Professional Documents
Culture Documents
Tania Li - Practices of Assemblage and Community Forest Management
Tania Li - Practices of Assemblage and Community Forest Management
Practices of assemblage
and community forest
management
Tania Murray Li
Abstract
Governmental interventions that set out to improve the world are assembled from
diverse elements discourses, institutions, forms of expertise and social groups
whose deficiencies need to be corrected, among others. In this article I advance an
analytic that focuses on practices of assemblage the on-going labour of bringing
disparate elements together and forging connections between them. I identify six
practices that are generic to any assemblage, whatever its specific contours: 1)
forging alignments, 2) rendering technical, 3) authorizing knowledge, 4) managing
failures, 5) anti-politics, and 6) reassembling. I demonstrate the power of this
analytic through an extended study of community forest management. This is an
assemblage that brings together an array of agents (villagers, labourers, entrepreneurs, officials, activists, aid donors, scientists) and objectives (profit, pay,
livelihoods, control, property, efficiency, sustainability, conservation). Its very
unwieldiness helps to sharpen analysis of how such an assemblage is, in fact,
assembled, and how it has been sustained for more than thirty years, absorbing
hundreds of millions of dollars in programmes from the west coast of Canada to the
eastern islands of Indonesia. I do not attempt to adjudicate the rights and wrongs of
this assemblage. Rather, I deploy an analytic of assemblage to explore the practices
that fill the gap between the will to govern and the refractory processes that make
government so difficult.
Keywords: governmentality; assemblage; community; forestry; politics; Indonesia.
264
Introduction
Dispositif, apparatus, technology of government, regime of practices, assemblage theorists have used various terms to refer to the way in which
heterogeneous elements including discourses, institutions, architectural
forms, regulatory decisions, laws, administrative measures, scientific statements, philosophical, moral and philanthropic propositions are assembled to
address an urgent need and invested with strategic purpose (Foucault 1980:
194).1 That purpose is the will to govern or, more specifically, the will to
improve: the attempt to direct conduct and intervene in social processes to
produce desired outcomes and avert undesired ones.2
In this article I argue for an analytic of assemblage to highlight some aspects
of government thus far neglected. I illustrate with an extended study of
community forest management. I propose an analytic of assemblage for three
reasons.
First, assemblage links directly to a practice, to assemble. The other terms,
in contrast, highlight the resultant formation (the apparatus, the regime, the
technology) which appears settled, potentially even complete.3 Assemblage
flags agency, the hard work required to draw heterogeneous elements together,
forge connections between them and sustain these connections in the face of
tension. It invites analysis of how the elements of an assemblage might or
might not be made to cohere.
Although the contingency of assemblage is often mentioned in studies of
governmentality its invocation is rather rhetorical. Nikolas Rose, for example,
writes that government is not a process in which rule extends itself
unproblematically across a territory, but a matter of fragile relays, contested
locales and fissiparous affiliations and hence inherently risky (Rose 1999: 51).
Yet he does not follow through to ask: if there is contestation, how is it
manifested? If relays are fragile, what are the practices that attempt to secure
them? If affiliations are fissiparous, how does this fact shape the resulting
formations? If there are risks, who apprehends these risks, and how do they
factor risks into their calculations? The answers to how questions such as
these deserve a more central place in studies of government than is currently
accorded. Yet I agree with Rose and others that the analytic of governmentality
is usefully distinguished from the ethnographic study of how governmental
interventions play out at a particular time and place.4 Examining practices of
assemblage enables an expansion of the analytic of governmentality without
loss of focus.
Second, thinking about assemblage as the continuous work of pulling
disparate elements together highlights a particular set of practices that are
often implicit in studies of government but seldom examined in a focused
manner. The practice that tends to take centre stage is problematization, as
scholars examine how problems come to be defined as problems in relation to
particular schemes of thought, diagnoses of deficiency and promises of
improvement. Yet there are other practices that are crucial to assemblage, and
265
equally generic, of which I have provisionally identified six. Here I simply list
them, deferring elaboration to the case study that follows.
1) Forging alignments : the work of linking together the objectives of the
various parties to an assemblage, both those who aspire to govern conduct and
those whose conduct is to be conducted.5 2) Rendering technical: extracting
from the messiness of the social world, with all the processes that run though
it, a set of relations that can be formulated as a diagram in which problem (a)
plus intervention (b) will produce (c), a beneficial result.6 3) Authorizing
knowledge : specifying the requisite body of knowledge; confirming enabling
assumptions; containing critiques. 4) Managing failures and contradictions :
presenting failure as the outcome of rectifiable deficiencies; smoothing out
contradictions so that they seem superficial rather than fundamental; devising
compromises. 5) Anti-politics : reposing political questions as matters of
technique;7 closing down debate about how and what to govern and the
distributive effects of particular arrangements by reference to expertise;
encouraging citizens to engage in debate while limiting the agenda.8
6) Reassembling : grafting on new elements and reworking old ones; deploying
existing discourses to new ends; transposing the meanings of key terms.
The third feature of assemblage is its potential to finesse questions of agency
by recognizing the situated subjects who do the work of pulling together
disparate elements without attributing to them a master-mind or a totalizing
plan.9 Although there are occasions when a revolutionary movement or
visionary leader announces a grand scheme for the total transformation of
society the kind of plan James Scott describes as high modern more
often interventions are assembled from an existing repertoire, a matter of
habit, accretion, and bricolage.10 Particular programmes come complete with
elaborate rationalizations but the elements from which they are drawn have no
singular rationality and no essence.
In addition to contingency, which I have already discussed, the diffusion of
agency is usefully elaborated by Deleuze and Guattari who use the term
assemblage to highlight the way in which material content (bodies, actions,
passions) and enunciations (statements, plans, laws) are linked not in linear
fashion but rhizomatically as reciprocal presuppositions and mutual insertions
play themselves out.11 The case I explore offers a striking example of how
reciprocal presupposition is implicated in the constitution of a social field fit
for government. Parties to the assemblage of community forest management
use the label ambiguously to describe material content (what social bodies do,
what they want) and to enunciate a prescription (what they should do, what
they should want). Further, Deleuze and Guattaris concept of assemblage
emphasizes both temporality and spatiality: elements are drawn together at a
particular conjuncture only to disperse or realign, and the shape shifts
according to the terrain and the angle of vision.
In the next section I introduce the assemblage of community forest
management and explain some of the features that make it especially fruitful
ground for the application of the analytic I propose. Most obviously its very
266
unwieldiness and the deep tensions that run through it highlight my central
question: how the elements of an assemblage are, in fact, assembled.
267
stakes mean that the switch from commanding to governing does not resolve
the struggle over access.
Community forest management is the assemblage that has emerged in the
space of struggle between villagers and forest bureaucracies on the forest edge.
Its central proposition is that people who live in close proximity to a forest can
manage it effectively over the long term. It begins, that is, from a proposed
solution rather than from a unified specification of a problem. The assemblage
is far from seamless. It is a product of popular demand as well as expert
diagnosis and prescription. It carries within it a will to govern that sits uneasily
with an argument that communities are capable of governing themselves. It
includes parties interested in upholding forest laws and parties who deem these
laws illegitimate. A rough map of parties to the assemblage of community
forestry and their various stakes would include the following:
. Official forestry departments seeking to retain or recover control over forest
land under their jurisdiction but claimed by forest villagers. By offering
these villagers limited benefits from the forest, officials hope to obtain their
compliance with forest rules or, in some cases, their signed recognition of
the legitimacy of the forest departments claim to disputed land. Benefits
yielded to villagers might include the right to collect non-timber forest
products; permission to plant food crops for a few years on logged over land
while the timber re-grows or payment in cash or kind for planting and
protecting stands of commercial timber.
. Conservationists who believe that community control over forests is or can
be more ecologically friendly than state-backed corporate logging and more
effective than conservation by centralized state agencies incapable of
enforcement. They draw upon theories about common property influenced
by the new institutional economics to argue that properly crafted rules
formulated by communities with expert assistance can sustain forests over
the long term.
. Agroforestry experts interdisciplinary teams of natural and social scientists
who work on combining crops with trees to demonstrate that forest
conservation and the livelihood needs of forest villagers can be met in one
felicitous package. They seek to validate indigenous agroforestry systems,
devise improvements and spread perfected models to new locations. They
hope that enhanced productivity will mitigate conflict over rights.
. Legal advocates convinced that forest villagers have legal rights as well as
moral claims to the forest. They cite customary laws, national laws and
transnational legal instruments such as the International Labour Organizations Convention 169 on Indigenous People or the Convention on
Biodiversity. They aim to secure the land rights of forest villagers from
appropriation by national regimes that continue the predatory practices of
their colonial forebears.
. Activists inspired by cultural difference for whom community forestry is an
object of passionate attachment embraced not only for its efficiency but
268
269
270
sound use and management, are involved in joint decision making and abide by
the rules that have been agreed. Therefore, it is imperative that decision making
power and responsibilities be entrusted to the primary users of natural
resources, the rural populations. There are compelling reasons to vest these
in strong community organizations and to reinforce participatory principles.16
Note the elision of what is and what should be. Resource users have a natural
interest in sustainable management, but they need to be educated and
reminded of this interest. Their agreement and compliance must be invoked,
and their organizations made strong and participatory. There is, in short, work
to be done.
271
know the needs of future generations, sustainable use thus involves not only
conserving biological diversity, fauna and flora, but also maintaining ecological
functions such as soil quality, hydrological cycles, climate and weather, river
flow and water quality (ibid.: 70). He then draws a link between natural
landscapes, livelihoods and local requirements. Sustainability also implies
maintaining supplies of natural products game, fish, fodder, fruits, nuts,
resins, dyes, basts, construction materials, fuel wood etc. which are essential
to the livelihoods of local people (ibid.: 70). He links forest dwelling with
cultural distinction and political marginality (ibid.: 71), connecting landscape,
livelihood and identity in one problem definition and one problem solution:
community forest management.
Note in Colchesters description the presumption that communities
exist for how else could they have their own institutions and a decisive
voice in planning? This assumption is elaborated in a strikingly technical
description of what forest-dependent communities want: the right to the
ownership and control of their territories; the right to self-determination; and
the right to represent themselves through their own institutions or, even more
specifically, the right to collective ownership of their communal territories.
They want legal recognition granting them inalienable freehold title to their
ancestral domains (ibid.: 72). It is, he states, denial of communal rights that
causes land ownership to be fragmented and individualized, and traditions of
custodianship and resource management undermined (ibid.: 76).
The critical omissions and erasures in this depiction of exemplary
communities, whose desires Colchester claims to know and represent, are
not hard to spot: farmers who cut trees to plant crops, timber workers, people
who have migrated to the forest edge or been forcibly resettled there, local
people who know nothing about the craft of weaving and dying cloth because
for generations they have bought their clothing in the local market, people
whose land is not and never was part of a communal territory, people who sell
their own land, people who welcome new markets for cash crops and take from
the environment not just to meet basic needs but to finance education and
access urban jobs.18 People such as these are difficult to accommodate in the
niche in nature to which villagers are assigned in Colchesters rendering of
community management. Refractory processes must also be excluded from the
diagram. In many parts of the world clearing forest is the practice that
establishes private rights to land, and land once cleared is rapidly commoditized and accumulated. Thus villagers interest in trees on a forest frontier
often has more to do with a competitive grab for land in the context of agrarian
class formation than management, still less management on a communal
basis.19 Community forest management is not equipped to address processes of
this kind so they do not figure in the narrative.
While some unruly forces are ruled out of Colchesters diagram, forces
that can be rendered technical are admitted. Colchester makes the difficulties
of implementing community management the subject of a detailed, technical
analysis focused on the troubled case of Papua New Guinea (PNG)
272
(ibid.: 86 7). In PNG collective land rights are indeed strongly protected by
law, yet communities have not conserved their forests. Instead, they have leased
their land to logging and mining companies in exchange for royalties. The
problem, Colchester argues, is the laws imprecise definition of who has rights
to negotiate land deals, leading to manipulation, bribery and debt leverage by
companies. He notes the rapid pace of social change that has made New
Guineans unrealistic about what is achievable, and led them to contemplate
futures that are not tied to the land. Tribal big men have become less
accountable and less responsive to community needs and rights than they were
under traditional circumstances, a phenomenon he calls lairdism and
recognizes to be widespread in indigenous societies (ibid: 87). The resolution
of this rather formidable set of counter-indications is to be found, Colchester
argues, in a technical intervention: replacing traditional institutions with more
democratic and accountable ones, tailored to new purposes.
As soon as an improvement is needed, experts are free to step in to rearrange
things. Colchester observes this freedom and finds it problematic. He quotes
Jack Westoby, former Director of Forestry at the FAO:
Only very much later did it dawn on the development establishment that the
very act of establishing new institutions often meant the weakening, even
the destruction of existing indigenous institutions which ought to have served as
the basis for sane and durable development: the family, the clan, the tribe, the
village, sundry mutual aid organizations, peasant associations, rural trade
unions, marketing and distribution systems and so on.
(Colchester 1994: 89)
273
274
275
but they seldom propose a radically different approach.22 Indeed, when the
alternatives appear to be reduced to community versus state, few scholars
concerned about social justice locate themselves on the state side of the divide.
For this reason, they self-censor their criticism. They are wary of supplying
ammunition arguments, data or the justification to use live bullets to
forest agencies or others who would exclude communities and coerce
conservation in forest areas.23 A few examples will illustrate the kinds of
contained critique made by social scientists who do not want their input to
unravel the assemblage.
One strand of scholarly criticism of community forest management focuses
on the nature of rural property relations. It is a debate between legal advocates
arguing that communities have distinct community-based rights, generated
autonomously by people interacting with each other and their environment,
and anthropologists who study law in practice. The anthropologists find that
peoples sense of entitlement and their frameworks for justification in property
disputes are plural and hybrid. Forest villagers sense of what is customary is
refracted through colonial and post-colonial legal regimes and the brute facts
of possession.24 The practical implication of the anthropologists critique is
that recognition of customary or community-based rights presents as many
questions as it resolves.
Just as the basis of community claims is complex and hybrid, so too with the
state: state agencies may claim to command large areas of forest but they are
not necessarily in effective possession. State assertions of eminent domain
become real to forest dwelling people only when forest guards take an
interest. In many forest areas there are long-standing accommodations that
enable people to farm in state forest while guards look the other way.25
Anthropologists and others who study these dynamics worry that de facto
compromises might be jeopardized by the new forms of visibility community
forest management programmes require, but stop short of condemning
the attempt. Containment works here at the level of empirical detail: in
a complex field of relations, community forest interventions may or may
not produce negative outcomes.
Anthropologists often critique assumptions of stability, coherence, homogeneity and localization. They argue, for example, that the community of
community management programmes:
compresses individual differences, intentions, and agendas, and presents people,
in terms of their corporate membership in communities, as somehow singular,
identifiable, and knowable entities. In assuming an unquestionable spatial
visibility and temporal viability, the notion that community can sometimes be a
specific social achievement or a particular political performance is completely
disregarded. In the novel form of essentialism, communities become the new
tribes, or even worse, the new races of development, environmental or resource
management discourses.
(Pannel and Benda-Beckmann 1998: 20)
276
277
Thus these advocates follow stricture with compromise: external initiatives can
qualify so long as they are eventually taken over and owned by the
community.
Donor agencies supporting community forest management programmes in
the global south routinely claim that their programmes are owned by a local
constituency, although such ownership is seldom achieved. Similarly activists
and non-government organizations (NGOs) claim that they merely facilitate
initiatives that are community based but this claim is contested. Sceptics argue
that NGOs sell the term community in order to access donor-funded
projects. Donors monitor the relations between NGOs and villagers to detect
signs of less than genuine participation. This puts NGOs in a bind. They
must service their client, the donor, and deliver the agreed outputs on schedule
and on budget. Their institutional survival depends upon successful projects
and further donor funds. The communities in which they work become the
target group or the beneficiaries of initiatives designed for them and
imposed, however gently. On the other hand, donors committed to concepts of
public participation and empowerment want to select NGOs with a credible
constituency, a public mobilized enough to define its own agenda. Spreading
donor largesse helps an organization to acquire and keep such a constituency,
but risks undermining its authenticity.29 The relay points in this assemblage
are fragile and held together by compromises of various kinds, as I will
demonstrate with some examples from Indonesia.
Public anger about the conduct of the forest department is widespread in
Indonesian villages within or adjacent to state-claimed forests. At one
national gathering of forest villagers convened by advocates there were
278
279
efficiency to the interaction between people and forests and to enhance the
welfare of the population. They insist that problems should be addressed not
by brutality but by enhanced technique of the kinds that experts are there to
supply.
In practice, contradiction is managed less by technique than by compromise
in its dual sense: parties to the assemblage make compromises and, in so doing,
they become implicated, their positions compromised and their critiques
contained (Li 1999). I have already examined the compromised position of
scholars, myself included, who set aside their scepticism about diagrams that
purport to render messy actualities technical and manageable in order to
protect the assemblage from attack. Compromise similarly characterizes the
position of the Sulawesi forest villager I quoted earlier who lamented the
omnipresence of forest guards. His village benefited from the work of NGOs
that helped persuade officials of their sustainable, indigenous forest management system, thus averting their eviction from a national park. After our first
meeting at his forest village, I met him again at the office of the NGO in the
provincial capital and again in the capital city where he was making
arrangements for a trip to China to represent forest communities at a
gathering of NGOs. He too had become an expert.
To point out the complex positioning of a forest villager is not to suggest
that he had become corrupt or somehow less authentic as if his only proper
place was in the village. In each of these sites he was a legitimate social actor,
positioning himself in a complex field of power that opened some avenues for
action just as it foreclosed others. Assemblages such as community forestry
cannot be resolved into neat binaries that separate power from resistance, or
progressive forces from reactionary ones. It is difficult to determine who has
been co-opted and who betrayed. Fuzziness, adjustment and compromise are
critical to holding assemblages together.
280
281
They also continue to evict people, the scandalous fate of tribal communities
caught in the wake of the massive Joint Forest Management programme in
India sponsored by the World Bank.33
Bilateral aid donors, multilateral development banks and foundations have
invested extraordinary resources in community forest management with the
twin objectives of improving communities and improving forest bureaucracies.34 The second relay is at least as difficult as the first. Donors must tread
softly, attempting to reform bureaucracies by engaging in policy dialogue,
capacity building and conducting pilot projects in which to nurture new
ideas and practices.35 Yet these pilots often end without donors securing
commitments from state forest agencies to scale up pilots into programmes,
still less to change their policies, which still include the harassment and
eviction of forest villagers. Further, endless workshops fail the test of new
audit requirements that expect donor agencies to manage by results.
Donors pursue their improving agendas within the limits set by their need
to continue to engage with the state apparatus. They can be pro good
governance and they can be suspicious of opportunistic officials seeking to
access project funds, but they need officials to sign off on memoranda of
understanding, issue visas and so on. Thus they cannot be anti-state.36 Nor can
they upset the status quo by engaging in an explicit political challenge. Baviskar
and Saberwals examination of the relationship of the Ford Foundation to
Indias ruling regime illustrates the anti-politics embedded in donor promotion of community management. In the 1960s the Foundation concluded that
involvement in Indian land reform would be futile. It was wary of accusations
of political interference and stepping too heavily on sensitive administrative
toes (Baviskar and Saberwal 2002: 18 20). In the 1980s the Foundation turned
to community management on marginal lands, a strategy that enabled it to
address poverty without tackling issues of land access or the mobilization of
rural people as workers. Indeed, the anti-politics of this strategy was explicit:
Robert Chambers, then a Ford Foundation programme officer, argued that
securing marginal resources for the poor was a politically realistic strategy
that would not be seen as threatening rural power structures (Baviskar and
Saberwal 2002: 54). From then on a focus on communities and institutions
closed down the space for an engagement with politics in terms of the class
character of the state and support for basic redistribution of land-related
assets (ibid.: 61). Rather than seeking to reform the refractory state apparatus,
Ford began to avoid it and work directly with NGOs (ibid.: 61).37 Later, it
stressed collaborations between groups with multiple perspectives and a
plurality of approaches (ibid.: 68).
The commitment to pluralism, to multi-stakeholder dialogues in which
diverse views and cultural differences are respected, is claimed by advocates of
community management as a moral charter. The same move displaces debate
about access to rural resources by focusing attention on their management.
Fords community forest management programme in Indonesia is a case in
point. It assembles forest villagers, NGOs, the forest department and others in
282
283
failure to recognize the rights and needs of forest communities. For the most
part, the officials acknowledged their collective failure to manage the forests
and stated their commitment to a better future working with communities.
In the small group sessions that followed participants were asked to share
their own experiences with community forest management, identify the
problems they faced and propose ways to overcome them. Time constraints
meant that it was not actually possible for a mixed group of officials, activists
and forest villagers from across the archipelago to share experiences, which
are hugely diverse and embedded in the complex politics of particular locales.
As a substitute, participants were asked to capture the key points by writing
with a marker on a large sheet of paper. The identification of problems and
solutions became similarly simplified. Problems were to be written down on
index cards which were then sorted by the facilitators into clusters, all
specificity reduced. What is one to make of a card that states lack of
communication between the relevant parties? What is the story behind it? The
solutions were equally banal: Increase coordination between all parties.
Participants were bored. The facilitators had to work hard to keep participants
on task and to corral the outputs into acceptable formats. By the third day the
anger and critique that dominated the plenary sessions had dissipated and lists
of ways in which each of the relevant parties should improve themselves had
taken their place. Villagers should cooperate with each other; officials should
be sympathetic to villagers livelihood needs; NGOs should be more effective
mediators; scientists should offer villagers agroforestry packages that are both
sustainable and lucrative.39
The report published following the workshop reveals the fractures within
the assemblage, even as it attempts to hold the assemblage together. It declares
on the first page that the participants agreed not to debate concepts and
definitions, instead declaring themselves willing to work together to reverse
forest degradation in Indonesia (CIFOR 2003: ix).40 It presents the harder
hitting critiques of forest policy and practice the unilateral declaration of
forest boundaries, impoverishment and displacement of villagers, timber
extraction that is out of control. Yet it combines these critiques with harmony
language, pointing to a way forward defined by more coordination, consultation, participation, constructive debate and mutual learning by all parties. It is
a masterful piece of assembly work.
From my insider location I came to know something of the struggle over the
recommendations regarding land rights. The Forest Minister had told the
organizers before the meeting that the issue of forest departments territorial
control was not up for debate and should not be discussed. His attempt at
containment failed at one level: virtually all of the written and verbal
submissions by the participants raised the problem of land tenure directly
and repeatedly, as the report observes. Yet he succeeded at another level. After
several drafts and iterations, in the final version the recommendations hedge.
Rather than a frontal critique they adopt the way of the ant. This is Indonesian
activist idiom for conducting a political campaign indirectly by identifying
284
285
the need for good governance, democracy and the rule of law. Second and
more contentiously, community forestry is being promoted as a way to promote
entrepreneurialism and market efficiency. Communities, advocates of this
version propose, can supply forest products and services at a competitive
price.41 Here I tease out some of the ways in which market thinking might be
grafted into the community forest management assemblage while pointing out
the stresses that make it difficult for it to take hold.
For some time, and with increasing intensity since about 2001, the World
Bank and other agencies have been promoting the formal registration and
titling of land in the global south, arguing that this measure improves tenure
security and transparency, encourages investment, enables landholders to raise
capital by mortgage and fosters an efficient land market. Proper land policies,
the experts argue, are essential to growth and poverty reduction (World Bank
2003). Security, in this instance, is the key word that has been re-signified. The
need for tenure security as a basis for livelihoods the concern of advocates
for forest communities is reframed as security for market efficiency, with a
crucial shift. Whereas advocates arguing on behalf of forest communities seek
security from arbitrary expropriation by the state and state-sponsored
corporations, proponents of the neoliberal version highlight the virtues of
security as a means to facilitate commerce. The overlapping discourse of
security is one relay point at which distinct agendas might or might not
be aligned. Critics have not been slow to point out that efficient land markets
equate, in practice, to the enhanced ability of investors to acquire land cheaply
from the poor. It renders them deeply insecure. There are popular
mobilizations to protest Bank land policies.42
A further re-crafting of community forest management to bring it in line
with market agendas hinges on the idea of issuing title to communities.
Although the World Bank has thus far emphasized the superior efficiency of
private individual freehold tenure, it has started to take up the argument of
advocates that communal tenure can also be private, as long as the community
has established a legal personality. The benefit, from a market perspective, is
that the community could then sell, mortgage or lease its assets to
corporations that can use it efficiently. This is a significant reversal of the
project of community forest advocates who seek to use communal tenure to
protect communities inexperienced with markets from the risk of losing land to
more savvy market operators, and to protect vulnerable villagers from the
actions of unaccountable leaders who might connive with outsiders to
dispossess them. Following the paternalistic approach of colonial regimes,
advocates argue that communal land should be inalienable. They worry that
the World Banks communal titling policy facilitates the bulking of land into
large parcels endowed with just the kind of legal certainty corporations require.
Suddenly communal tenure looks risky and the fuzzy boundaries and ad hoc
compromises wrought between villagers and guards on the forest edge look
oddly secure. Advocates still press for the legal recognition of communal
286
tenure but demand tight restrictions, insisting that a market in communal land
is the last thing forest villagers need.
A third element that grafts market thinking into the assemblage is the idea
that community forest management can be linked to conservation and poverty
reduction by making conservation pay. The organization Forest Trends works
to accelerate development of markets for forest ecosystem services (such as
watershed protection, biodiversity conservation and carbon storage), to expand
markets of sustainably produced forest products and to advance markets that
serve the interests of forest communities (Forest Trends 2003).43 Its discourse
aligns the elements in a new sequence: good governance requires that
authorities recognize the property rights of forest communities including
their right to sell forest products and services. Thus communities will manage
forests sustainably not because it is their custom to do so, or because they have
direct use for forest products, but rather because experts have set the
appropriate conditions and market incentives to encourage the approved
behaviour.
Finally, it is worth observing that transnational conservation organizations
such as Conservation International and the Nature Conservancy, frustrated by
the difficulty of governing through communities and recalcitrant forest
bureaucracies, have also shifted to a market logic, buying up large tracts of
forest land or acquiring long-term conservation concessions in which they
can do as they please and neither communities nor forest bureaucracies have
any say.44 Yet these concessions still face the problem of enforcement. It is
possible for a transnational corporation to fence off a high-value enclave such
as a diamond mine and police it with para-military force, but much more
difficult to control the action of tens of thousands of forest villagers in remote
locations, especially when they dispute the legitimacy of their exclusion from
the forest estate. Corporations will still need ways to govern as well as
command, and they might well draw on the community forest assemblage even
as they reconfigure it.
Conclusion
My goal in this article has been to advance an analytic of assemblage. I have
taken up the Foucauldian emphasis on how questions and deployed it to
examine how heterogeneous elements discourses, institutions, laws,
administrative measures, scientific knowledge, moral prescriptions, material
interests are assembled to constitute a technical field fit to be governed and
improved. I delineated six practices that contribute to holding disparate
elements together: forging alignments, rendering technical, authorizing
knowledge, managing failures, re-posing political questions and reassembling
as the ground shifts. I emphasized agency the work that situated individuals
do without re-inscribing the self-sovereign subject with a master-mind, a
master-plan or a singular interest and intention. I also emphasized contingency
287
Acknowledgements
Thanks to John Clarke, Donald Moore, Jesook Song and the journals
reviewers for critical comments on earlier versions. My work has been
supported by Canadas Social Science and Humanities Research Council
through standard grants and through the Canada Research Chair program, and
by a writing grant from the John D. and Catherine T. MacArthur Foundation
programme on Global Security and Sustainability.
Notes
1 See also Dean (1999: 21) and Rose (1999: 52).
2 For an outline of the concept of government as the conduct of conduct and
intervention in the relations between men and things to secure the welfare of the
population, see Foucault (1991). See also Dean (1999), Gordon (1991) and Rose 1999). I
explore the will to improve in Li (2007).
3 Barry (2001: 10 11, 218, fn. 38) similarly stresses agency, process and emergence
over the kind of completed order suggested by Foucaults term dispositif. See the
extended ethnographic treatment of assemblage as emergence in Moore (2005).
4 See Rose (1999: 19) and Dean (1999: 31). In Li (2007) I argue for bringing studies of
government into dialogue with histories and sociologies of rule without collapsing one
into the other. My argument builds upon those of Clarke (2004) and OMalley et al .
(1997) who critique the neglect of practice and instability in studies of government.
288
5 Rose (1999: 48) calls this process translation, a term that suggests the unidirectional conversion of something that is fixed. I prefer forging alignments, which
leaves open the question of how lines are adjusted to accommodate divergent agendas.
See OMalley (1996) and Li (1999) for empirical studies of how programmes were
altered by their target groups.
6 See Rose (1999: 33) on the bounding and dissecting of a technical domain.
7 See Ferguson (1994: 270) and Mitchell (2002).
8 On the attempt to constitute active citizens while containing critique, see
Cruikshank (1999).
9 For a helpful discussion of changing concepts of agency in the earlier and later
Foucault, see Dupont and Pearce (2001: 144).
10 Scott (1998) outlines the combination of utopic dreams and concentrated powers
under which high modern schemes emerged. I situate Scotts approach in relation to
studies of governmentality in Li (2005). Compare Cruikshank (1999: 42) on
government as an accretion of small things rather than totalizing systems. Karl
Polanyi neatly captured the contrast when he argued that the creation of an unfettered
market for land and labour in the nineteenth century was a grand utopian scheme
ideologically driven then as it is in its contemporary neo-liberal guise. Yet the array of
interventions needed to keep populations alive and productive under market conditions
emerged in piecemeal and pragmatic fashion: Laissez-faire was planned, he argued,
planning was not (Polanyi 1957 [1944]: 141).
11 Deleuze and Guattari (1987: 85 91, passim ).
12 Scott (1995: 202) citing the preeminent governmentalist Jeremy Bentham.
13 For an example, see Banuri and Marglin (1993).
14 See Rose (1999: 174).
15 See Gauld (2000) on the Philippines and Fairhead and Leach (2003: 142 5, 150)
on Guinea.
16 Huijsman and Savenije (1991: 25, quoted in Benda-Beckmann and BendaBeckmann 1999a: 1, n. 3).
17 Similar strategies that blend discussion of rights with projects for improvement are
present in texts by Owen Lynch and collaborators (Lynch and Harwell 2002; Lynch and
Talbott 1995) and the advocacy group Down to Earth (DTE 2002).
18 See, for example, the study of invisibilities in the forests of British Columbia by
Willems-Braun (1997).
19 Struggles of this kind in Southeast Asian contexts are described in Hirsch (1993)
and Li (2007). See also Rocheleau and Ross (1995).
20 In Indonesia, the cases of Krui in Sumatra and Bentian in Kalimantan loom large
in the advocacy literature. See Fried (2000) and Michon et al . (2000).
21 See Agrawal and Gibson (1999).
22 My own critical engagement with community forestry fits this pattern (Li 2002).
23 On coercive conservation see Neumann (1998) and Peluso (1993).
24 See Lynch and Harwell (2002) for the legal advocates position and BendaBeckmann and Benda-Beckmann (1999b: 19, 37 8) for the anthropologists critique.
25 See Baviskar (2001), Li (1999), Pannel and Benda-Beckmann (1998: 12, 18), Peluso
(1992) and Vandergeest and Peluso (1995).
26 Campbell (2002: 115) and Li (2001).
27 On the rise of managerialism in environmental discourse to the displacement of
politics, see Brosius (1999), Hajer (1995) and Scoones (1999). Harmony practices are
described in Keeley and Scoones (1999).
28 Schroeder (1999) describes the numerous and stringent conditionalities applied to
community forest management in the Gambia.
29 The donor demand for NGOs to have authentic constituencies is discussed by Jim
Igoe (2003).
289
30 See Anau et al . (2002), Lynch and Harwell (2002: 87 90) and McCarthy (2000).
31 Fairhead and Leach (2003: 151 3) describe the role of donor funds in assembling
community forestry in Guinea.
32 I make this argument in the context of Indonesia in Li (2001).
33 The World Banks new version of the programme has attempted to correct the
problem, apparently without success (World Rainforest Movement 2005).
34 The Joint Forest Management programme in India has received massive donor
investments, mainly in the form of loan funds from the World Bank (Sundar 2000;
World Rainforest Movement 2005). In Francophone West Africa a similar approach,
known as gestion des terroires (territorial management), has been widely pursued.
There have been forty projects covering 2500 villages at a cost of US$280 million in
Burkina Faso alone (Batterbury 1998: 872).
35 An evaluation of the community management programme of Canadas International Development Research Center applauds its provision of well-documented case
studies of pilot projects as learning opportunities that prove the validity of the
concept and demonstrate to policy-makers that alternatives to (failed) state management do exist (Gonsalves and Mendoza 2003: 4). The authors note that the programme
has not yet produced policy shifts or commitments by government agencies to scale up
pilot projects, although they expect this pay off to occur in future years (Gonsalves
and Mendoza 2003: 14, 28, 37, 41). Nowhere do they raise the possibility that existing
state management regimes serve vested interests. All that is lacking for improved
management is perfected technique, generated in projects and packaged in methods
briefs and policy briefs (see Gonsalves and Mendoza 2003: 24 5, 38). Fairhead and
Leach (2003: 48) also note the intrinsic anti-politics of donor-sponsored projects that
supply resources without threatening entrenched interests.
36 Fairhead and Leach (2003: 150) describe donors narration of their role in
protecting forest villagers from corrupt state foresters.
37 Fords work on community management in India was dominated by an exemplary
place a village called Sukhomajri in which the model was perfected. The model
proved very expensive to create and maintain and could not be replicated in large part
because the state bureaucracy saw little to gain from it. The institutional imperatives of
the village were studied exhaustively, but those of the state apparatus were overlooked.
See Baviskar and Saberwal (2002: 54 9).
38 Reform signals the period that began in 1998 with the resignation of Suharto,
president for thirty-two years. Decentralization of control over natural resources to
hundreds of district governments was a measure introduced by his successor, Habibie,
with donor support.
39 Fairhead and Leach (2003: 101 5) describe a corralled consultation process
convened by Conservation International that was set up to confirm prior framings,
although various sets of participants saw through the process leaving it unclear
whether consensus and support had in fact been produced.
40 This document is in Indonesian. All translations are my own.
41 See McCarthy (2006) on the neoliberal inflection of community forestry on
Canadas west coast.
42 See critiques in Food First (2002).
43 See Bhttp://www.rightsandresources.org/ (accessed 30 January 2007) for an
updated statement of this platform.
44 See Fairhead and Leach (2003: 42).
290
References
Agrawal, A. (2001) Common property
institutions and sustainable governance
of resources, World Development 29:
1649 72.
Agrawal, A. and Gibson, C. (1999)
Enchantment and disenchantment: the
role of community in natural resource
conservation, World Development 27:
629 50.
Anau, N., Iwan, R., Heist, M.van,
Limberg, G., Sudana, M.
and Wollenberg, E. (2002) Forest,
Science and Sustainability: The Bulungan
Model Forest: Technical Report Phase 1
1997 2001, ITTO Project PD 12/97
Rev.1(F) , Bogor: CIFOR.
Banuri, T. and Marglin, F.A. (eds)
(1993) Who Will Save the Forest?
Knowledge, Power and Environmental
Destruction , London: Zed Books.
Barry, A. (2001) Political Machines:
Governing a Technological Society,
London: Athlone.
Batterbury, S. (1998) Local
environmental management, land
degradation and the gestion des
terroirs approach in West Africa: policies
and pitfalls, Journal of International
Development 10: 871 98.
Baviskar, A. (2001) Forest management
as political practice: Indian experiences
with the accommodation of multiple
interests, International Journal of
Agricultural Resources, Governance and
Ecology 1: 1 21.
Baviskar, A. and Saberwal, V. (2002)
Fording the Waters: The Ford Foundations
Agriculture and Water Resources Programs
in India and Nepal: A Review, The Ford
Foundation.
Benda-Beckmann, F. von and
Benda-Beckmann, K. von (1999a)
Community based tenurial rights:
emancipation or indirect rule?, In K. von
Benda-Beckmann and H. F. Finkler (eds)
Papers of the XIth International Congress
of the Commission on Folk Law and Legal
Pluralism, Moscow August 18 22, 1997 .
(1999b) A functional analysis of
property rights, with special reference to
Indonesia, in T. van Meijl and F. von
**
**
291
**
**
**
**
292
**
293
**