Beyond argumentation: a practice-based approach to environmental policy

ABSTRACT

We propose that a practice-based approach to environmental policy can help consolidate theoretical understanding of and empirical focus on practices in IPA. Doing so counteracts a tendency to privilege knowledge and discourse in IPA and environmental policy analysis. We draw on multiple strands of practice theory to inform three sensitising concepts: situated agency, logic of practice, and performativity. These concepts provide the analytical tools to investigate how social order and social change originate from the entanglement of meaning and action in practice. We illustrate these concepts by applying them to practices of community forest management in the countries of Ethiopia and Tanzania. Our analysis suggests that a practice-based approach is able to offer nuanced and empirically grounded accounts of political struggles and democratic practices. Its potential is especially strong in cases where policy-making and policy ideas and outcomes are less clearly linked to argumentative processes. We conclude by arguing for further inclusion and consolidation of practice-based approaches within the tradition of interpretive policy analysis.

KEYWORDS:

• Practice theory
• IPA
• discourse analysis
• community forestry
• situated agency
• logic of practice
• performativity

Introduction

What if our language does not simply mirror or picture the world but instead profoundly shapes our view of it in the first place? (Fischer & Forester, 1993, p. 1)

Thinkers once spoke of ‘structures’, ‘systems’, ‘meaning’, ‘life world’, ‘events’, and ‘actions’ when naming the primary generic social thing. Today, many theorists would accord ‘practices’ a comparable honor. (Schatzki, 2001, p. 10)

Over the last two decades, interpretive policy analysis (IPA) has brought us new ways of thinking about environmental problems, solutions, and environmental policy as a whole. The edited volume by Fischer and Forester called The argumentative turn in policy analysis (1993) has contributed greatly to this. Since its publication, a wealth of research articles has challenged the once dominant paradigm of realist conceptions and positivist approaches to policy analysis, especially in the environmental field (e.g. Behagel & Turnhout, 2011; Feindt & Oels, 2005; Hajer & Versteeg, 2005). Even in the relatively conservative field of forest policy, we now see a substantial increase in published articles using IPA (Arts, 2010).

IPA has challenged positivist approaches to policy analysis that are based on philosophical realism (Hajer & Wagenaar, 2003). IPA scholars argue that realist accounts of policy, based on positivism, tend to neglect the important role that values and norms play in defining policy problems and solutions and tend to favour powerful voices over marginalised ones (Yanow, 2007). Interpretive approaches, based on social constructivism, instead hold that values, norms, and facts are entwined and that it is the job of policy analysts to bring this to light (Fischer & Gottweis, 2013). Accordingly, IPA has successfully shed light on practices of deliberative democracy, public participation, local expertise, collaborative planning, and environmental activism, while highlighting their democratic and/or undemocratic aspects. Importantly, these analyses are of high practical value because they capture the messy realities of policy practices (Arts, Behagel, Turnhout, de Koning, & van Bommel, 2014; Fischer & Forester, 1993).

IPA has approached democratic practices and environmental issues from the viewpoint that ‘public policy, formulated through language, is the outcome of argumentative processes’ (Fischer & Gottweis, 2013, p. 489). We argue that this viewpoint has led to an implicit preference for discursive and formalised descriptions of policy processes over more practice-based and informal ones. In our view, this is problematic. Democratic practices of empowerment and autocratic practices of disempowerment are not only found in (mostly Western) experiments with citizen juries, collaborative planning, or competing discourse coalitions. These practices can equally be found when studying the use of local expertise, implementation of community councils, or established traditions. Furthermore, we find examples of environmentalism not only in discourses of conservationism or ecological modernisation, but also in practices of indigenous communities, in local traditions, and so on. To address this shortcoming, we argue that IPA should give more attention to the social and localised practices that are part and/or object of policy implementation (Pülzl, 2013).

In this article, we propose a practice-based approach as a way to counterbalance the privileged position of communication and discourse in many interpretive accounts of environmental policy. Such an approach to environmental policy can help consolidate both a theoretical understanding of and an empirical focus on practices in IPA. The next two sections will set the scene by discussing (1) the roles of discourse and practice within IPA and (2) the practice turn in policy studies. Subsequently, we introduce our practice-based approach and the three sensitising concepts that it employs: situated agency, logic of practice, and performativity. This is followed by two case studies of community forest management (CFM) to illustrate our approach. Finally, we draw our conclusions and argue for a broader application within IPA of a practice-based approach to environmental policy studies.

IPA, discourse, and practice

At the heart of IPA is a ‘focus on meaning as central to individual and collective endeavors’ (Yanow, 2007, p. 117). The idea that language profoundly shapes how we view the world leads IPA scholars to use meaning as a research focus to understand policy processes. Moreover, as ‘interpretation is part of the practice of governing’ (Colebatch, 2014, p. 350), a focus on language is not only a way to gain access to empirical reality, it is also a strategy to directly study that reality. For these reasons, discourse analysis has become an important analytical method within IPA. Discourse analysis has allowed researchers to highlight how the articulation of meaning is closely entwined with political dynamics that include production and normalisation of specific truths, how discourses compete for power, and how discursive practices discipline society and individual subjects (Hajer & Versteeg, 2005), amongst others.

Discourse analysis is – in light of the above – key to the study of environmental politics. Yet, discursive approaches ‘have consistently grappled with the dilemma of how to reconcile meaning and materiality, discourse and practice’ (Neumann, 2002, p. 629). Discursive approaches often engage with the concept of practice in terms of the ‘non-discursive’ (see Behagel, 2012). For example, they conceptualise the non-discursive as social structuration and/or institutionalisation (e.g. Hajer, 1995). Alternatively, they consider discourse itself as an activity or ‘discursive practice’ (Howarth, 2000) and do away with the non-discursive altogether. This is problematic:

… by putting knowledge forward as the major, privileged, or exclusive way of relating to the world, we overload it. We leave out the whole living, experiencing, interacting, embedding material and social environment that often gives rise to what we know in the first place and without which it simply wouldn’t make sense to us. (Cook & Wagenaar, 2012, p. 14)

The role of practice in policy-making has received explicit attention within the broader IPA community. The work of John Forester on practice stories and the ‘deliberative practitioner’ (Forester, 1999) is exemplary, as he teases out how democratic ideals of deliberation become part of the daily practice of public planners. Wagenaar has also contributed to strengthening of the role of practice within IPA, by developing an account of practice as ‘of knowing and doing’ (Hajer & Wagenaar, 2003, p. 25) and ‘meaning in action’ (Wagenaar, 2011), drawing on posthumanist theory (e.g. Pickering, 1993) as well as pragmatic philosophy (Cook & Wagenaar, 2012). Yanow focuses on practice as well as she discusses how various qualitative methodologies produce specific accounts of reality (Yanow, 2007). Finally, recent work on relational ontologies and action research (Bartels & Wittmayer, 2014) highlights the relational, situational, and performative aspects of policy practice (Bartels, 2013).

Practice research is thus an integral part of IPA, yet it has never become a consolidated approach in the way that discourse analysis has. This is particularly true for environmental studies and forest and nature conservation policy. In the latter field, discursive dynamics surrounding forest and nature conservation policy on multiple scales have been well documented (Leipold, 2014). Studies of implementation processes and local processes, however, remain dominated by rational-institutional approaches in the tradition of common pool resource management (Ostrom, 1990). Accordingly, we mostly lack accounts of forest and nature policy that are sensitive to the influence of non-discursive and material aspects of practice, such as how biodiversity responds to global efforts of knowledge-making (Turnhout, Waterton, Neves, & Buizer, 2013). Insight into how institutions structure practice and how local agents respond to institutions is moreover mostly limited to the analysis of design rules, path dependencies, and autonomous human agents. A notable exception is critical institutionalism, that uses a focus on practice to describe how local ‘bricoleurs’ actively accept, change, or reject institutional interventions (Cleaver & de Koning, 2015).

The practice turn

Within social theory, the concept of practice has been key to move beyond problematic dualisms, including object and subject, agency and structure, power and knowledge, mind and body, and nature and society. According to the editors of The practice turn in contemporary theory,

[d]uring the past two decades, practice theory has emerged as a potent challenger to prevalent ways of thinking about human life and sociality, which have until now focused either on individual minds and actions or social structures, systems and discourses. (Schatzki, Knorr-Cetina, & von Savigny, 2001, p. i)

Influential sociologists such as Bourdieu (1977, 1990) and Giddens (1984) have used the concept of practice to argue that social structures do not simply ‘exist’ or influence actors ‘from the outside’, but are produced and reproduced in interaction with them. Foucault also invokes the concept of practice. He highlights the specificity of discourse and argues that discourse cannot be resolved through a system of significations, but instead should be seen as a contingent way of imposing order on the world (Foucault, 1972, p. 229), supported by non-discursive practices.

In the last 15 years – following the publication of The practice turn in contemporary theory– many scholars have embraced practice theory. We cannot find a singular practice-based approach, since a practice perspective ‘emerges from the coming together of several distinct scholarly traditions’ (Nicolini, 2012, p. 9). Freeman, Griggs, and Boaz (2011, p. 130) even speak of a ‘plurality of turns to practice’. While acknowledging that we cannot do justice to this broad range of practice approaches, we will briefly discuss three important practice traditions that have informed our approach: the sociological, the posthumanist, and the pragmatic tradition.

The sociological tradition of practice builds on Bourdieu (1977), de Certeau (1984), and others (also see Bueger, 2014). This tradition focuses on day-to-day practices, such as office culture (e.g. Shove, 2014) or daily practices of consumption (Spaargaren, 2011). By focusing on the role of subjectivity within practices, the sociological tradition contributes to an understanding of practice that is open to agency while taking routines and traditions into account. In the words of Spaargaren (2011, p. 815): ‘[p]ractice theories go beyond individuals but emphasise the fact that human subjectivity is at the heart of processes of structuration, reproduction, and (also environmental) change’.

The posthumanist tradition is especially known for attributing agency to objects and things (Kipnis, 2015). From the perspective of a relational ontology, this tradition posits material agency as semiotically related to and symmetric with human agency (Haraway, 2008; Latour, 2009). From a posthumanist perspective, both material and human agency are ‘temporally emergent in practice’ (Pickering, 1993, p. 564) and constitutive of each other. The posthumanist tradition therefore decentres agency to include material objects.

The pragmatist tradition is closely related to authors such as Bevir (2004), who applies a decentred notion of ‘situated agency’ that highlights local context. Others apply the notion of performativity (e.g. Butler, 1997; Wagenaar & Wilkinson, 2013) to highlight how reality emerges from our practical engagement with it. These concepts all emphasise that ‘performances are not merely the result of conscious acts nor indeed of human agency alone’ (Waage & Benediktsson, 2010, p. 6), but rather shaped in social relationships and material contexts, thus including affects, bodily dispositions, tacit knowledges, and skills (van der Arend & Behagel, 2016).

Common characteristics of what a practice-based approach entails that are shared across various schools of thought include a special attention for qualities of decentredness, emergence, materiality, and relation, amongst others (Freeman et al., 2011; Nicolini, 2012; Schatzki et al., 2001). Practically, we emphasise a number of characteristics for analytical purpose. This list includes broadly shared guidelines for applying a practice-based approach¹:

1. The basic unit of analysis is practice. Analytical focus is neither placed on the social system nor on individual agency, but rather on the entwinement of agency and structure in practice (Schatzki et al., 2001).

2. Social structures such as rules and institutions are produced and reproduced in practice (Bourdieu, 1990; Giddens, 1984). As such, policy implementation is not understood as a linear application of a set of external objectives, but as a dynamic process of interpretation and negotiation in specific sites (Fischer & Forester, 1993).

3. Social practices include both how people relate to other people and to things, artefacts, and other forms of life in their environment (Latour, 2009).

4. Knowledge and discourse are ‘performative’: rather than merely representing reality, they are also constitutive of reality (Callon, 1998) as they are aspects of practice themselves.

5. Practical knowledge is crucial for understanding the workings and endurance of practices (Adler & Pouliot, 2011). Knowledge and action are intertwined, as aptly captured by phrases such as ‘meaning in action’, ‘thinking on your feet’, or ‘practical reasoning’ (Yanow & Tsoukas, 2009).

6. Stability of social practices, in terms of internal logics, patterns, and practical knowledge, is historically informed and not based on universal necessity. Consequently, social practices are always open to change and ‘contingency’ is an important aspect of practice (Behagel, 2012; Schatzki, 2001; Wagenaar, 2011).

7. Practice-based research operates from a ‘flat ontology’ (Collinge, 2006), where order does not precede practice. Specifically, practice-based approaches should neither be reduced to micro-sociology, nor be limited to specific practices (for example policy implementation).

A practice-based approach and sensitising concepts

In our edited volume Forest and nature conservation policy: A practice -based approach (Arts, Behagel, van Bommel, de Koning, & Turnhout, 2013; also see Arts et al., 2014), we outline the basic fundamentals and concepts of a practice-based approach to study forest and nature conservation policy and illustrate these with a number of case studies. Our practice-based approach draws on three sensitising concepts to guide analysis. First, to account for agencies as not only discursively informed but also mediated through material objects and events, we adopt the concept of situated agency. Second, to account for both discursive structures and non-discursive routines, we adopt the concept of logic of practice. Third, to be able to describe the entanglement of meaning and action following the ideals of ‘symmetry’ and to account for tacit knowledge, we make use of the concept of performativity.

The concept of ‘situated agency’ is inspired by Bevir (2004). The concept has two key characteristics. First, it understands agency partly to result from context, or ‘the situation’. It assumes that actors’ ideas, identities, and behaviours are shaped in the social practices in which they are situated (van der Arend & Behagel 2011; Bevir, 2004; Hay, 2002; Giddens, 1984). This is to say that actors follow routines, behave according to specific ideas or discourses, and are constrained by both formal and informal institutions in the pathways of action that they choose. At the same time, their subjective and contingent role in the practice in which they are situated allows them to choose and act otherwise, in particular when confronted with social disruption, political dilemmas, or shock events. In the absence of disruptive events, actors can improvise upon existing situations and also affect change. Second, actors are not autonomous agents. Although they have the capacity for change, they always need to act within and upon a certain practice. The concept of situated agency is therefore useful to describe political struggle and/or resistance without the need to immediately invoke a dominant role for knowledge or discourse that is extraneous to practice, while still recognising the limiting or exclusionary effects of social structures.

The concept of ‘logic of practice’ is adopted from Bourdieu (1977, 1990). We use it to steer attention to how routines and traditions shape practice. First, a logic of practice does not represent a formal structuring of practice. As Bourdieu puts it: ‘practice has a logic which is not that of the logician’ (Bourdieu, 1977, p. 109). The concept allows us to acknowledge that the social is ordered according to various principles (be they discursive, institutional, or otherwise), without awarding this order the status of a pre-designed model, logical structure, or coherent plan. Moreover, a logic of practice is understood to organise doings, sayings, and things by means of a few generative principles. These principles may be discursive, but also non-discursive (i.e. as routine), and may be formal (i.e. as a governmental rule) or informal (i.e. as a cultural practice). The logic of practice is therefore not strictly bound to institutional and/or administrative boundaries and scales. Applying this concept also implies that the analyst does not award institutions too much explanatory power (Cleaver & de Koning, 2015) and thus prevents purely institutional explanations of social phenomena.

The concept of performativity originates from the philosophy of language that argues that language is not a neutral vehicle for the articulation of facts or interests but an active intervention into the world that it seeks to represent (Austin, 1962). We apply it to highlight how practices have emergent properties and include the ‘redoubling’ of meaning and act (Butler, 1997, p. 11). The concept highlights the productive relation between meaning and action and helps flesh out more practical forms of (tacit) knowledge. Performativity also implies that knowledge can be performed through improvisation and gives rise to social change from within a practice. As such, we consider knowledge to be an aspect of practice and do away with the classical notion of representation that views language, science, and knowledge as mirrors of nature’s and society’s realities (Brown, 2009). Finally, knowledge need not only show itself in the form of discourse. Instead, ways of acting upon the worlds can also be seen as tacit forms of knowledge (Polanyi, 1966).

In terms of methodology, a practice-based approach follows the main tenets of IPA with a preference for qualitative methods (although quantitative methods can also be used for the purpose of explorative research), in-depth case studies, and participatory research methods (for a more elaborate discussion, see Arts & Babili, 2013; van der Arend & Behagel, 2011). Below, we use the example of CFM in two countries to illustrate how the practice-based approach can be applied.

Applying the practice-based approach: practices of community forest management

Community forest management

CFM holds that local management of forests, either by communities or jointly with forest departments, is more effective than management by central state institutions, because a ‘sense of ownership’, either legal or practical, and hence responsibility, is given back to the people (Charnley & Poe, 2007). Already in the early 1970s, the idea of ‘participation by local people’, both for better forest management and improving livelihoods, was practised in a few countries, advocated by NGOs and scientists, and intensively discussed in the FAO at the global level (Arnold, 2001; Umans, 1993). Later on, this idea guided international law and agreements on forests, for example, Agenda 21, the Rio Forest Principles, the Convention on Biological Diversity and the Non-Legally Binding Instrument on All Types of Forests. Simultaneously, universalised principles of how to design effective CFM institutions emerged (Agrawal, 2001). Particularly the ‘design principles’ of Ostrom (1990) are now famous. Ostrom derived these principles from a comparison of successful cases of common pool resource management around the world. They include: clearly defined spatial boundaries to the resource; locally adapted access and management rules; participatory decision-making; monitoring of compliance and sanctioning of non-compliance; and procedures for conflict resolution; amongst others. These and similar principles have been applied to many CFM projects around the world, with varying success (International Forestry Resources and Institutions [IFRI], 2015).

CFM has become an internationally acknowledged ‘package’ of universalised ideas, norms, and design principles that has travelled to national and local levels, where it has become embedded in forest law, policy, programmes, and projects, or where it has strengthened already existing CFM-like practices in countries (Arts, Kleinschmit, & Pülzl, 2016). Today, about 15% of the world’s forests is (co)managed by local communities (FAO, 2010). Below, we will discuss two examples: one village in Ethiopia and four villages in Tanzania (all field studies which have recently been supervised and co-published on by the second author of this paper; see Arts & Babili, 2013; Arts et al., 2016; Ayana, Vandenabeele, & Arts, 2015; Vandenabbeele, 2012). These examples show how despite similar CFM ideas, norms and principles, local practices differ considerably. Whereas institutional scholars generally consider such differences as expressions of implementation successes, or failures, due to for example (lack of) group homogeneity, clear rules, and sufficient capacities (IFRI, 2015), we apply a practice-based approach as outlined above. Thus, whether and how a CFM package is adopted in a community is the result of a mixture of: (i) local translations of this package by situated agencies; (ii) deeply rooted routines in established forest practices; and (iii) emergence of unexpected events in social-material fields, as we will show below.

CFM in Ethiopia

Agama is a small village consisting of four hamlets in the south-western part of Ethiopia and located in one of the few remaining forested landscapes in the country: Bonga forest (Vandenabbeele, 2012). Fifty-two families live in these hamlets who are from different cultural and regional descents (Kaffa and Manja indigenous groups and the Kambata settler group). Their livelihoods depend on agriculture, forests, and some off-land income (like migrant labour or small pensions). Historically, local people have been closely related to the forests; economically, socially, and culturally. The forests hold manifold timber and non-timber resources for livelihoods and income, for example, honey, spices, and coffee. Particularly the Coffea Arabica agro-forestry systems in this region are famous. Moreover, the forests are important hotspots of biodiversity (Afromontane rainforests containing numerous rare monkeys, birds, frogs, trees, and plants). Over time, parts of the forests have been degraded and converted to other landscapes. In the region, up to one third of the forests have been lost since the early 1970s and standing forests are often degraded due to human use, including coffee production (Vandenabbeele, 2012). This threatens both biodiversity and local forest-dependent livelihoods. Consequently, Bonga forest was brought under various forms of protection, from a National Forest Priority Area in the 1980s to a UNESCO Biosphere Reserve today.

In Ethiopia, state forestry has been the dominant approach under various political regimes (the Empire till the end of the 1970s, the socialist Derg in the 1980s, and the current semi-democratic state from the 1990s onwards; Ayana, Arts, & Wiersum, 2013). In the early 1990s, CFM started as an experiment in the country. The label that was used for this experiment was PFM, or Participatory Forest Management, because Ethiopia took a co-management approach in which the state forest service and local communities jointly manage public forests. This approach was strongly driven by international NGOs and donors, although endorsed by subnational state agencies. The latter became convinced of the need of such a participatory approach halfway the 1990s for a number of reasons: (i) strongly increasing deforestation rates in the country after the Derg regime was overthrown, (ii) influential international discourses that advocated an integrated approach towards rural development and forest conservation, and (iii) positive examples of this integrated approach elsewhere in the world (India, Nepal, and Tanzania) (Vandenabbeele, 2012). PFM has only very recently become a formal part of Ethiopian national forest policy.

Bonga forest was one of the first experiments with PFM in the country, about 20 years ago. It was initiated by the UK-based NGO Farm Africa and financed by the German donor organisation GIZ. The experiment had substantial consequences for the people of Agama and the forest. In line with the CFM norms and principles, the village was expected to establish a village forest committee; to (re-)demarcate and zone the forests; to set up forest restoration projects; to install new rules on forest access, use, and management (through permits, sanctions, fees, and fines); to set limits on timber harvesting, grazing, and charcoal production; and to implement a system of monitoring (‘forest guards’). All this became a new reality for the villagers in Agama. Its implementation, however, was not a smooth trajectory (Ayana et al., 2015).

In Agama, Farm Africa was officially in charge of the first phase of the PFM project (1996–2000). Farm Africa’s first act was to re-demarcate the forest. This revived old and painful memories of forest nationalisation in the 1980s, when the Derg regime relocated many people out of the forests to villages and deprived them of many customary forest rights. Moreover, the establishment of a forest user group (FUG) also met resistance. It fuelled old conflicts amongst the various groups in the village, related to different levels of (dis)trust by locals vis-à-vis external state and non-state actors; rivalry – including violent acts – amongst indigenous groups and settlers on cultural conventions and forest use; and different perceptions on customary and legal ownership of land and forests. Such came as a total surprise to the project developers, being unaware of these conflict-prone histories of the forest–people interface in Agama. On top of that, the district forest department did not live up to many of its organisational and financial promises under the PFM arrangement. Also, forest officials hardly changed their ‘arrogant’ attitude in the eyes of the villagers (Vandenabbeele, 2012).

After its failed attempt to re-demarcate the forest, Farm Africa wished to ease social conflicts, while decreasing social pressure on the forest. To do so, the NGO started to invest resources in so-called alternative livelihood projects in Agama, such as tree nurseries and small cattle husbandry. Villagers noted that these initiatives indeed brought additional income (Vandenabbeele, 2012). They also acknowledged that Farm Africa increased their awareness of the need of forest conservation. Over time, these livelihood projects started to overtake the original aim of PFM, namely improving local forest management. Moreover, villagers mainly used PFM as a means to (re)emphasise forest land claims and user rights vis-à-vis the state and settler households in the village, while forest management continued to be mainly regulated by customary norms and traditional authorities. Whereas the former points at creative use of the PFM package by local people, thus translating forest management requirements into forest rights claims, the latter expresses the strength of established traditional forest practices.

In 2003, a PFM Agreement was signed between the FUG and regional and local forest officials. FUG was obliged to sustainably manage the Agama forests, about 1200 ha in size, while forest officials had to monitor the management and support it organisationally and financially. This was also the moment that Farm Africa withdrew from Agama, leaving the PFM project behind without ‘roots and wings’ (Vandenabbeele, 2012, p. 132). Later, the FUG was transformed into the Agama Forest Cooperative. Irrespective of its unfortunate prehistory, this body became a rather successful local business in non-timber forest products (NTFPs), although privileging its board and – to a lesser extent – its 220 members. Still, forest management remained in customary hands. All in all, the PFM project ‘never really got into the forests’ (Vandenabbeele, 2012, p. 189). Consequently, the project hardly affected forest quantity and quality around Agama. Even though it definitely induced social change, its outcomes were markedly different from its original intentions.

CFM in Tanzania

In Tanzania, decentralisation of forest management started in the early 1980s (Blomley et al., 2008). The Local Government (District) Authorities Act of 1982 gave village councils the authority to propose bylaws related to natural resources management, while district councils were given the authority to approve these bylaws. This process of decentralisation was also reflected in the National Land Policy of 1995 and the National Forest Policy of 1998. These two policies recognise the active participation of local communities and local government in the management of land and forest resources. CFM was adopted in Tanzania under this umbrella of decentralisation. Over time, the Tanzanian CFM approach developed into two approaches: joint forest management (JFM) for public forest reserves and community-based forest management (CBFM) for village lands (Babili & Wiersum, 2012). Under the JFM arrangement the right holder is usually the government, while the communities are the local managers with support of the district forest office. Under the CBFM arrangement, on the other hand, local people are both owners and managers.

This case study focuses on four neighbouring villages in Babati district in Northern Tanzania (Ayasanda, Boay, Endanachan, and Haraa) that share a forested landscape (Bereku forest reserve and Duru-Haitemba forest reserve) (Arts & Babili, 2013; Babili & Wiersum, 2012). CFM started as a national pilot in Babati in 1994. Before, the forests nearby the villages practically functioned as ‘open access regimes’ and deforestation and degradation were widespread. The villagisation process in the 1970s and 1980s had – as in Ethiopia – led to a formal abandonment of customary forest authority and had spatially separated people from their forests. Traditional connections got lost and forests appeared as ‘no man’s land’, although formally owned and to be managed by the state. The hope was that through the CFM programme these trends could be reversed by returning forests rights to communities – fully or partially – in exchange for active and responsible co-management. In the set-up of the programme in Babati, international donors were strongly involved, both from Sweden (SIDA) and Norway (NORAD), besides district forest officials and representatives of national ministries (Arts & Babili, 2013). The donors provided ideas, methodologies, and tools for the participatory approach to design the projects in the villages and forests. The officials facilitated the negotiation process and endorsed its outcome.

Initially, the projects faced opposition since CFM would imply restrictions on the hitherto more or less free access to forest resources. Accordingly, many villagers feared for their forest-dependent livelihoods. Moreover, the villagers claimed that deforestation was largely due to government (in)activity and requested forest rehabilitation and financial compensation before CFM could start. This initial opposition was softened through meetings, consultations, social pressure, and amendments of the programme to improve the role of and compensation for local people. Particularly the possibility to harvest NTFPs was expanded in the agreements. In the end, the CFM projects were adopted in the village councils.

CFM brought many new things to the villages (a committee, new rules, projects, restrictions, bans, permits, sanctions, fees, fines, monitoring, etc.). This ‘CFM package’ interfered with daily practices and local institutions. Particularly restrictions on the harvesting of timber and a ban on cattle grazing in the forests met with resistance. People continued to ‘illegally’ use the forest for these purposes. When people were caught and fined, the forest guards often met with resistance or refusal to pay. As in Agama, established forest practices were not so easy to change, irrespective of the adoption of new forest management agreements. In an attempt to deal with this, village forest committees started to improvise on rules, for example allowing the harvest of dead wood, lifting or softening the grazing ban, and lowering some fines. Other new rules and norms did not match well with traditional, socially embedded practices either. For example, the village forest committee challenged the role of the aldermen in sacred forest management, diminished the importance of ritual forest patches, and replaced certain customary laws on the protection of spring forests and certain tree species with formal laws. Consequently, the village councils and the forest committees ran into conflicts over who holds authority over what. Also, as an unintended consequence of CFM, some of the traditionally protected trees and forest patches now turned into resources for extraction. A final observation is that the behaviour of district forest officials did change as was the intention: from giving commands and fines to facilitating a new management system and providing guidance to local forest guards. Such roles were of course not immediately internalised, so conflicts about these new roles for forest officials arose now and then.

Contrary to Agama, where traditional forest management largely continued, the CFM projects in Tanzania did affect forest management and ecology (Arts & Babili, 2013). The forest became zoned and partially protected; use of resources were now better regulated (through permits and fines) and monitored (through forest guards); forest fires were now more actively prevented and fought; and reforestation and rehabilitation and regeneration projects were initiated. Some regulations were also softened, as outlined in the above, and some ‘illegal’ use continued. Over time, though, resistance decreased. As a consequence, both villagers and officials observed an improvement in the general forest condition, with the exception of traditionally protected trees and patches. They mentioned an increase in forest cover and in availability of various NTFPs (particularly fodder for cattle), the reintroduction of some lost tree species (for example African teak), less soil erosion on forested slopes, and an increase of wildlife, particularly monkeys and leopards (which was welcomed by the forest rangers, but received with mixed feelings by the villagers). A time series of satellite images of the Bereku and Duru-Haitemba forest reserves under CFM of the four villages – about 2800 hectares in total – revealed an annual increase of about 0.36% in the 2000s. Besides the improved forest condition, villagers also experienced better relationships with forest officials. However, villagers hardly noted any income change as a result of CFM. The latter is a problematic outcome, because improved livelihoods and increased income are considered crucial cornerstones of CFM, particularly so to guarantee long-term support by local communities (Arnold, 2001; Charnley & Poe, 2007).

A practice-based comparison

The case studies show that similar CFM packages affect local community practices very differently. They are ‘unpacked’ in different ways, depending on the situation at hand (Shove, 2014); they are translated to the insights and needs of local officials and peoples (Arts et al., 2016); they interfere with established forest practices, inducing a ‘struggle’ between strongly rooted routines and rationally crafted rules of the game (Bourdieu, 1990); and they might unleash hidden histories of local forest-people conflicts (Vandenabbeele, 2012). Whereas the Tanzanian case showed a positive impact on the forest, the Ethiopian case did not. These outcomes are paralleled by observations of more trust and better relationships with forest officials and less internal conflicts in and around the forest committees in the Tanzanian villages, compared to more oppositional relationships in Agama. Although the new local forest institutions in the Tanzanian villages did trigger social conflicts, the degree of conflict in Agama was more severe. In terms of improving forest-related livelihoods and income, the impact of the projects in both countries was limited, which is in line with findings from the CFM literature (Charnley & Poe, 2007). Yet, there are remarkable differences in how CFM changed local forest management and governance. We argue that a better understanding of these differences requires a practice-based approach.

First, the concept of situated agency shows how CFM projects supported the agency of local communities. In Agama, CFM was partly used to work out a historical issue of forest rights, which had been more difficult under previous regimes. Moreover, villagers used the livelihood projects introduced by Farm Africa to shift focus from forest conservation to livelihood improvement. This agency was partly the result of CFM as a concept (e.g. it includes a focus on forest rights and livelihoods), but equally to be attributed to the positions of local actors, including that of Farm Africa as an implementing agency and the choices it made ‘along the way’. In the case of Babati, situated agencies were also at work. In particular, land use rights connected to harvesting timber and NTFPs and to cattle grazing were repeatedly questioned. Apparently, CFM as a new component of practice was forceful enough to institute new formal rules, but not to (fully) change the activities of local people. In addition, CFM empowered members of the forest village committees to challenge the authority of other ethnic groups and of aldermen, as relations with ritual forest patches lost strength and significance. Hence, it reshuffled power and authority relations.

Second, the concept of logic of practice shows how the new ideas and principles of CFM became part of a local practice that already had its own ordering principles. It offers insight in how CFM could not be integrated in all aspects of existing practices. In Agama, CFM was influential in changing activities to improve livelihoods, but less so when it came to changing forest management. Here, customary norms and traditional institutions were more dominant in guiding forest management activities than the new norms and rules of CFM. Equally, CFM did little to change the roles of forest officials, showing that it could not change the existing social order. That change was more present in Babati, where customary rules and authority on forest management were challenged and partly replaced by the new forest village committees. Moreover, a visible struggle to deal with harvesting and grazing rights shows that CFM became integrated in the daily activities and local norms in Babati. This included the roles of forest officials, which gradually changed from a command-and-control role to a facilitating one.

Third, the concept of performativity highlights how the ideas, norms and principles of CFM are performed very differently across multiple contexts and can bring about unexpected results. In Agama, CFM ideas and norms were performed in terms of land rights and by emphasising livelihood improvement over forest conservation and restoration. Moreover, the establishment of a forest committee led to the unleashing of older conflicts amongst various indigenous and settler identities present in the village. In Babati, the establishment of a village committee did not so much tie in with ethnic conflict, as it did with the division of authority between village alderman and other community members. Furthermore, customary norms of harvesting and grazing were highlighted here over for example customary management of forest. This shows how newly introduced norms are redoubled in acts that invoke the emergence of conflicts in both cases. However, the response to the conflict, and the entanglement of the meanings of CFM with its actions led to very different practices: one focusing on new livelihood strategies, another focusing on power struggles and new practices of forest management.

The analysis according to the three sensitising concepts of our practice-based approach shows that privileging practice in social analysis can offer new insights into how meaning and action are entwined. The cases illustrate that agency need not be ascribed to rational and autonomous individuals and/or organisations alone. Instead, we can see how the introduction of new norms and rules empower actors in some domains, while restricting their actions in others. This empowerment is expressed equally through human subjectivities as it is through more material elements such as cattle grazing and timber harvesting. Moreover, the cases illustrate how practices are ordered according to various logics, some based on ideas and argumentation and others on tradition and custom. We believe it important to show how practical logics can give order to a specific practice, or fail to do so. The examples of debates over forest rights in Agama and over authority in Babati are cases in point. In addition, the concept of performativity shows that a ‘social-ecological field’, in which a project is supposed to be implemented, can respond to its implementers in many unforeseen ways. The revival of old ethnic and insider–outsider conflicts in Agama and the prevalence of modern CFM ideas over sacred forest management in the Tanzanian villages, thus abandoning some forms of forest protection, are clear illustrations. Finally, the comparison between cases show that the way in which policy models such as CFM are performed is decided by practice. Although we only discussed these cases summarily, they illustrate the types of nuanced and empirically grounded accounts a practice-based approach is able to offer.

Conclusion

The argumentative turn in environmental policy analysis has highlighted the role of language in shaping our world in general and environmental policies in particular. However, we have also witnessed a greater focus on processes of policy change and politics than on policy implementation and local practices (Pülzl, 2013). A practice-based approach that consolidates theoretical work on practice within IPA and that is applied to a broad range of empirical practices can correct this imbalance. A practice-based approach allows detailed accounts of various types of political struggle and agency, locally embedded structures and routines, and the performative aspects of knowledge.

The cases on CFM above illustrate the analytical prowess of our sensitising concepts. First, they highlight how the concept of situated agency can shed light on changing power relations as one type of agency is empowered while another is not. These agencies were found to depend on the complex of relations between various traditions, actors, norms, and so on. The concept of logic of practice draws attention to how responses to institutional interventions take place in a spectrum between acceptance and rejection and shows how day-to-day struggles carve out new traditions and roles for actors. The concept of performativity highlights how newly introduced ideas can tie in with historical norms and socially embedded values and bring these to the surface in unexpected ways. It offers a way of considering knowledge as an equal aspect of practice amongst other aspects, rather than knowledge having a privileged position.

In this article, we have argued for a practice-based approach as an additional way of doing IPA, while consolidating various strands of practice research. That is not to say that we propose a practice-based approach to simply replace discourse analysis as the approach of choice in IPA. Instead, we argue for the added value of practice-based research to account for the empirical richness of policy phenomena and to highlight contingent and emergent properties in an ‘eternally unfolding present’ (Cook & Wagenaar, 2012). Specifically, we draw attention to the entwinement of materiality and meaning rather than privileging meaning over materiality. To do so is to expand critical inquiry in democratic and undemocratic practices from argumentative modes of policy-making – and the bias towards specific models of society (i.e. liberal democracies) that these arguably produce – to global and local practices of policy implementation and the way these are shaped by local traditions, indigenous knowledges, and so on.

We recognise that practice-based research is still lagging behind when compared to other research approaches in terms of methodological and conceptual development (Bueger, 2014). The practice-based approach that we apply can assist in ‘catching up’ and contribute to the broadening of the spectrum of research approaches in IPA. There is no single approach to fleshing out all of the important themes and processes that make up environmental policy. Whether it is the rise of new environmental discourses, the integration of indigenous and local knowledge in global biodiversity policy, or questions over inclusion and exclusion of various actors in community forestry, we advocate for a broad understanding of such phenomena that approach both political dynamics and established practices from an interpretive stance.

We argue for further inclusion and consolidation of practice-based research within the tradition of IPA. As a continuation of that tradition, a practice-based approach brings new analytical concepts – specifically situated agency, logic of practice, and performativity – to bear on the analysis of the entwinement of meaning, materiality, and action in environmental policy. Most importantly, practice-based research directs our attention to those practices that may not be at the centre of political debate, but where democratic struggle and the exercise of agency are nevertheless very much present.

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1 This list has previously been used by the authors in a similar form in Arts et al. (2013) and Arts et al. (2014).