Introduction

2Most of the literature about nature conservation focuses primarily on the ‘what’, in other words, on the matter that justifies the implementation of policies and the creation of PAs (Brechin et al., 2002): the biodiversity loss at a global scale. Without questioning the importance and urgency of such issue, it seems equally important to challenge the ‘hows’ and ‘whos’ of conservation, in other words “to question the process by which biodiversity conservation is commonly conducted” (ibid. p. 57). For over a century, conservation policies are not the fruit of a philanthropic ideal but the result of an ongoing effort to separate spaces for nature from those reserved for humans (Adams and Hulme, 1998). Political ecologists have analyzed such policies from a critical point of view to better understand the nature-people relationship involved in the establishment of PAs (Adams and Hutton, 2007).

3Political ecology has investigated the field of conservation for nearly two decades, both through case studies and works seeking to question what is called an “environmental problem” (e.g. the biodiversity loss), along with the “solutions” that are suggested to solve it (e.g. the establishment of a global network of PAs), the social actors involved (e.g. state and non-state actors), and the consequences of the whole process at a local level (e.g. conflicts over access to and control of natural resources). In other words, this approach interprets the conservation issue as a highly political and social one, entailing power struggles between the actors who participate in the definition and implementation of policies (Adams and Hutton, 2007; Brechin et al., 2002; Neumann, 1997; Agrawal and Gibson, 1999; Benjaminsen and Svarstad, 2009 and 2010). Brown (1998) shows for instance, how the adoption of a political ecological analytical framework provides a way for analyzing the influence of science on the policies related to biodiversity management, and their entanglement with understandings, perceptions and values of the most influential actors and interest groups. She concludes that these policies raise issues of environmental justice, as they inevitably create ‘winners’ and ‘losers’.

4The political ecology of conservation mainly relies on post-structuralism. Peet and Watts’ Liberation ecology has been particularly influential: they proposed a political ecology approach “which integrates politics more centrally, draws upon aspects of discourse theory which demand that the politics of meaning and the construction of knowledge be taken seriously, and engages with the wide ranging critique of development and modernity” (Watts and Peet, 2004, p. 5). Briefly said, the authors argue that conflicts for resources are not only conflicts for the control of material resources, but also ideological struggles manifested through discourses (Benjaminsen and Svarstad, 2009; Adger et al., 2001).

5Discourse analysis, particularly important for this post-structuralist branch of political ecologists, draw upon the work of the French philosopher Michel Foucault. In his view, discourse is a “system of representation” (Hall, 2001) which results from the interaction of the binomial knowledge-power. In a Foucauldian understanding, discourse goes therefore beyond the text or discussion to embrace both actions and practices (Svarstad, 2012). Foucault argues that discourses ‘do’ things: they produce meaning, truth and subjectivity while being dependant on the “regimes of knowledge” that circumscribe, at a certain point, what is true from what is false. Power is at the core of these regimes given that it determines which knowledge can be applied and defines what has been called a “field of knowledge” (Hall, 2001). Nevertheless, the philosopher considers discourse as a place of struggles and negotiations between actors who support each specific discursive formation linked with various strategies of power. In the field of environmental policy, the Foucauldian discourse theory has a strong echo, since it brings new perspectives for the understanding of policy processes (Sharp and Richardson, 2001). This work particularly follows Maarten Hajer’s reading of Foucault, who uses discourse analysis to bring out the “secondary discursive realities” of environmental politics, that is to say the close links between policy outcomes and discourses (Hajer and Versteeg, 2005).

6On the basis of these theoretical insights, political ecologists dealing with nature conservation have thus highlighted the existence of three main discourses supporting the implementation of PAs (Adams and Hulme, 1998; Hutton et al., 2005; Benjaminsen and Svarstad, 2010; Rodary et al., 2003). The first one, the “fortress” conservation, relies on the idea that conservation initiatives should be conducted against the local populations, who are blamed for environmental degradation. Stemming from colonial policies, this paradigm dominated conservation discourse until the 1980s. The second discourse, “community” conservation, appeared in the wake of the emergence of the famous concept of "sustainable development". This allowed the broadening of the definition of PAs to development issues: from then on, conservation policies are asked to reach a win-win situation between biodiversity conservation and poverty reduction, in wich local participation is perceived as the key element of this balance. The latest discourse is known as “back to the barriers”: observable from the late 1990s, in a context of accelerated environmental globalization, it promotes conservation policies based on the commodification of nature (Wunder, 2005). It is characterized by a return to exclusive and authoritarian forms of conservation, which put a particular emphasis on a global scientific knowledge (Milian and Rodary, 2010), in response to both the alleged “failure” of community-based approaches and to the emergence of new urgent concerns such as the biodiversity crisis and climate change. Although this political ecological literature has identified a specific historical period for each discourse, it also insists on the fact that these three discourses have to be understood in the sense of “leading discourses”: while being highly influential on policies and practices, none of them exclusively dominated one specific period (Svarstad, 2012).

8The analytical framework defined by political ecologists, relying on a “genealogy” of discourses of conservation starting from the early twentieth century in Africa, is at the basis of this work. The main objective is to shed a light on the evolution of the Beninese policy-making process of nature conservation. More precisely, the contribution of this paper consists in extending the political ecologist’s analytical framework to Benin to assess whether discursive elements exist that can be related to the three aforementioned discourses. This paper is structured as follows. The second section describes the relevant literature and the methodology according to which the discursive analysis is carried out. The third section presents and discusses the discursive elements supporting nature conservation policies in Benin in chronological order. The last section summarizes the findings and draws conclusions.

Methodology

9The analysis of the case study relied on secondary data, mainly collected in 2011 in Benin: policy documents at the national and international level, legislative documents, reports of major international organizations, and scientific papers. The focus of this paper is on the two national parks on the country, the W and Pendjari, and, whenever possible on the particular case of the Pendjari since it is considered as an example of success conservation in Benin and in West Africa in general(Kiansi, 2011; Vodouhê et al., 2010).

10These materials comprise the literature that is subject to discourse analysis based on the Foucauldian approach and political ecological interpretation (Hajer and Versteeg, 2005; Adger et al., 2001; Jensen, 2009). More precisely, the following elements are investigated: 1) the understanding of the problem and the underlying perception of the human-nature relationship, 2) the explanation of the causes of the problem, 3) the solutions suggested and, 4) the actors involved and their roles.

11Researchers in political ecology have shown that nature conservation discourses, and hence the evolution of the network of PAs, have always been shaped by the international environmental agenda and particularly by its understanding of the relationship between poverty and development (Adams et al., 2004; Rodary and Milian, 2008). Jensen (2009) paid special attention to the definition of a generational framework which reflects the evolution of the international environmental debate throughout the 20^th century. The discourse analysis carried out in this paper starts from this useful generational framework in order to define a new periodization, based on the discourses effectively found in Benin.

Discourse analysis of nature conservation policies in Benin

1^st generation (1900-1960)

13From the mid-century, the discourse analysis reveals a progressive broadening of the problem: the destruction of nature and its resources takes a central place in the debate, at the expense of the deforestation one (Aubréville, 1954). The study carried out by Calendra (2000) shows a similar change of approach of the French administration, which passes from a specific focus on the productive sector (forests) to a stronger ideological engagement towards nature protection.

14The objects of protection, from forests to nature more broadly, are clearly understood in a dualistic vision of the human-nature relationship (Castree and Braun, 2001) that encourages a definition of nature as anything that is not human, the so-called African “wilderness” (Neumann, 2004). Human actions are considered as uniformly destructive, referring to the image of the “reckless” indigenous (Castree and Braun, 2001). Looking at the corpus of colonial texts on the subject, the practices of African people are invariably blamed by the authors as the causes of the aforementioned problems (Aubréville, 1937; Roure, 1956). Initially embedded in a civilization discourse, where Africans are cruel and lack of knowledge because of their inferior race, this argument later becomes part of a modernization discourse, in which Africans have to be educated in order to liberate themselves from the so-called “weight of traditions” (Monod, 1950; Aubréville, 1937).

16In the literature in question, the proposed solutions are oriented towards the creation of PAs. The objectives given to such areas mainly reflect the old 17th century French philosophy of “patrimonialism” (Rodary et al., 2003) which promotes exploitation of forest in a perspective of conservation. Its benefits are initially seen as privileging the Métropole but, from the 1950s onwards, the need to address the needs of local populations through conservation actions became increasingly urgent (Aubréville, 1954; Monod, 1950).

17In line with the goal of the object of protection, scientists and the state initially encouraged the establishment of classified forests (RF, 1935) and, in a second phase, of national parks (RF, 1954). In Benin, most of the current protected area network was established during the colonial period and on the basis of Aubréville’s assessment of the state of forests in Dahomey (1937). As elsewhere in FWA, parts of these classified forests were later converted to national parks under the influence of the growing environmental concern (Calendra, 2000). This strategy of PAs creation’s goes hand in hand with highly repressive policies towards local people, whose cornerstone is the 1935 forestry code. This famous decree allowed the extension of the colonial land tenure over the whole forest areas and increased the control of the French colonial administration over people by prohibiting (in classified forests) or restricting (in protected forests) the rights of indigenous people over natural resources (RF, 1935, chap. II). The repression is additionally implemented through exclusionist practices, carried out within the purpose of emptying spaces to protect their inhabitants. As underlined in the literature, these displacements entailed substantial material and social costs, affecting the livelihoods of the hardest hit populations (Brockington and Igoe, 2006). In Benin, the signs of such evictions can particularly be found in Fiasson’s work (1937) on the W Park (Benoit, 1999). Kiansi (2011) also discusses the displacement of people to particular centers at the periphery of the Pendjari park until the 1960s and demonstrates the land inequalities that this process has caused, which are today an important source of conflicts.

18The state, namely the French colonial administration, is the key actor of the protection of nature in FWA. The creation of PAs obviously contributed towards the reinforcement of its control over spaces and people (Rodary, 2008). Neumann (2004) shows how the establishment of “conservation enclosures” is concomitant with the emergence of new spaces called “artifactual wilderness”, which are the product of an idealized conception of nature with state control over the territory. He argues that “wilderness”, and therefore its protection, is part of the process of modern state formation. Throughout the colonial period, foresters played an increasingly important role as their missions expanded. At the end of the period, the structure and activities of forestry services is made more complex since nature protection is part of their remit (Roure, 1956). However, it is the scientists who particularly increase their power during this first generation, thanks, for instance, to the refocusing of the conservation issue on the reasonable use of natural resources (Monod, 1950). Moreover, at the dawn of the independences, they rely on an international coordination as evidenced by the creation of IUPN (Roure, 1956).

19The discourse analysis conducted for this first generation highlights several elements which are typical of the “fortress discourse”: a focus on the “wild”, here the forest, at the expense of human activities, perceived as the only responsible for nature degradation; the promotion of ‘coercive’ policies of conservation and strict protected areas, whose establishment clearly limited the access and rights of indigenous people to resources; lastly, a central place given to the State, that can easily “make rules about who can use nature and where, when and how they can do so” (Adams and Hutton, 2007, p. 152).

2^nd generation (1960-1992)

20After its independence in 1960, Benin experienced considerable political instability noticeable by a succession of military coups. The latest, in 1972, was led by Commander Kérékou, who remained in power for nearly 17 years. In the years immediately following the Kérékou coup (1972-1974) the system of government took the form of a military dictatorship. The second period (1974-1980) is characterized by a radicalization of the regime and the adoption of Marxism-Leninism. The last decade of the Kérékou regime was marked by a certain “deradicalization”, concomitant with a strong recession of the country, which compelled it to institute economic liberalization reforms and to initiate a democratic transition (Bierschenk and Oliver de Sardan, 2003).

22The Beninese conservation program put in place at this tame focused on the economic potential of tourism in the Pendjari National Park. The priority given to the development of sport hunting and the international aid is directed towards technical measures and anti-poaching regulations (Sayer et al., 1979). This post-independent period is especially characterized by the establishment of a legislative arsenal that has limited access to the two national parks and their three hunting zones to any other activities that are neither touristic nor scientific (ibid.). The denial of indigenous people’s rights to the forest is also observable when one sees that “traditional” or “local” hunting is automatically equated with poaching, and that commercialization of all hunting products is prohibited (RPB, 1980 and 1987). Violent repression became common. For example a 1980 edict allows the foresters to use military personnel to assist them in law enforcement actions against poachers (ibid.).

24Here again, we have all the elements characterizing a “fortress discourse”: reinforcement of the barriers of the previously created protected areas, conservation priorities defined according to the potential benefits they can bring to the central state, profusion of the legal instruments aimed at achieving these goals at the same time they marginalize local people in an even more violent fashion than during the colonial period. The novelty is that the authoritarian central state is supported by new influent actors: the international NGOs and intergovernmental agencies.

3^rd generation (1992-2002)

26Discourse analysis of conservation policies in Benin identified one central problem starting from the early 1990s: conservation initiatives are too expensive for the state. This argument is particularly obvious in the feedback on the EEC development project of national parks (PAPN) which highlights the disappointing conservation results compared with the important financial investments (Mahe and Touré, 1990). In parallel, the fact that conservation failed to provide benefits for local people was identified as a serious problem (Green and Chardonnet, 1988). On the ground, this is particularly visible through an increase in the number of park-population conflicts, the latter claiming for more land for agriculture and denouncing parks as exclusively oriented towards international tourism (Aladji Boni, 1984). The “non-involvement of local communities” (RB, 1995) is consequently seen as a major cause of the mismanagement of PAs in Benin, exacerbated by the “return to land” policy advocated in the late 1980s by the Kérékou government which resulted in higher rates of population growth in the rural areas of the country (Bierschenk and Olivier de Sardan, 2003).

29This generation of conservation policies is evidently marked by a decline in the role of the state in conservation. Analysis reveals that the promotion of the participatory approach is more a result than a cause of this change. In other words, the emergence of this discourse is a consequence of the inability of the central state, both in terms of political legitimacy and economic performances, an observation made elsewhere in Africa from the late 1980s (Rodary, 2008). The Renouveau Démocratique has thus been an opportunity to defend a new type of discourse, aligned with the international one. In practice however, the analysis brings out that the change of strategy of the state in favor of local population is not as obvious. This supports the findings of Bierschenk and Olivier de Sardan (2003), who demonstrated that the Beninese democratic turn does not constitute a profound break in the functioning of the state and the logic that it unfolds (bureaucratic-administrative, factional-clientelist, etc.). In their opinion, the decentralization process, yet considered as a condition and a guarantee of good governance, only reinforces the complexity of local-level politics, and speeds up the fragmentation of local political arenas in which the state is profoundly entangled.

30At the same time, this 3^rd generation is characterized by the confirmation of the role of NGOs in conservation policies. As argued by Rodary and Milian (2008), the concept of sustainable development enabled them, from the late 1970s, to re-legitimize their actions in a context of strong criticism about PAs. At the global scale, the number of NGOs increased exponentially from the 1990s, and a multi-level governance system has been progressively established in order to articulate the global strategy with local actions based on PAs (Dumoulin and Rodary, 2005). In the PAs management strategy of 1995, the delegation of some activities to NGOs and private actors is clearly encouraged (RB, 1995). As shown by Bierschenk and Olivier de Sardan (2003), the proliferation of NGOs and private actors should also be linked with the new space left by the decline of the central state. Progressively, the NGOs place themselves as new actors: they both compensate for the lack of the state at the local level (e.g. public services) and are seen as privileged interlocutors by aid donors thanks to their supposed apolitical position towards “civil society”. The greater number and roles of NGOs has served to increase the complexity of local politics.

31In summary, the discourse promoted from the early 1990s is clearly new compared with the previous “fortress” one. It has been identified in the literature as the “community conservation discourse”. While starting from a financial concern about conservation, it relies on the idea that conservation initiatives cannot be carried out against the interests of local people, who should instead reap their benefits (Benjaminsen and Svarstad, 2010). The participation of local population is thus considered as the guarantee of the success of conservation. The literature argues that how this participation occurs is a key element of the sustainability of conservation projects. However, the interests of the central state did not allow for a real reconfiguration of power in favor of local people, and the multiplication of NGOs contributed to the increasing complexity of the situation, where frontiers between the actors of conservation became more and more blurred.

4^th generation (from 2002)

32More than a decade after the democratic turn, the decentralization process in Benin was officially completed in the early 2000s. At the international level, the conservation issue became progressively integrated into more global concerns such as the biodiversity crisis and climate change. PAs seem to have, once again, gained legitimacy in this debate since they support both the global strategy of mitigation and adaptation to climate changes (Dudley et al., 2010).

33The analysis of conservation texts echoes this evolution of the international environmental agenda. The new Beninese strategy 2011-2020 about conservation of PAs is of particular interest. It presents the biodiversity crisis as a serious preoccupation and stresses that the conservation efforts are insufficient given the rates of species extinction and habitats destruction, especially in classified forests overrun by farmers and loggers (RB, 2011; RB, 2002 and 2012). Climate change is considered to be exacerbating biodiversity loss and thus identified as a core problem. Moreover, the discourse analysis puts forward a financial concern in nature conservation, and underlines more precisely the lack of mechanisms to ensure sustainable financing which is not only based on development aid programmes (RB, 2011).

36The solution suggested to solve the financial problem mentioned above can be found in the new PAs strategy. Its primary objective is, for instance, to increase the ability of the Beninese PAs to mobilize sustainable funding. It thus calls to the participation of Benin to initiatives such as the Clean Development Mechanisms (CDM) and the REDD UN programme, and encourages the establishment of a trust fund, the Foundation of West African Savannas, to ensure a sustainable financing of the areas to be protected (RB, 2011).

37The discourse analysis shows that from the 2000s, the actors involved in the conservation discourse are the same as during the previous generation. However, the gap between the national actors and trans-national ones is increasing. The later, such as NGOs or companies, are seen as central interlocutors in a context in which conservation issues are embedded with global concern. As Milian and Rodary (2010) demonstrate, prioritization tools are eloquently produced and advocated by NGOs themselves. In addition to gaining ground on activities previously attributable to the state, NGOs also supplant the so-called communities, as evidenced by the fact that they are in charge of the definition and management of the new PAs created in the southern part of the country (UNDP and GEF, 2008 and 2009).

38To sum up, this last generation positions the protected areas issue on a global scale. The problem is global (the biodiversity and climate change crises), the solution is thus global (protected areas defined by prioritization tools) and the actors privileged are those with global scale connections (NGOs). The discourse analysis therefore clearly puts forward elements that could be linked to the “back to the barriers” discourse, identified by Hutton and colleagues (2005). However, in parallel, the analysis shows a significant persistence of the community conservation discourse since the participative approach is now legally framed (2004 Wildlife Act) and advocated in the management of all the protected areas of the country.

Conclusion

39The discourse analysis informing this work brings a critical perspective to the nature conservation issue in Benin. Three discourses have been found, that match with those identified in the analytical framework: the fortress discourse (1900-1989), the community conservation discourse (from 1990), the “back to the barriers” discourse (from the late 2000s). A deeper comparison with the analytical framework shows that the fortress conservation discourse ended in Benin almost ten years later than elsewhere in Africa, and that the community conservation discourse is nowadays a leading discourse, while being in competition from the late 2000s with a “back to the barriers” one. The discourse analysis also brings some interesting elements about the nature of the shifts in discourses: the transition from a fortress discourse to a community conservation one is mainly the result of external factors and not from a genuine political will to involve populations. More recently, the transition towards a “back to the barriers” discourse arises from the promotion of new dynamics among actors and new issues and not from a real critique of the community-based approach. Finally, the discourse analysis reveals a certain status quo in nature conservation policies during the twentieth century with the persistence of two elements: the economic interest is the first driver of nature conservation policies, and the place allotted to the local populations remains minor, despite the implementation of a community-based conservation approach. Such elements, which may be confirmed by further empirical studies focused on the practices of conservation, will be the subject of future investigations.