■: 7 Rethinking the ! :,. securitization, of the I ; environment i ' {' : Old beliefs, new insights i Maňa Julia Trombetta Other chapters in this book have pointed at the limits of a Formalistic and de-con-textualized approach to securitization. They have shown, in different ways, that die social construction of a security issue is a more dynamic, nuanced and complex process dian the one described by the Copenhagen School (CS). This chapter deals with the implications ofadopting a narrow, textual approach to securitization when analyzing the social construction of global environmental problems as security issues. It thus amplifies die diird assumption developed in Chapter 1, while taking seriously some of the methodological precepts offered thereof The chapter does this by reconsidering some of the debates surrounding environmental security. The first one is related with the opportunity of speaking environmental security: What are the consequences of evoking security? Are they always as problematic as the CS assumes? The case of the environment is a relevant one because the debate is divided between diose supporting the term environmental security, suggesting diat is a good way to promote action and diose who warn against its implications. The second debate is about the practices brought about by securitization: Are they fixed and unchangeable or can diey be transformed by securitizing non traditional issues? The environmental sector is relevant because several appeals to environmental security have been made with the intent of challenging existing security practices and provisions and yet many contemporary security discourses — mentioning precaution and resilience - seem to have been influenced by die environmental debate and concepts. An approach, like securitization, which considers the discursive formation of security issues, provides a new perspective to analyze die environmental security discourse, its potential to transform what counts as security and the ways to provide it. It allows, for instance, an investigation of die political process behind the selection of threats, exploring why some of diem are considered more relevant and urgent dian odiers. In tiiis way, die focus shifts from supposedly objective direats to die collectivities, identities and interests diat deserve to be protected and the means to be employed. In diis chapter, however, it will be shown that die possibility of understanding die transformation of security practices and provisions is precluded because, by focusing on tile textual, formal aspect of speech acts, die CS imposes a problematic fixity on security as a form of social practice. For the School the label security brings with VMi Mmiajuliil Tnmikilii it a specific mindset and a aei of problematic practices associated \viih the logic of war and emergency. For the CS. these practices are not open 10 negotiation or political debate. Accordingly, transforming an issue into a security issue is not always desirable. In the case of the environment, the warning seems clear: "When considering securitizing moves such as "environmental security' ... one has to wcigli the' always problematic, side effects of applying a mind-set of security against the possible advantages of focus, attention, and mobilization"' (Buzan, Waiver and de Wilde 1998: '29). The solution suggested by the CS is to avoid the transformation of an issue into a security issue or to '"dcsccuHtizc" as many issues as possible. This however cannot always be possible or desirable, as the debate about environmental security has shown. First tire performative, constitutive approach suggested by the speech act theory implies that even talking and researching about security can contribute to the securitization of an issue, even if that (and above all the practices allegedly associated with it) is not the desired result (see Huysmans 2002: 4-3). Second, attempts to show that something is not a security issue can lead to the mar-: ginalization and the minimization of urgent threats, especially when several attempts to transform environmental issues into security issues seem to have mobilized actions and produced forms of cooperation rather than conflict. The second reason that makes the environment a relevant case to investigate is that the CS has dealt specifically with it. For the School the environment sector is one that need to be considered to analyze contemporary security dynamics. In tills way several tensions emerges between an empirically driven approach adopted by the CS. which is attentive to the peculiarities of the environmental sector and the attempts to identify the quality7 that makes an issue a security issue or the "security-ness" of security. Amongst the peculiarities of the environmental sector the CS observes diat few attempts to evoke security within die environmental sector have not passed the border oľ ordinary politics or brought about exceptional measures and the logic of confrontation. The School has dismissed those appeals as failed securitization moves that are appeals to security that did not lead to securitization. Against this perspective, or old beliefs, tiiis chapter argues that the securitizations of the environment were indeed successful since they brought about measures and policies diat probably would not otherwise have been undertaken, and yet they contributed to transform the logic and the practices of security. The chapter is in three parts. The first part deals with the limits that a textual approach to security creates in the case of the environment. It introduces the key elements of the dieory of securitization and their relevance for die analysis of environmental security discourses and their implications. This part shows that a discursive approach like securitization can potentially capture several aspects of the transformative intent that characterizes many appeals to include environmental' issues in security analysis, and yet, it points out diat die fixity imposed on security practices by the CS creates an impasse that leads to the problematic suggestion of keeping the label security away from as many issues as possible, including the environment. The necessity oľtliis fixity is challenged by the second part, which outlines a tension between the empirically driven analysis of the environmental sector and the conceptualization of the "sccuriryness of security" and suggests that the R,thmliws> III, yfamlrjtib-n «/. miivimmil I j i securitization of the environment has contributed to bring about u transformation of security practices, i he final | tart provides some examples from die environmental sector. Jt argues for a more comextualizcd approach, which suggests that in a process of securitization not only arc issues transformed into security issues, but also the practices associated with security are challenged and sometimes transformed (Balzacq, 2009a: Balzacq, this volume). More specifically, the chapter deals with two cases of securitization of environmental issues. They are the hole in the ozone layer and environmental conflict. Securitization and the environment: potential and limits In order to explore the potential and limits of securitization theory in dealing with the social construction of environmental problems as security issues, it is necessary to briefly review die key elements of securitization theory: the performative power of evoking security, its inter-subjective nature, and the "specific rhetorical structure" (Buzan, Wsever and de Wilde 1998:26), and analyze them in relation with, the environmental problem. Waver, drawing on Austin's work, considers "security" as a speech act. "In this usage, security is not of interest as a sign that refers to something more real; the utterance iisejfa die act. By saying it, something is done (as in betting, giving a promise, naming a ship)" (Wsever 1995: 55). While this is not the place to discuss whether Weaver's understanding of speech act is appropriate (see Balzacq in this volume), it is relevant to emphasize that Wtevcr is interested in Austin's theory because it captures the power of language in transforming situations and provides a perspective in which the problematic distinctions between "true" and "false" or objective and subjective threats become irrelevant. Accordingly, to say: "global warming is a security issue" is not considered as a constative (that can be true or false - the point, in tiiis perspective, is not to decide whether global warming is a real threat or not), but a performative (that can be felicitous/successful or not). What matters for the School is whether saying that global warming is a direat transforms die way of dealing with it. In this way, the CS does not focus on the truth of a statement but on the "truth effect" of it. Considering the performative power of speaking security opens a new perspective to analyze die development of environmental security discourses and their consequences. Many environmental problems are uncertain and will fully-manifest their consequences in a more or less distant future; this makes the political process of constructing insecurities crucial to understanding why some problems are considered as more relevant and urgent than others or why some issues mobilize action while others are largely ignored. However, focusing on die security utterance only can be problematic because diis could suggest that everything can become a security issue when someone names it that way. Indeed not all the appeals to security transform an issue into a security issue. To avoid this problem the CS distinguish between securitizing moves (Buzan, Wtever and de Wilde 1998: 25), which are appeals to security that can be successful or not, and proper securitization. The School then qualifies securitization in two Ways: first, securitization is a collective phenomenon, "a specific form of social 1 3H Marin Julia Trtiiiiwllti praxis"' (Buzan. Waiver unci dc Wilde 1998: 201}; and second, it lias a specify-rhetoric structure and follows speciiic rules. Secui itization, lor die CS, is a collective phenomenon in two respects. First, it js an inter-subjective practice. One actor can try and say that something vital is at risk and can point at a threat, but a successful securitization is not decided bythc": speaker alone, but by die audience as well: "[SJecurity . , . ultimately rests neither with die objects nor with the subjects but among die subjects" (Buzan. Waiver and de Wilde 1998:31). Securitization in diis way reflects die values and interests of apolitical community. In die case of die environment, its securitization suggests a grow-ing relevance and awareness ofenvironmental problems and a shared aspiration to do something about them. Second, security is about collectivities not individuals: For die School, this is relevant because it allows scholars to "historicize securitv. :o study transformation in die units of security affairs," an opportunity diat for the: School is precluded both to traditionalists, who focus only on die state, and Critical Security Studies that Focuses on the individual (Buzan, Waiver and de Wilde 1998; 206-7). Tliis is quite relevant for die environmental debate since it opens up the possibility of transforming political community through die social construction 'op common direats, and several attempts to link security and the environment embody cosmopolitan intents. As Beck suggests "threats create society and global direats create global society" (Beck 2000: 38).The CS, however, is sceptical about die possibility of a security unity as large as humankind,1 and the reasons have to do not with historical or sociological analyses dial could ottdine die enduring relevance of the state as a security actor, but witii other assumptions of die theory. These aspects are diose related to an antagonistic logic of security and are the same as those that determine die problematic fixity of security practices, which precludes the possibility of analyzing the transformations of security units, at least in universalistic terms. In order to clarify why it is difficult to imagine a security unity as large as humankind, it is necessary to explore the other qualification of securitization provided by die CS, namely that security is a specific kind ofspeech act; it has a specific rhetoric structure and brings into existence a specific set of practices. Security is about "die staging of existential issues in politics to lift them above politics. In security discourse, an issue is dramatized and presented as an issue of supreme priority; Thus, by labeling it as security, an agent claims a need for and a right to treat it by extraordinary means" (Buzan, Wrever and de Wilde 1998:26, emphasis in die original). For the CS this appeal to survival carries widi it a set of connotations that, invokes the logic of "direat-defence," die identification of an enemy and eventually die logic of war (Wtever 1995: 54). The mechanism diat identifies die "securityness of security," die "quality.. . that makes somediing a security issue in international relations" (Buzan 1997: 13), recalls the understanding of the political provided by Schmitt, for whom "die political is die most intense and extreme antagonism, . .. diat of die friend-enemy grouping" (1996: 29). Securitization is identified witii die exceptional decision that constitutes enemies and brings into existence die logic of war. Even if die School does not share diis vision of die political, it suggests diat diis logic characterizes die security mindset. Accordingly the problem with the broadening of the security agenda is diat this mindset is spread as well. Ildhmhug tin m ■'" ili.-jiln'ii aj iiinnmmml 139 In this way. the proWrms with icruritizuiiuu. when tlif- environment is involved, starts to appear. On die one hand, an approach thai considers the discursive fbr-I maiion of security issues provides a new perspective to analyze the environmental security discourse and its transformative potential, h allows, for instance, an analy-! * sis of die political process that leads to prioritizing some issues instead of others, die ! transformation of the political communities that are supposed to be protected, the | jEgittmizing of security practices and the empowerment of the actors that can con-trast specific threats. On die other hand, securitization is problematic for die set of practices it is supposed to bring about, which are supposed to be fixed and based on j, very narrow understanding of what security is about, which is identified as the [iscription of enemies in a context. While die securitization of an issue is open to ■legotiation and political debate, die practices it brings about are not, and they will necessarily come into existence once an issue has been successfully securitized, and, ; moreover, are those practices diemselves diat allow' us to decide whether an issue has been securitized or not. This tension is evident in die long term debate about environmental security, which opposes diose who suggest considering the environment as a security' issue in order to promote action, to focus on die issues that really matter and to adopt a cooperative rather titan a confrontational approach to security, and those who argue that security has a tradition it cannot escape and thus appeals to security should be avoided. The latter argument has been reinforced by securitization theory and the sense ofnecessity it seems to impose. Several commentators have tried to bridge this divide and avoid what Dalby "■. (2001), talking about environmental security, has described as "die dangers in a good idea." Floyd (2007) has suggested diat diere are positive and negative securitizations and diat diis can be decided on die basis of tiieir results. This is largely • based on die consideration that within the environmental sector not all the appeals to security have introduced a confrontational logic, identified enemies or allowed exceptional measures against diem; on the contrary, some or diem have promoted quick and effective actions. However, without challenging die logic of security suggested by die CS, die solution proposed by Floyd seems to imply diat, in some cir-rumstances, die logic ofcreating enemies can be the most appropriate. However, this seems to contradict die attempt to overcome die divisions between the CS and Critical Security Studies since die latter adopts a positive understanding of security (sec Booth 1991, 2007). Another example is provided by Jon Barnett. He first argued diat the securitization of the environment can have perverse effects and shown diat several attempts to transform environmental problems into security issues have resulted in a spreading of die national security paradigm and die enemy logic, even if die intentions behind diem were different. Then, to avoid these problematic developments, Barnett has suggested promoting a "human centered" understanding of security. However, if one accepts die ineluctability of die security mindset and logic evoked b\ securitization: "environmental security is not about die environment, it is about security; as a concept, it is at its most meaningless and malign" (Barnett 2001: 83) one cannot expect diat an appeal to a human centered security will provide 1 -10 Mumi Julia Trumhlm dillVn-iu outcome;.. II tin: practices evoked by speaking security arc fixed and unchangeable, why sin mid the son of claim made by Barncn be different from similar ones? Why should his appeal to a ""human centered security" be ditlerent from the appeals to environmental security, iithe intentions oftlie speakers or the contest arc-' • irrelevant? These dilemmas, however, are based on the idea that security practices are inescapable and unchangeable and the theory of securitization, as elaborated'" by the CS, has contributed to suggest so. Failed securitization or changing security practices? The CS has contributed to making a specific, negative understanding of security-which has characterized the dominant Realist discourse within IR - appear as "natural" and unchangeable since all the attempts to transform it appear to reinfc-.,1-. ;\. logic as die examples from the environmental security debate have shown.2 This perverse mechanism, however, can be challenged by showing that the social construction of a security issue docs not necessarily follow die formal mechanism described by die CS, and die environmental sector provides relevant examples. More specifically, it will be shown that die attempts by die CS to combine an empirically driven approach, which is attentive to the actual processes of securitization and the specificities of different sectors with a de-contextualized "securityness of security" create several tensions and inconsistencies. These tensions will be analyzed by considering the peculiarities of die environmental sector as described by the CS itself. The CS explores die specificity of the environmental sector in Security: A-Framework for Analysis (Buzan, Waever and de Wilde 1998). die theoretical book in which die CS illustrates die theory of securitization and analyses the dynamics of securitization widiin five relevant sectors. For each sector die School identifies the actors or objects (referent objects) that are threatened, specifies the relevant threats and the agents diat promote or facilitate securitization. The environmental sector is radier different from die odiers. Amongst the peculiarities of the environmental sector described by the School, two deserve a specific analysis for dieir implications: first, die presence of two agendas - a scientific and a political one; second, the multiplicity of actors. They both stress die relevance ofa contextualized analysis and the importance of factors which suggest diat die social construction of security issues is more complex than die successful performance of a speech act. This will lead to the final characteristic of die environmental sector, namely the consideration that several securitization moves lead to politicization, radier dian to securitization, since they do not exceed die "normal bounds ofpolitical procedure" (Buzan, Waever and de Wilde 1998: 25). Against diis problematic compromise it will be argued diat the securitization ofnon traditional issues like environmental problems is challenging and transforming existing security practices, but die focus on die fixity of security practices does not allow die CS to account Tor this process. The diree aspects are analyzed in turn. "One of the most striking features of the environmental sector," it is argued in Security, "is die existence of two different agendas: a scientific agenda and a political IMItiiihiKlht •.oirilhalhin vlmiromimit l-l I •inendn" t'Bu/.itii. Wever : 111 < I dr Wilde 1991:1: 711 he-v cxplmn iii.il the scientific ■ ageiid» relets mainly to natural m"it nee and iiun-j>itv<-nmirmul activities and it 'is ibotit the authoritative assessitii-ni of threat" i Buzan. Wti-vcr and de Wilde 1991!: ■ 72). In the case of the environment: the relevance of the scientific agenda is evident jri die attempts to legitimize dillercnt competing claims with the authority of science, but it is often present in other sectors, such as health issues related with the spread of pandemic or other diseases. Buzan, Wtevcr and de Wide argue that "the extent to which scientific argument structures environmental security debates """strikes us as exceptional" (Buzan. Wtevcr and de Wilde 1998: 72), but, quoting Rosenau, diey admit that "die demand for scientific proof is a broader emerging . characteristic in die international system" (Buzan, Wtevcr and de Wilde 1998: 72). This, however, has two implications. First, it seems to challenge die possibility of transforming die way of dealing with an issue by appealing to security and focusing on die "trudi effect" of a statement. In odier words it questions the "self referential-ity" of die speech act security (Balzacq 2005). That is, if one starts to admit that a successful securitization widiin the environmental sector requires specific conditions, namely die presence of authoritative knowledge, or scientific proof, one has also to admit that the specific nature of an issue, an environmental problem, for instance, requires a context and issue-specific analysis. This calls for a more contextualized approach that considers the peculiarities of each case and challenges the possibility of translating the dynamics ofsecuritization from one sector to another. Second, the existence of two agendas has implications for the suggestion of desectiritizing as many issues as possible. Is it possible and what does it mean to "clesecuritize" an issue which is on the scientific agenda? If scientific research outlines the dangerousness of an environmental problem, how is it possible to provide security? This suggests the importance of an epistemic community and experts in a process of securitization, and shows diat some actors are in privileged positions to perform a successful securitization, an argument suggested by Bigo (1994, 2002) to oudine die importance of security" experts and argue against a de-contextualized approach. This leads to the second peculiarity of die environmental sector: the presence ofa multiplicity of actors. The environmental sector is characterized by sccuritizing actors, supporting actors and veto actors. This suggests die political su-ugglc and the complexity of die social construction of threats. This contrasts widi Waiver's suggestion that "security is articulated only from a specific place, in an institutional voice, by elites" (Wtevcr 1995: 57). In die environmental case the multiplicity ofactors is largely justified by the School with the relative novelty of the securitization of die environment. "The discourses, power struggles, and securitizing moves in the odier sectors are reflected by and have sedimented over time in concrete types of organizations - notably states... nations (identity configurations), and die UN system" (Buzan, Waiver and de Wilde 1998: 71). However, this is not the case widi the environment: "It is as yet undetermined what kinds ofpolitical structures environmental concerns will generate" (Buzan, Waiver and de Wilde 1998: 71). This suggests that die logic of security described by die CS refers to a specific one diat has developed with and contributed to die development of specific institutions and, with them, of the actors, practices 14'J Mmia'/iilk Tioiiikltu Liud means that are supposed to provide security. The presence ol several actors is neii only a prerogative of the environmental sector, but it also characterizes otiier nov sectors in which no institutional arrangements arc in place. These considerations lead to die final peculiarity, which can also be considered as the solution adopted by the CS to deal with die problem dial, within the environmental sector, several appeals to security have not brought about die logic or security and the practices associated with it. The third peculiarity is that many seen-' ritizing moves result in politicization. This is problematic for the School, which argues diat "transcending a security problem by politicizing it cannot happen through diematízarion in security terms, only away from such terms" (Wtever 1995. 56). For dre School, once die enemy logic has been inscribed in a context, it is very difficult to return to an open debate. Nevertheless, the various politicizatiurs uf environmental issues diat followed the appeal to security - diose the CS disiH >j j as failed securitizations — seem to suggest that diere is a tendency to politicize issues drrough their securitization. Securitization theory, for die CS, is meant to be descriptive; however, the environ-1' mental sector suggests dial die focus on die formal aspect of die speech act security prevents it from providing an adequate instrument for analysis. A de-contextualized, self-referential approach to security underestimates two aspects: Grst different contexts can have different logics andpractices of security, and diey can influence and challenge each otiier; diis process is not one way only or from die military to die odier sectors. A Idi of work has been done on die implications of applying die (realist) logic of security to environmental issues, while little has been done on how the environmental logic (and which one) influences security practice. This cransfonnation is likely to occur dirough sccuritizing moves - that is, dirough appeals to security in different contexts and for different needs - radier than away from diem. Second, the logic of security itself can change, as new principles, actors, capabilities and direats gain relevance and different security discourses emerge (Huysmans 2002: 58). Environmental security is about transformation and diis is die reason why die environmental sector is so problematic. In order to provide an account of die discursive formation of security issues and of die process of transformation diat securitization implies, it is necessary to move away from die emphasis on die self-referential character of die speech act security to move into the realm of communicative action (Williams 2003: 512) and social change. This is in fine widi die suggestion proposed by de Wilde diat securitization "triggers two debates: one about the underlyingrisk assessment, one about die strategic answer to it" (de Wilde 2008:596). Two cases from the environmental security debate Tins section describes two securitizations within die environmental sector, namely that of die hole in die ozone layer and diat of environmental conflicts. It emphasizes die relevance of a more contextualized analysis and - contrary to die conclusion reached by die CS, which considers environmental securitization as failed securitization moves - suggests diat diese securitizations have challenged and somehow transformed some of the practices associated with securitization. Reilnnkmii fh,- uiuutt -jtUvn <.'f < u;;i annual 143 for each ink- tin- thivaln. tin: st/iiiritizing actor.- ami tin- % ansa! mechanisms invoked arc explored to show how the social cuiisirueiii >n wf'a tin vat i> a more complex matter that relies on diilcicnt actors, shared understanding and symbolic references. The emphasis, however, will not be on showing that securitization does not jelv on formal, linguistic aspects only, but on die implications of this for security provisions and the transformation of die practices of security. In this, respect, the choice of die cases is relevant. In the case of the ozone, appeals to security7 have determined die first international agreementbascd on the precautionary principle. In die case of the environmental conflict, die debate has contributed to promoting preventive approaches. They bodi suggest die relevance of security practices based on prevention, risk management and resilience, which have recendy gained relevance in the climate security discourse. They somehow contrast widi die logic of emergency and exception which characterizes die speech act security as described hy die CS. In diis sense, the logic of security captured by die CS represents a very specific case. An empirically driven, sociological approach can ottdine when it occurs and why. avoiding subsuming all die construction of direats to this logic. The securitization of the depletion of the stratospheric ozone The depletion of die stratospheric ozone is one of the global environmental problems often mentioned as a direat to security (Prins and Stamp 1991; Madiews 1989; Barnett 2001). Waever Buzan and de Wilde mention it in Seamlv and Clinton in die National Security Strategy considers it as a direct direat to die healdiofUS citizens (1998: 13). This prompts several questions: how was it conceptualized as a direat, which actors were involved in die process, which measures resulted from diat conceptualization? The eardi is protected from dangerous high energy radiations by a layer of ozone in die stratosphere. Ozone is a molecule constituted by three atoms of oxygen; it adsorbs die energy of die radiation by splitting into two compounds - a molecule of oxygen and a radical - and tiien recombining again. In die 1970s, concerns emerged diat, in die high atmosphere, exhaust gases could destroy ozone by preventing its recombination. The initial debate was prompted by environmental concerns related to die construction of a fleet of supersonic airplanes by die US, the UK and France and heated by die dispute on landing permits and accusations - on bodi side of die Adantic - of trying to export environmental standards. The issue was largely framed as an environmental problem which might have implications for die national economy, and was not considered as a security' issue. This initial framing I'Litfm 1994:62) contributed to die selection of the actor who became the legitimate scientific autiiority in die field of atmospheric research. Since space expeditions were also suspected of interfering widi die stratospheric ozone, NASA convinced Congress it was die best agency to study die stratospheric ozone's depletion and it soon became a major audiority in die field, providing about 70 per cent of spending on stratospheric research (Litfin 1994:63), In 1974 Rowland and Molina, two chemists at die University of California Irvine, suggested diat CFC gases, widely used in industry for tiieir inertia in die lower atmosphere, can release chlorine into die stratosphere, thus acting as a catalyst in a 1 H Maria Julia Tivmktki :,(.'L ufivuctions thai have the liiinl result ofimpeding the recombination of the ozone moltrulf s>. Tl icy forecasted the depletion ofbetween 7 and 13 per cent of the ozotjj layer. Ai that time, CFCs had an impressive diffusion, bodi as aerosol propellants fur deodorants and as coolers in refrigerators and air conditioners. They were alsQ used for blowing poiyurediane foams, sterilizing medical equipment and for a variety of other uses. They were considered wonderful chemicals, very useful in a variety of settings and with no side-effects. Ozone depletion started to become one of the emerging global environmental problems. The problem was first discussed by UNEP (United Nations Environment Programme) in 1976. The following year a meeting of experts on die ozone layer was convened and UNEP and YVMO (World Meteorological Organization) created a committee to periodically assess ozone depletion (Litfin 1994: 73-5). While research on the atmospheric dynamics was still in its infancy, dierc-w; s _ relevant body of research on the impact of ultraviolet radiation on life. Ultraviolet" radiations is dangerous for people and for various forms of life, causing cancer and blindness. It was the possibility of an impact on human health, diat heated the debate on CFCs and shaped states' actions in die international arena, even before: consensus emerged on the relevance of die thinning oftlie ozone layer and it causes. The debate widiin UNEP and WMO was characterized by the creation of two committees. The choice of two committees (one discussing the economic dimension and the other die health issues) suggested how two contrasting constructions of threats were emerging: die first one considered the direat to die economy ofcutting CFC production, the second the threat to human life posed by the production of these chemicals. Securirizing actors were NGOs and environmental groups, which tried to mobilize states to act collectively. Scientific research on die health impact of high energy radiation played a relevant role in transforming ozone depletion into a threat to human health and in promoting international efforts. In this sense die issue was securitized in the scientific agenda. Despite tile lack of consensus on die extent of die problem and its causes, in 19B1 inter-governmental negotiations to phase out ozone-depleting substances started, Their result was the signature of die Vienna Convention for the Protection ordie Ozone Layer in March 19B5. The Vienna Convention was a framework convention; it did not set up specific targets or incentives but called for common research. For several states involved in die negotiation die main concern was the protection of dieir industries radier dian diat of die ozone layer, which appeared to be a distant, uncertain threat. In this respect this was a failed securitization that did not mobilize exceptional measures. In May 1985, a British research team discovered what was immediately labeled as die "ozone hole." The term hole is actually a metaphor since it refers to a depletion of about 30 per cent of the ozone in the Antarctic region, something rather unexpected and not forecast by any scientific model. Despite the initial scepticism, tlie alarming results were verified by NASA. The authority orscience was somehow challenged since it had not been able to predict such a dramatic development and the relevance of acting on the precautionary principle gained relevance. The broadcasting of NASA images of a computer model representing the polar zone J" !\. ■<>-< ■nih^!i!"n <y ',(i jM-r «-«*m .lepkiii>n in a Iiriglii Hliniiitisoilournimribmed «> the visualization < >l the "link" .umIiIm- j jcreeptkm of a tlir.'iii i*v Lilltii 1994: %-9). This led to a process of securitization of a new kind of threat. While in the negotiation of the Vienna Convention what was supposed to be protected, at least by a number of states, was the industry producing ozone depleting substances, the symbolic representation of a hole in humanity's stratospheric protecting blanket mobilized action. The ozone layer was considered as a fragile asset to lie protected (Litfin lyu !: 97). This created a sense of crisis and die transformation of the depletion of ; stratospheric ozone into an existential threat to the whole of humankind. . - Several sccuritizing actors were involved, from states to civil society and the * scientific community. Boycotting of spray cans and food packages followed. *H Sflh Nevertheless no measures outside die borders of normal politics were taken. ( • Even if scientists cautioned against basing international negotiations on the discov-ry of die ozone hole because there were other plausible causes for this occurrence, it ' - iHtfficulttoimaginediatitdidnotplayaroleintliesubsequentagreements.Therep-i. -entation of ozone depleting substances as a threat to human life contributed to the ' quick signature of the Montreal Protocol, in which 50 countries agreed on a gradual phase-down of CFC production and consumption and set a target of 50 per cent of their 1986 levels by 1998-99, with a ten-year grace period for developing nations. A i'cw months later new scientific evidence confirmed diat the Antarctic phenomenon was likely related widi CFCs and consensus mounted for a total phase-out. The Protocol was amended and strengthened at Conferences or the Parties in London (1990), Copenhagen (1992), and Vienna (1995). The number of controlled substances was increased from the original eight to over eighty. By 1995 most of diem were phased out by die industrialized countries while substantia] steps were taken by sev-I eral developing countries. As Kofi Annan stated: "Perhaps the single most successful international agreement to date has been die Montreal Protocol." The case oftlie Montreal Protocol seems to represent a case in which the politi-cization of an issue occurred through its securitization and not outside it. The rep-■■■■■ resentation of die direat was die result of a social process in which different interests . wete shaped and transformed. The process was characterized by the interplay between the scientific and the political agenda and oudined die dialogical radier than formal nature or the process of constructing an issue as a security issue. : Symbols and images played a determinant role but they had to be framed in a context l haracterized by the production ofcumulative knowledge suggesting a causal : Enk between CFCs and ozone depletion. Decisions however were taken widiout die legitimizing authority ofscientific research and ozone negotiations are die first case ofrnternational agreements based on the precautionary principle. And yet the security measures and provisions were based on cooperation rather than confrontation and included economic sanctions and incentives. Environmental conflict In the aftermath or die Cold War, the number of environmental problems which were argued to have security implications was quite large, including I4li Maria Julia Tiimibrlln problems like climate change, pollution and depletion of natural resaur (Mathews 1989). In die 1980s die emergence of global environmental probleh. like global warming and ozone depletion, determined one of the first attempt i broaden the international security agenda. The Brandt Report (1980) sugycn,.^ that Tew threats to peace and survival of die human community are greate- .]| .n those posed by the prospects of cumulative and irreversible degradation of tin. iu, pherc on which human life depends" (quoted in Brauch 2003: 81) an ■ „ Brundtland report (1987) used the expression "environmental security." [n 1( 1980s the tendency to frame environmental problem in security terms was em aged by peace movements interested in mobilizing action on die issues thai ri j |. matter and by the attempts to promote a non confrontational approach to Lp i r,. military dimensions of security. However, it was with the end of die Cold Wi die debate on environmental security gained relevance. Even if the initial interest for environmental issues was quite broad, rangin , pollution to global warming it narrowed down in a few years. An example s. ,rn vided by the negotiation of die United Nation Conference on Environment u Development; while security was an issue broadly discussed in die prepai. iup. work, by die time of the Conference, held in Rio in 1992, it was no longer nn the agenda and die term security was carefully avoided in die official documents. Several reasons lie behind these de-securitizing moves, such as the concerns of developing countries about green imperialism and interference in dieir se: ni v agendas, the diminishing concern for environmental security in the former i urr-munist countries where the slogan was used to mobilize political action against die So\iet Union. There is, however, an exception, which is die debate about environmentally induced conflict. In this perspective environmental degradation is a security issue since it may contribute to triggering and sustaining violent conflicts. Tl i-argument was ratiier persuasive in the post Cold War environment. It resonated with the more familiar understanding of national security and opened up a nt >.: i for die military. ;! The academic discussion was largely shaped by the work of Thomas Homer-Dixon., who chaired a series of research projects which aimed to study die relatioii-sliips between environmental degradation and violent conflicts (Homer-Dixon 1991, 1994). Even if Homer-Dixon was cautious in suggesting a straightforward connection between environmental degradation and conflict, his argument was spread and amplified by Kaplan's article "The Coming Anarchy" (1994), which forecast massive population displacement and violent conflict, and baldly labeled die environment as die "national-security issue of die early twenty-first century' . (Kaplan 1994: 58). The argument was quite influential witiiin die Clinton administration. As Matthew reports the dien US Undersecretary of State for Global Affairs, •• • ■. ■ Timodiy Wirtli, sent a copy of Kaplan's article to every US embassy and the alarming picture it provided seemed to give an account of die crises die US had to face in Somalia and was struggling to address in Haiti (Matthew 2002: 111), This contributed to die securitization of environmental conflict within defence and political circles, at least in die US. This has promoted furtiier research and political initiatives, in both die US and Europe. This further restMivh .mhiiit impact iind change and thai the best results art* associated with t'arly intervention and preventive measures (Dulfidd and Waddell 200b: 10,'. The debate on environmental conflict has been criticized on normative grounds (Daiby 1999: Barnctt 2001) because it shifted the focus of research on developing: countries and represented people in the Third World as "barbaric Southern Others" (Barnctt 2001: 67); erased the responsibility of developed countries ("of causing environmental change; and tried to frame environmental problem in terms of national security. Nevertheless one has to consider diat this debate and the policies it has determined have achieved two dungs: first, they have legitimized new actors and instruments to develop forms of security governance, which play clnwn'" die role of the state and of traditional reactive responses; second diey have pl0. moted die development of human security and of a new paradigm of preventive measures which are often legitimized by die use of die concept. This does not mean denying the relationships of power or even domination diat are behind die environmental conflict discourse or even the human security one. Duflield and Waddell have considered that discourse as an attempt to broaden the neo-libcral govern-mentality on a global scale. This, however, suggests that die security practices are different from those identified by die CS as are die means to resist or challenge diem. Conclusion Securitization theory has the great merit of conceptualizing the power or discourses, and, more specifically, ofthe word security, in transforming a situation, but die CS!s focus on self referential speech act and die emphasis on the de-contexiual-ized "sccurityness" of security, while providing an elegant theory which captures die structural and social dimensions embedded in language and the problematic: persistence of a set of practices which associate security widi die identification of an enemy and die confrontational logic of war. does not allow us to explore die complexity of die social construction of security issues and explore the potential of a discursive approach. This tension is evident in die analysis of the environmental sector provided by die School itself in Security: a New Frameworkfor Analysis, In diat case, die empirically driven analysis which characterized the original approach of the School and which pays attention to die multiplicity ofactors involved in the process ofsecu-ritization, dieir different rules and capabilities, and emphasizes the importance ol"a scientific and a political agenda, contrasts with the self referential understanding or security suggested by considering securitization as a speech act. The point however is not only about providing a more accurate picture of the social process of die social construction of a threat and ofits implications. The problem is diat the approach suggested by die CS tends to essentialize a specific logic of security and die practices associated with it. This is problematic because die possibility of exploring any transformation in die logic and practices ofsecurity is precluded, and diis is particularly problematic within the environmental sector. Moreover, the CS, in questioning the opportunity of inscribing enemies in a context, suggests the desecuritization of as many issues as possible, leaving unchallenged prulilemuiu n\ \ iraniec:- asM