse effects of research. Consent forms
refer the ethnographer to sites that inform- thus rest on the researcher’s conceptions of
ants and gatekeepers deem sufficiently safe security, rather than the informant’s actual or
or appropriate for an ethnographic outsider’s perceived risks.
19
DIIS WORKING PAPER 2015:01
Further, Bosk and de Vries (2004) stress that out others. Presenting both formal and in-
whereas institutional review boards offer formal knowledge and practice, as ethnogra-
best practices for routine review and approv- phy does, means that the ethnographer not
al, qualitative research will never move with- only defines the field of study, but also devel-
in the confines of such models (Bosk and de ops its narrative. To triangulate data, the idea
Vries 2004: 252). A rigid insistence on ano- of member checking15 is useful, i.e. sharing
nymity and confidentiality, they argue, may notes, analyses and drafts with informants to
hamper ethnographic fieldwork of for in- ensure proper understanding and clear poten-
stance government agencies, because political tially sensitive information. But it also points
environments do not appreciate controversy to the power vested in the ethnographer.
or power play being made public (Bosk and This leads to another consideration of ‘eth-
de Vries 2004: 254). And in terms of ano- ics in practice’ in the ethnography of global
nymity, it can be virtually impossible, despite policy, namely the politics of writing and rep-
good intentions, to hide the identities of re- resentation. As Mosse (2009) points out, the
search participants, as Fujii points out, or anthropology of policy produces knowledge
there may be instances in which it is necessary and text that more often circulates among in-
for the researcher to lie to preserve confiden- formants than other ethnography. Being a
tiality (Fujii 2012: 720). Whereas institution- knowledge producer among knowledge pro-
al ethics is an important standard of practice ducers accentuates a set of reflections and
and is useful as a means of sensitising the re- challenges, which echoes the critique of the
searcher, pre-approval procedures and adher- power differentials between ethnographers
ence to ethical guidelines do not safeguard and their subjects-interlocutors that were
research subjects from the actual impacts of raised in anthropology in the 1980s. With this,
the researcher’s engagement in the field and handling the privacy of actors, the politick-
the repercussions of her knowledge produc- ing among them and secrets in and between
tion. The responsibility in action still remains sites grants consideration. Mosse describes
with the researcher beyond procedures and the challenge of ethics in the ethnography of
approvals. policy as a ‘problem’ because informants – in
Here, the second strand of ethical consid- his case policy makers and development ex-
erations is of import, namely the approach perts – do not share epistemology and pur-
to ethics as an on-going consideration dur- pose with the ethnographer. The former re-
ing research. The ethnographer is confronted lies on official knowledge, universal logic and
at all times with choices of relationships and consensus on appropriate technologies of
foci in studying and representing global poli- problem-solving; the latter lives on the mean-
cy processes and its actors. As mentioned ear- ings that escape authorised policy interpreta-
lier, in defining the field(s) of global political tion and which shows both personal interest
processes, one might argue that the ethnog- and unintended consequences of policy and
rapher inevitably constructs the field, rath- practice. The policy makers and practition-
er than locating it – both offline (Stepputat ers, he argues, thus may find ethnography a
2012: 443) and online (Forte 2004: 224). But threat to professionalism – their profession-
it entails influence over knowledge produc- alism. Likewise, Stepputat (2012) argues that
tion, where the chosen sites and relationships
constitute the building blocks, while leaving 15
We thank Dvora Yanow for the idea of member checking.
20
DIIS WORKING PAPER 2015:01
21
DIIS WORKING PAPER 2015:01
is a term for the overlapping academic/field- icy areas are necessarily flawed by lofty and
work space in contemporary ethnograph- contested notions (site, place, space, locality,
ic projects. Deeb and Marcus (2011) explain level, scale, network, apparatus, assemblage).
how they, in relation to Marc Abélès’ ethno- However, we hope that this working paper
graphic mega project on the WTO (Abélès has shown that it is worthwhile to take up the
2011), met with the Director-General of the challenge.
organization to ‘gain insights by trying to
share a perspective on the problems of and
at the WTO, which ethnography might elu-
cidate, as well as the forms that ethnograph-
ic data and knowledge might take’ (Deeb and
Marcus 2011: 53). While this encounter could
be interpreted as a sort of ‘member checking’
in a case of ethnographic ‘studying up’, Mar-
cus emphasizes that para-sites and ‘collabora-
tories’ can help giving account of what kind
of questions contemporary ethnography can
answer, what results it might be expected to
produce on the basis of what data (Marcus
2010: 14).
CONCLUSIONS
22
DIIS WORKING PAPER 2015:01
LITERATURE
23
DIIS WORKING PAPER 2015:01
24
DIIS WORKING PAPER 2015:01
25
DIIS WORKING PAPER 2015:01
26
DIIS WORKING PAPER 2015:01
27
DIIS WORKING PAPER 2015:01
28
DIIS WORKING PAPER 2015:01
29
DIIS WORKING PAPER 2015:01
30