Hidden alliances : rethinking environmentality and the politics of knowledge in Thailand ' s campaign for community forestry

Recent debates about the politics of environmental knowledge have often invoked the terms, ‘eco-governmentality’ or ‘environmentality’ to analyze how states manage knowledge and political processes in order to regulate citizens and make problems governable. In this paper, we argue that these approaches need to pay more attention to the active participation of society in producing authoritative knowledge, and to the democratic challenges that arise when state and social actors agree on authoritative knowledge. Adopting recent debates from Science and Technology Studies (STS), and especially the concept of co-production, we argue that environmentality should not just be seen in terms of the state invoking ‘power over’ citizens, but as the active engagement of both state and social actors in establishing authoritative knowledge claims simultaneously with claims about appropriate social order, which also exclude other significant social perspectives and environmental management options. We illustrate this debate with the history of legislation and political debates about community forestry in Thailand, where different state agencies and non-governmental organisations have—knowingly or otherwise—collaborated to uphold unchallenged statements about the hydraulic functions of forests in order to strengthen different political positions, but which exclude more diverse development and livelihood options. We therefore argue that political debates about community forestry should therefore pay more attention to how political opponents agree—and the social groups and policy options that are excluded from these agreements—rather than only analyze how one party might have power over another.


INTRODUCTION
In Thailand, official frameworks for community forestry have been discussed for years, but are still largely in paralysis (RECOFTC, 2011).One cause of the slowness in legislation is the opposition between two sets of actors in public debates.On one hand, a broad alliance of the state departments and conservationist non-governmental organisations (NGOs) claim that Thailand's forests perform a crucial role in regulating water supply to the lowlands and protecting biodiversity.These groups therefore argue that there is a need for a strong state presence in managing forests.On the other hand, another alliance of development-and community-based NGOs have argued that community forests offer an important opportunity for devolving power to local communities; that communities have demonstrated an ability to protect forests; and consequently that community forestry represents an important test of democratisation (Hirsch, 1997;Chayan and Aquio, 2000;Pagdee et al., 2006;Ting et al., 2011).
In this paper, we argue that this paralysis in Thailand's community forests needs to be understood in deeper terms than the clash between these two opposing alliances.Rather than reducing the debate about community forestry to these different positions, we call for more attention to how and why these interests have emerged and have become powerful.We also argue that these positions have emerged simultaneously with different knowledge claims about appropriate forest landscapes and social orders.It is only by analyzing the co-evolution of environmental knowledge and political interests can the barriers to community forestry be understood more fully.
By so doing, this paper also questions the approach of many existing frameworks for analyzing environmental knowledge in disputes such as community forestry.In particular, in recent years the approaches known as 'eco-governmentality' or 'environmentality' have emerged as ways to understand how states can engage with knowledge production and policy processes in order to regulate citizens, shape environmental subjectivities, and accordingly influence overall forest policy (Agrawal, 2005;Li, 2007a;Fletcher, 2010;Cepek, 2011).While this paper agrees that there is a need to analyze the impacts of knowledge on policymaking and the formation of environmental subjects, we propose that there is a need to understand more how these actors in opposition to each other can also agree on certain forms of authoritative knowledge, which also exclude social perspectives and alternative land use options.
Consequently, rather than focusing only on how state and social actors disagree, there is also a need to identify the hidden alliances between and within state and society that restrict democratic debate about environmental policy.
In order to do so, this paper first summarises some recent debates about environmental knowledge in community forestry, and especially the various discussions about environmentality as an analytical framework.The paper then reviews alternative Rethinking environmentality in Thailand's community forestry perspectives arising from recent debates in Science and Technology Studies (STS), and especially the concept of co-production (Jasanoff, 2004), which offers the possibility to consider how political interests and knowledge co-evolve and exclude.A third section then applies this debate to a historical analysis of community forestry in Thailand using historic newspaper reporting and a chronology of legislation and activism.Finally, the paper draws lessons for how social science can help make hidden alliances in the production of authoritative knowledge more transparent, and hence contribute to a more inclusive and transparent form of environmental governance.

ENVIRONMENTALITY AND COMMUNITY FORESTS
For many analysts, community forestry represents an institutional challenge of achieving shared behavior in 'forests used in common by a large number of heterogeneous users' (Chhatre and Agrawal, 2008: 13286).Usually, community forests are taken to mean forest landscapes that allow locally determined benefits such as livelihoods, non-timber forest products, pasture, burial sites, or social activities.But these general statements also contain uncertainties about what kind of forest results; the variety of land uses allowed; and appropriate types of state-society relations (Charnley and Poe, 2007;Dove et al., 2011).Indeed, experience in Orissa and the Nepal Terai has indicated that community forests can be administered by local communities in accordance with state laws; or be regulated by communities in defiance of state strictures (Springate-Baginski and Blaikie, 2007;Nightingale, 2009).Moreover, community forests can include a variety of forestry and land-management styles, including plantation forestry and forest conservation, rather than comprising traditional preservation such as sacred forests, which some analysts have suggested should not be upheld as a universal, nor especially robust, model of community forestry (Flint et al., 2008;Charnley and Poe, 2007;Dove et al, 2011;Bullock and Hanna, 2012).
Much research on community forests has, unsurprisingly, analyzed the negotiated relationships between state and communities.
For example, common pool resource management theory has considered how states use institutions and regulations to encourage individualistic resource exploiters within communities more willing to follow management regimes in accordance with other land users (Mukhopadhyay et al., 2007;Thoms, 2008;Larson, 2010).Also, there has been attention to developing plural land uses through a more deliberative and inclusionary approach involving levels of decentralisation of natural resources management (Ribot, 2002).
But in addition, various analysts have considered less cognitive forms of political control involving community forestry, and especially the Foucauldian concept of governmentality.Michel Foucault ([1978] 1991) used this term to refer to the indirect, often unseen, means of influencing 'the conduct of conduct'-or the shaping of political decisionmaking or social behavior by influencing the generation of expert knowledge, regulatory practices, and social expectations.Governmentality may be undertaken by the state, or it can comprise a wider trend (or assemblage) of neoliberal forms of economic growth and development assistance (Fletcher, 2010;Li, 2007a), and hence might be a 'pluralised' form of governance (Gordon, 1991: 36).A key aspect of governmentality, therefore, is in legitimizing world visions, or the information and advocacy arising from expert organisations, in order to persuade citizens, and reduce public debate, about the assumptions and objectives of proposed policies (Ferguson, 1990).Indeed, Agrawal (2005: 223) states, 'an analysis of governmentality orients attention toward the concrete strategies to shape conduct.'Analysts applying governmentality to environmental themes have coined the terms eco-governmentality or environmentality to consider how 'environmentalist programmes and movements operate as forms of governmentality' (Luke, 1998;Cepek, 2011: 501).
A key concern is how environmental knowledge shapes the responses and behavior of citizens.Agrawal (2005: 8) again writes that environmentality 'stands for an approach to studying environmental politics that takes seriously the conceptual building blocks of power/knowledges, institutions, and subjectivities'-and that the objective of studying environmentality is 'to understand and describe how modern forms of power and regulation achieve their full effects not by forcing people toward state-mandated goals but by turning them into accomplices' (Agrawal, 2005: 216-217).
These themes have clear applications in the political analysis of community forestry, as they offer opportunities for analyzing how states can influence communities to adopt formal or informal national guidelines for forest conservation.Agrawal's (2005) study in the northern Indian region of Kumaon, for example, shows how the post-colonial Indian government sought to change the individualistic tree-felling behavior of hill communities through a long-term process of shaping citizens' perceptions and beliefs.A key step in this process was the establishment of forest councils that gave villages a sense of 'intimate governance' that allowed citizens to feel included by the state, but also undertake the state's wishes.Accordingly, environmentality in this example refers to how the state persuaded communities to consider interests other than their own immediate profit-seeking motives, and hence to respect, and indeed enforce, the regulations proposed by the state.
In other studies, Li (2007b: 270), for example, argued that, in Indonesia, 'experts deploying what Foucault called a governmental rationality have sought to manage process and relations, balance diverse objectives, and conduct the conduct of individuals and groups, all in the name of improvement.' Li (2007b: 12) in turn, uses the term, 'provokasi' to refer to the local forms of knowledge and resistance that might challenge the expertise posed by states and other development interventionists.Jepson et al (2012: 851) Rethinking environmentality in Thailand's community forestry also define environmentality as 'the process through which individuals and communities align their environmental practices with state ends'-although these authors note that this process can be both cognitive and non-cognitive.Meanwhile, other analysts have proposed that 'environmentality' might not be occurring because villagers (such as in Amazonia or Indonesia) have not adopted the outside conceptions of environmental change and management (Cepek, 2011;Erb, 2012).
Accordingly, the term 'environmentality' has been used by various analysts to describe the shaping of environmental knowledge and subjectivities simultaneously; usually in terms of how far outside interventionists (chiefly the state) can change more localised, or community-based, perceptions and behavior.

HIDDEN ALLIANCES AND KNOWLEDGE
The works cited above have all offered important insights into the relationships of environmental knowledge and the subjectivities of citizens.We argue, however, that these analyses can be enhanced by looking more critically at how authoritative environmental knowledge is made; how various actors legitimise knowledge; and how other social perspectives and policy options are excluded from this knowledge.
First, there is a tendency in some analyses of environmentality to focus on how states or other actors change the environmental behavior and subjectivity of citizens, rather than also considering how this process changes wider environmental understandings in problematic ways.Agrawal in particular tends to define environmental behavior in terms of regulating individualistic self-interest and accepting the state's information about environmental protection.He writes, 'a desire to protect commonly owned or managed trees and forests, even with the recognition that such protection could enhance one's material self-interest, subscribes to environmental subjectivities' (Agrawal, 2005: 165).Yet, for many political ecologists, this statement overlooks the long-standing debate about the potential negative effects of state-led plantation forestry on biodiversity, water supply, and land rights (Gerber, 2011).
Agrawal's specific approach to environmentality within this statement, therefore, tends to emphasise the transitional impacts of knowledge and governance on citizens' behavior, rather than also acknowledging the implications of the knowledge itself on environment, or aspects of landscape that citizens might value.Other research on eco-governmentality, however, has acknowledged the co-evolution of subjectivity and concepts of environment dynamically.For example, research within feminist political ecology in India and Nepal has highlighted how gender roles, environmental goods and hazards, and political authority Rethinking environmentality in Thailand's community forestry evolve as a process (Nightingale, 2006(Nightingale, , 2011)).In this sense, there is a fruitful discussion about how subjectivity and environmental knowledge are connected; but some well-known applications of environmentality have tended to focus upon how the state regulates a priori definitions of individualistic behavior rather than highlighting how different understandings of environment and citizenship co-evolve, and with what implications for governance, livelihoods, and environmental management.
Second, there is a tendency in various analyses of environmentality to consider the role of environmental knowledge in influencing and controlling citizens, often in tacit ways.There are three challenges to seeing environmentality in this way.First, some political analysts have expressed unease at the assumption that state-society relations should only be understood in terms of opposition and resistance.Instead, there is a need to emphasise the ways in which actors 'bind sources of power into relationships of productive exchange rather than to resist, subvert, or evade them' (Walker, 2012: 58).
The second challenge is that this approach treats knowledge as akin to interests, but various debates from Science and Technology Studies (STS) have argued that authoritative knowledge should be seen as something that is shared, despite the different cognitive positions adopted by political opponents.For example, it is widely acknowledged that 'knowledge' is not simply technical expertise from official state bodies, but comprises a diverse set of information, beliefs, and discourses consultation that-in Foucault's words-'produce[s] domains of objects and rituals of truth' (Foucault, 1977: 194).But, recent STS analysts have argued that authoritative knowledge in itself is usually composed of the tacit, unchallenged assumptions that are adopted by different political opponents, rather than the claims that are clearly in opposition to each other.Consequently, focusing on how one party (such as the state) might wish to dominate others by generating and using knowledge might blind the analyst to the ways in which authoritative knowledge is shared by all parties.
Expanding on this argument, Jasanoff (2004: 20) writes, 'society can be unproblematically conceptualised as composed of interest groups with clearly articulated (exogenous) positions and preferences… newer work recognises the inadequacy of interests as a primary explanatory category.'Instead, Jasanoff proposes the idiom of 'co-production' as an organizing framework to acknowledge how knowledge and visions of social order co-evolve.The crucial point here is that environmental knowledge should not be seen simply as the claims and counter-claims of state and social actors, but that these interactions involve various elements of agreements, or what Hajer (1995: 65) calls 'discourse coalitions,' that give authority to an image of the world, but which can actually exclude various social viewpoints or biophysical understandings of environmental change that do not match this image.
The third challenge is that a co-productionist framework, therefore, does not focus upon how relatively powerful actors generate knowledge in order to have 'power over' less powerful actors, but instead on how authoritative visions of the world are coproduced by this political interaction, and-crucially-on what, and who, is excluded as a result.A key concern here is the representation of authoritative knowledge as 'scientific' by different parties.Science, under a co-productionist framework, is viewed as authoritative and legitimate knowledge in various contexts, rather than simply the domains of laboratories or formally identified experts.Indeed, this kind of authoritative knowledge can even avoid formal scientific research if this research does not reflect visions of social order.
Accordingly, a co-productionist framework of analysis can still be similar to analyses of governmentality because they both seek to understand how knowledge invokes power, and vice versa.But a co-productionist analysis might differ from some prominent applications of environmentality because co-production focuses on how different political actors combine to create authoritative and unchallenged visions of the world, which are exclusionary of other perspectives and knowledge, rather than focusing on how knowledge is generated by different parties in order to dominate each other.Perhaps most importantly, the concept of coproduction also implies that the legitimisation of knowledge about technical concepts such as community forestry occurs simultaneously with a process of social ordering.Hence, the legitimisation of different forms of forest management can only take place when social structures are also legitimised (and delegitimised).In turn, this process of 'ordering nature, and ordering society' (Jasanoff, 2004) poses important questions for debates about governmentality.For example, it is now less clear how environmental subjects 'become' environmental, without also asking how environmental goods and bads are defined (and with which exclusions).
Simultaneously, are definitions of environmental subjects co-produced (i.e.pre-defined) by what is considered environmental?The biggest question asks how can social science reveal how co-production occurs, and hence allow political debate to pay attention to people and environmental knowledge that are excluded?

COMMUNITY FORESTRY AND KNOWLEDGE IN THAILAND
This section asks the following questions:  How does the debate about community forestry in Thailand reflect the shaping of environmental knowledge and environmental subjectivity simultaneously? How do these forms of environmental knowledge and subjectivity exclude other possible viewpoints and environmental management options?
 What are the implications of this analysis for debates about co-production and environmentality, and for community Rethinking environmentality in Thailand's community forestry forestry in Thailand?
In order to do this, the paper focuses first on the apparent social divisions in the political debate about community forestry, and then on the hidden alliances between actors that authorise and give legitimacy to visions of forests and society.The paper refers to a historical analysis of proposals about community forestry legislation and public debates going back to the national ban on logging in Thailand operational in 1989.

Environmental knowledge and community forestry in Thailand
The debate about community forestry in Thailand is usually dated to the national logging ban, passed in 1988 (effective in 1989).
The logging ban was a response to public concerns about rapid rates of deforestation during the 1980s, and forbade cutting of trees, or the planting of tree plantations, by agents other than through official state policies.The campaign for so-called community forestry was an attempt to allow a more flexible form of access to forestland for rural communities, without also losing the ability to control exploitative or unsanctioned deforestation (Pinkaew Leungaramsri and Rajesh, 1992).Since 1989, various bills and proposals for community forestry have been considered within formal legal debates and parliament, but a formal framework has not yet been agreed.In 2003, it was estimated that some 1-2 million people in Thailand lived in zones bordering official protected areas such as national parks and wildlife sanctuaries, but these people had no official land certification or recognition under the law (Sato, 2003).
But, as with other countries, community forestry in Thailand is not simply a technical debate about agroforestry or forest management; it also is an arena in which roles of appropriate citizenship, blame, and responsibility are allocated (Hajer, 1995;Johnson and Forsyth, 2002;Ting et al., 2011).Some important themes are national security, democratisation, and the concern of many citizens about Thailand's rapid transformation and industrial development since the 1960s.In the 1970s, military governments closed many forest areas to public access in order to protect against feared communist insurgency.In the 1980s, an alliance of middle-class conservationists and lower-class peasants successfully campaigned against the Nam Choan hydroelectric dam in the rainforests of the western province of Kanjanaburi.The national logging ban followed.At a time when Thailand was ruled by authoritarian, military regimes, environmentalism offered a means of resistance to state plans.Moreover, analysts have noted how Thai environmentalism has included both lower-and middle-class activists, and wide-ranging concerns such as protecting forest heritage and biodiversity as well as livelihoods for poorer farmers (Hirsch, 1996(Hirsch, , 1997;;Fahn, 2005).
Over time, however, these concerns have also demonstrated different approaches to environmental subjectivity-or allocations of Rethinking environmentality in Thailand's community forestry blame and responsibility-in the debate about community forestry.One important pole of the debate emphasises centralisation and control over forest resources, usually on the grounds of national security and environmental protection.Another visible pole is the desire to decentralise, and offer autonomy to communities, often rooted in wider discussions of democratisation.These viewpoints emerge despite the varying agendas of particular state agencies or non-governmental organisations (NGOs).For example, the main government agencies involved in debates about community forestry are the Royal Forest Department (RFD) (with concerns about forest protection and planting); Royal Irrigation Department (focusing on water supply to agriculture); Department of Land Development (including the control of erosion, and soil productivity); and Department of Public Welfare (which engages with employment, health, and well being of rural populations).Some prominent NGOs include the Wildlife Fund Thailand, and Seub Nakhasathien Foundation (emphasizing environmental protection), and an array of public intellectuals and development organisations that emphasise questions of democratisation and livelihoods within rural development.Indeed, this paper discusses in more detail how the specific agendas and shared beliefs of all these actors represent agreements and/or disagreements about environmental knowledge and appropriate behavior.
The gene and legitimisation of environmental knowledge has been key to establishing different definitions of environmental subjectivity.The Royal Forest Department (RFD) was established in 1896 and became the key governmental actor in defining official forest policies.Information gathered by the RFD was originally targeted at administering and increasing commercial logging; it was only since the 1960s that it was charged with protecting forests (Vandergeest and Peluso, 2006).In 1941 the Forest Act declared 40% of the land area of Thailand as 'forest,' though the actual extent of forested land did not always correspond to this category.
The National Park Act of 1961 started the process of identifying protected land (and forbade the collection of timber and non-timber forest products in these zones).The 1964 Forest Reserve Act, started to acknowledge conservation by gazetting so-called 'permanent' forests as forest reserve land by royal decree.In 1985, the National Forest Policy reinforced the aim of maintaining at least 40% national forest cover by setting aside 25% of Thailand's landmass as economic forest and 15% as conservation forest.
Since the National Park Act of 1964 more than 110 national parks have been declared in Thailand (Vandergeest, 1996;Forsyth and Walker, 2008: 28-29).
Land classification was also undertaken alongside statements about the ecological sensitivity of forests, watersheds, and valleys.These arguments often described negative impacts of upland agriculture, especially in northern Thailand, where various ethnic minority groups practicing different forms of shifting cultivators have lived for centuries, or migrated into Thailand during the twentieth century.During the 1980s, the Watershed Classification Scheme identified five watershed 'classes ranging from Class 1 Rethinking environmentality in Thailand's community forestry headwater areas that are identified as the highest priority for protection, to Class 5 areas on gently sloping or flat land where intensive agriculture is considered appropriate (RFD 2004c).The Royal Forest Department's Watershed Management Division states that Class 1 areas must be strictly kept permanently as head water sources and immediate reforestation programmes must be undertaken on the abundant shifting cultivation area.
The classification of sensitive watersheds, however, cannot be easily separated from the military desire to control strategic border areas, often with histories of insurgency in the 1960s and 1970s.In Phop Phra district of Tak province, for example, 157 Hmong families were moved to lowland sites allegedly to protect the headwaters of local rivers and to maintain the integrity of a newly declared wildlife sanctuary (Pratya Sawetrimon, 1987).Interestingly, this area had been notorious during the late 1960s as a site of Hmong insurgency, and the government responded by building a road to increase military access.This road became so plagued by attacks that it became known as 'The road that cost nine men per mile' (The Nation, 1973).By the 1990s, the Land Development Department's plan for Chiang Mai province classified almost 600 upland villages as 'communities which do not have the potential to become permanent villages and which should be evacuated to more suitable areas' (Bandith Tansiri et al., 1993: 38).
The arguments used to classify land have also been challenged by ecological scientists who have used different data sources (including socioeconomic variables) to calculate the same categories (Pandee and Maathuis, 1990).Moreover, various ecological scientists have challenged the assumptions used to justify these classifications on the grounds that they simplify the cause-andeffect of water shortages and erosion; and diminish the consideration of other options for land management (Bruijnzeel, 2004).For example, it does not always follow that upland agriculture damages the ability of soil to hold water, or that erosion only occurs on agricultural land.Upland agriculture can adopt measures to reduce impacts; plus many non-anthropogenic factors can cause changes to water supply and sedimentation, such as gully erosion under forests, or even forest plantations (Calder, 1999).Despite these arguments, however, government officials have often invoked the language of science and certainty to make public statements about community forestry appropriate forest use.In 1999, the director of the Royal Forest Department, Plodprasob Surasdi, announced, 'A virgin forest is an untouched forest but that's a utopian notion so we have to find a way to mingle the two [forests and human occupation] with minimum impact.But please don't ever say we need humans in the forest to protect it.That's a lie' (Uamdao Noikorn, 1999).
Unsurprisingly, the government-and especially the RFD-made various statements during the 1990s that community forestry Rethinking environmentality in Thailand's community forestry should be adopted on a highly centralised basis, where specially sanctioned communities are allowed to administer forestland in accordance with RFD directives and advice.The RFD, and other government agencies, however, also changed their positions over time, and under different governments.Under the temporary military regime of the early 1990s, the Khor Jor Kor reforestation programme sought to establish eucalyptus and pine plantations on officially sanctioned 'forest' land, but where villages in the northeast of Thailand had settled or grown crops for decades.Unsurprisingly again, villagers resisted these proposals by marching on Bangkok (Pye, 2005).Later, under democratically elected governments, more than 2,000 hill farmers from six northern provinces marched in protest in 1995 against the RFD's desire to evict them from forest reserve land (Bangkok Post, 1995).A similar protest occurred in May 1999 when some 5,000 lowland and hill tribe farmers formed a Rally for Rights in front of the Chiang Mai provincial hall, demanding better land tenure, less state-sponsored reforestation, and better access to citizenship.This, and many similar protests were organised partly by NGOs such as the Assembly of the Poor, which was created from various pre-existing networks of NGOs and trades unions in 1995 to represent poorer farmers and workers in informal Thai politics (Missingham, 2003).Substantial support also came from sectors of the Thai academic community-especially from Chiang Mai University-and a number of prominent public intellectuals (Chayan and Aquio, 2000).

Environmental subjectivities and community forestry
Community forestry therefore became a testing ground for political debates about appropriate forms of democracy and decentralisation in Thailand.Accordingly, activists in favor of more decentralised community forestry justified their position as democratic progress; especially after the 1997 Constitution of Thailand mandated local participation in natural resources policy (which, in turn, was supported by the Decentralisation Act of 1988) (Anan Ganjanapan, 1997).One editor of the English-language daily, The Bangkok Post wrote that the RFD comprised 'gun toting rangers, at the invitation of forest authorities' who seek to evict 'peasants from their ancestral homes' (Sanitsuda, 1999).The limits to citizenship for ethnic minorities in the hills of northern Thailand were also linked to racialisation as a key factor behind the allocation of rights (Vandergeest, 2003).Indeed, a so-called 'People's Version' of the community forestry bill was proposed in 1998 by various members of NGOs and national social advisory organisations, which emphasised the ability of local communities to govern forest land without strong intervention by the state (Anonymous, no date).But this bill also carried implicit definitions of appropriate communities simultaneously with definitions of appropriate forests.In particular, the bill diminished attention to agriculture within community forests as opposed to activities such as planting trees, or maintaining forest areas.The bill also required eligible communities to be those that be 'original local communities' 'that live together as a society in the same area and pass down their culture together' Rethinking environmentality in Thailand's community forestry (Anonymous, no date).These statements present an image of tradition and isolation within villages that avoids the migration and transience of citizens between villages, or the levels of modernisation that have already occurred in many villages.Indeed, the 1997 Constitution of Thailand also implied that 'persons so assembling as to be a traditional community' could participate in resource management 'in a balanced fashion' (Government of Thailand, 1997: article 46).
Statements like these suggest that community forestry (or participation in natural resources management) can only be for appropriate communities, which act in an appropriate fashion.Indeed, civil rights lawyers criticised the People's community forestry bill for prescribing community roles rather than allocating rights.One NGO representative said the bill 'aims to make us responsible for protecting nature in our communities.It doesn't allow a person or group of people to live in, or to make a living in the forest' (Supara Janchitfah, 2002).And another academic advocate of community forests argued that the bill 'gave local communities the right to manage the forests, not to occupy forest land' (Kultida Samabuddhi, 2002;Forsyth and Walker, 2008: 53).The People's Version did not become law.In 2007, an initial community bill, reflecting aspects of the People's Version was passed by the National Legislative Assembly in Thailand.However, this bill has been described as 'contradictory' by some observers, who also argue that Thailand effectively still lacks a community forest framework (RECOFTC, 2011).In particular, NGOs and government agencies still disagree over which livelihood activities are permissible in forests.Article 34 of the 2007 bill stated that community forests should be defined as protected forest that allows local communities to collect non-timber forest products (Rights and Resources Institute, 2008).For many NGOs and intellectuals, this statement is too restrictive of other possible livelihood activities.
By 2009, 7,306 officially registered community forestry sites existed nationwide (amounting to about 400,000 hectares), plus some further 3,000 unregistered sites (Somroutai Sapsomboon, 2009).These sites, however, include a variety of tree-planting projects or communal management of pre-existing forests in both officially protected and unprotected land.Meanwhile, there are still estimated one million or more villagers living in forest reserve land that still have no official recognition (RECOFTC, 2011).

Hidden alliances and exclusions within community forests
The political debates about community forestry in Thailand since 1989 have given the impression that there are two main positions that are engaged in conflict.On one side is a coalition of state agencies and conservationist NGOs who fear that community forestry will decrease state control over threatened forests and resources, and encourage illegal logging and damaging forms of agriculture.
On the other side are development NGOs and public intellectuals who see community forestry as a test case for democratisation Rethinking environmentality in Thailand's community forestry and decentralisation.In turn, these debates have led to a further discussion about appropriate forms of community and land use.
These opposing positions are visible because newspaper reports refer to them, and the statements made by activists, politicians, or academics on either side tend to speak against each other.But in addition to these divisions, there are also areas of agreement between these different positions in the community forestry debate, which have given authority to shared understandings and visions of social order.These hidden alliances are not examples of 'forcing people toward state-mandated goals …by turning them into accomplices' (Agrawal, 2005: 216-217).Instead, they show how agonistic environmental politics can allow political opponents to strengthen environmental knowledge together-and hence exclude mutually other social perspectives and environmental management options.
We argue that there are three aspects to how alliances and exclusions occur.First, the dominant opponents in community forestry debates often use the same framings of forestry and ecology in order to achieve different objectives.In particular, various NGOs and government departments tend to emphasise the same understandings of forest fragility while seeking alternative policies to address this.In one study, Forsyth (2007) analyzed the different contribution of lower and middle classes to the definition of forests in Thailand, and the objectives of their activism, using historic newspaper reports from 1968 to 2000.This research, unsurprisingly, found that most discussions of forests were dominated by middle-class actors who emphasised concepts of wilderness and ecological fragility (or the so-called 'green' agenda).Lower-class actors (such as peasants or factory workers), and the so-called 'redgreen' agenda (emphasizing livelihoods) were less significant.This research showed that forest-based activism might not be as inclusive as claimed by some activists (e.g.Hirsch, 1997;Fahn, 2005).
But this research also showed that during the mid-1990s, newspapers reported that lower-class activists (such as poorer forest users) were more willing to describe forests in terms of the 'green' agenda of fragility and wilderness, rather than in terms of livelihoods and development.Was this a genuine change in sentiment among forest users?A cunning decision by poorer land users to change their stated subjectivities during the community forestry debate?Or did this trend represent a change in how journalists reported farmers, regardless of what farmers thought?This research cannot answer those questions, but it is also clear that the number of official community forests is relatively low, and so actual behavior has probably not changed greatly.
A second set of alliances occurs over the adoption of environmental knowledge about ecological fragility.Various NGOs and state agencies-who otherwise disagree over how community forestry should be implemented-also make statements about underlying hydrological or biodiversity functions of forests in Thailand.First, these actors refer to the damaging impacts of upland agriculture on lowland water supply, flooding, and sedimentation.And then, these statements are used to legitimise forest protection and reforestation as the objectives of environmental policy in highland and other areas.There are many possible reasons for this stabilisation of scientific claims that underpin these questions of land use.The Thai state is composed of various agencies that can agree on reforestation as a convenient solution.The RFD aims to restore, and maintain control over, forest cover.The Royal Irrigation Department emphasises the supply of water to lowland farmers.The military and police have sought to control national security through managing landscapes, especially in border zones.And since Thailand's logging ban in 1989, the need for timber has been managed through establishing state-controlled plantation forestry.As with other developing countries, Thailand's forest policy has been influenced by the use of widespread plantation forestry as a means of managing watersheds, often using Scandinavian pine (Brechin, 1997;Pinkaew Laungaramsri, 2000).
i These viewpoints have coincided with the rise of concerned environmentalists who have, as discussed above, framed forests in terms of ecological fragility and destructive development.Consequently, much debate about environmental degradation in Thailand's watersheds, has analyzed lowland problems such as water shortages through the context of water supply rather than demand, and hence sees restoring forests as a solution (Alford, 1992;Hamilton, 1988).
Political debate therefore might start at discussing how reforestation should occur.One activist scholar wrote: 'if plantation forestry is a logical extension of colonial sustained yield logging, then conservation forestry is its mirror opposite' (Usher, 2009: 10).
And thirdly, these hidden alliances in the framing of debates about forestry result in various exclusions of knowledge reflecting alternative visions of social order.Most pertinently, it arises in a crude classification of different communities as either 'forest destroyers' or 'forest guardians' in ways that might not reflect what these different land users do (Forsyth and Walker, 2008).
Moreover, it also excludes long-standing scientific research about the impacts of upland agriculture and forests on the ecosystem services highlighted by Thai official policy such as flood and sedimentation control.Indeed, the relationship of land-cover change and lowland water shortages and floods is widely debated within hydrology, and varies according to factors such as climate, geology, and valley morphometry, rather than land cover alone.In South Africa, New Zealand, and Nepal, governments have acknowledged a more diverse set of relationships (Calder, 1999).Yet, in Thailand, various state agencies erect signs informing citizens that forest plantations will increase dry-season water supply, and reduce wet-season flooding, even when research has suggested that industrial plantations can absorb more water than they release-or that careful management of soil infiltration capacity under agriculture will reduce runoff occurring on cultivated fields (Ziegler et al., 2001;Calder and Aylward, 2002;Bruijnzeel, 2004).
Many conservationist NGOs repeat these views even if they oppose state policies on community forestry.Often NGOs invoke the Rethinking environmentality in Thailand's community forestry dichotomy of forest guardians and destroyers in order to demonstrate that it is possible for communities to live within forests as defined by the government and conservationist NGOs.In particular, activists and government officials have placed attention on one ethnic group in northern Thailand, the Karen, because of their historic practices of allowing forests to grow back quickly following cultivation and occasional use of traditional rituals that pay respect to forests.For example, Usher (2009: 110) describes how one Karen elder in the Hot district of Chiang Mai province opposes state-led pine plantations.She quoted the elder as saying: 'If they remove the [original] trees from the watershed, the rivers will run dry, the soil will lose its fertility, and we won't be able to grow rice.How will we eat? ' Walker (2001), however, questions this 'Karen consensus' and proposes this representation is driven by the desire of other actors to demonstrate that 'forests' and 'communities' can co-exist, in the terms that conservationist NGOs and the Royal Forest Department portray.For Walker (2012: 192) this kind of representation 'relies on an imagery of local cultural identity, self-sufficient agriculture, and ecologically-friendly lifestyles… that is largely disconnected from the livelihood aspirations of Thailand's commercially connected middle-income peasantry.'Indeed, many Karen upland settlements are actively engaged in commercial agriculture, waged-labour on farms, and migration to cities (Forsyth and Evans, 2013: 103).
In contrast to the Karen, the Hmong ethnic group are typically vilified in NGOs reports and the popular press for being environmental destroyers on account of their different shifting cultivation, and more recently, the use of upland cabbage cultivation and agrochemicals (Forsyth and Walker, 2008).These classifications of appropriate social groups alongside appropriate forms of landscape and environmental behavior are also akin to the People's Version of the Community Forestry bill, which defined community forests simultaneously with communities that had been in place for some years, which did not engage in migration, and which only used forests for non-timber forest product collection.This representation of the Hmong, however, reduces discussion about the ways in which upland agricultural activities can enhance livelihoods without damaging forest ecosystems; and places blame and responsibility for environmental degradation with the Hmong largely because of how their ethnicity is portrayed.

CONCLUSION
This paper has sought to investigate the politics of environmental knowledge in relation to community forestry in Thailand, and examining the benefits and limitations of analytical approaches known as eco-governmentality or environmentality.Clearly, these terms are broad, and it is not our intention to suggest that analysts should either adopt or reject these terms.Rather, our intention is to highlight that recent debates about environmentality have analyze how environmental knowledge and political processes can be used to regulation individualistic community behavior by making citizens accomplices with the state (e.g.Agrawal, 2005: 216-217).While we acknowledge the benefits of this analysis, we argue that there needs to be more attention to how environmental Rethinking environmentality in Thailand's community forestry knowledge becomes authoritative; to see social actors as more actively engaged in legitimizing knowledge; and to look more critically at what social groups and policy options are excluded from authoritative knowledge.
The case of community forestry in Thailand has shown that public debate can become dominated by different visions of community forestry that are frequently presented as the only options.In Thailand, these have been the idea that forests are so threatened and important that the state must be involved; versus the more people-oriented approach that sees decentralisation as an important part of democratisation.Instead, we have presented evidence that these viewpoints have given authority to a vision of forests and their hydrological functions that are highly challenged by scientific research and reduce land-use options.Moreover, these statements are made in coordination with simultaneous categorisations of appropriate forms of community and economic activities, which also restrict livelihood options for ethnic minorities who farm in upland watershed zones.
At the root of this analysis is the need to consider the relationship of interests and knowledge more carefully.There is a tendency for studies of eco-governmentality to analyze environmental politics in terms of how knowledge allows different parties to project knowledge in order to gain political advantage.Indeed, states might do this to regulate citizens; or social groups might appropriate certain knowledge claims or dominant discourses for their own ends to make very different claims on the basis of that knowledge, including as a way to subvert the existing order.Yet, a more representative analysis and outcome also needs to ask how these interactions reify visions of social order simultaneously with knowledge about environmental change.A co-productionist approach (Jasanoff, 2004), instead seeks to see which truth claims are strengthened as the result of political debate, and which alternative knowledge and visions of order are excluded.
Consequently, instead of seeing environmentality as an arena where states manipulate knowledge to hoodwink citizens; there is a need to see how both state and social actors form hidden alliances in the generation and legitimisation of knowledge, which in turn provide reference points for further state-society conflicts.In the case of Thailand, both state and social actors uphold knowledge claims that restrict options for livelihoods and environmental management: looking at what is excluded, rather than who persuades whom, might lead to better outcomes.