The discursive battle for Lysekil

(1)

Department of Political Science

The discursive battle for Lysekil

An argumentative discourse analysis of the Swedish Preem refinery debate

Torsten Blad

Independent Research Project in Political Science, 30 credits International Master’s Programme in Political Science

Fall, 2020

Supervisor: Karin Bäckstrand

Word count (excluding appendices): 20,003

(2)

Abstract

This thesis investigates the role of discourse in our understanding of environmental issues, using an argumentative discourse analytical approach developed by Maarten Hajer. The case under study is the public debate around the Preem refinery expansion in the Swedish town of Lysekil, which mainly took place between 2019–2020. Particularly interesting about this case was the puzzling fact that all actors involved used climate-change mitigation as their main argument, regardless of being for or against the expansion, thus signalling differences in problem perception and representation. The ensuing public debate furthermore provided the opportunity to study Swedish environmental discourse in a non-traditional political arena. The results show the existence of four distinct discourses in the debate: one reformist, one pragmatic, one legalist, and one industrialist. All four refer to climate-change in their arguments but use it to promote different and conflicting actions. Interestingly, the results also show that both the reformist and the industrialist discourses were equally frequent in the debate, indicating the presence of two dominant discourses rather than one. Furthermore, despite the reformist being strictly anti-expansion and the industrialist being for expansion, they showed similarities in their ideological standpoints. Both discourses showed signs of ecomodernism, indicating a diversity within the Swedish ecomodernist discourse. Lastly, I conclude that the different discourses can be understood through the concept of discursive closure, where the ambiguity and simplification of concepts like climate-change lead actors to interpret it differently despite agreeing on the realness and severity of the issue.

Key words: Preemraff, Lysekil, argumentative discourse analysis, discourse coalitions, storylines

Thanks

I would like to thank my main supervisor Professor Karin Bäckstrand, and my secondary supervisor Professor Jonas Tallberg, for their support. Your excellent advice and words of encouragement were greatly appreciated during times of covid-isolation.

(3)

1. Introduction

In March 2020, the small town of Lysekil on the Swedish west coast became a battleground for climate debate. The Saudi-owned petroleum company Preem had in 2018 gotten permission from the Supreme Land and Environment Court (SLEC) to expand its oil refinery – Preemraff – on the town’s factory grounds. However, following appeals from concerned environmentalist groups and individuals the government was prompted to take over the case. The fate of the refinery was to be decided at state level after being judged in court a second time in the town of Lysekil. As the deliberation process began in March, protesters were bused in from all around the country to express their opposition, representing several NGOs and associations for climate and environmental protection. But several voices were also raised in defence of Preem, primarily from company representatives, concerned factory workers and townspeople. The issue further became a partisan conflict, as municipal politicians from both sides of the issue faced off in local newspapers. From a discourse analytical perspective, what was particularly interesting about the ensuing debate was that both the advocates for and against the expansion used climate preservation as their main argument. Those against argued that an expansion would increase Swedish greenhouse gas (GHG) emissions as more oil would be processed in the country. Those in favour argued that while Swedish emissions would increase at first, global emissions would decrease over time as the refinery produced cleaner fuel for ships to use. They also argued that the refinery, if built in Sweden, would produce fuel more efficiently than if built in another country, again resulting in decreased global emissions. Despite this dissention, both sides actually agreed on the severity of the climate crisis. This indicates that the problems of climate-change and other environmental damage was constructed differently by different groups of actors. Both sides claimed to have emissions reduction as their primary objective despite using different narratives of Preem’s role in it (TT, 2020), and therein lies the research problem which is central to this study. What leads different actors with seemingly similar goals to have such different understandings of certain phenomena? The Preem refinery debate is an interesting example of how people use the same concepts to mean different things depending on which discourse they subscribe to.

(5)

2

On September 28, 2020, Preem made the surprising announcement that they would withdraw their application for expanding the refinery in Lysekil (SVT, 2020). After months of speculation in the media about what the turnout would be and increasing pressure on the government to make a decision while preventing what seemed to be an impending crisis between the two ruling parties, Preem cut the debate short by suddenly cancelling their plans.

With a reference to “future economic prospects” (Preem, 2020a), the company presented a new narrative of why an expansion was not economically feasible, displaying a clear shift from their previous discursive standpoint. Regardless of the reason, the issue was seemingly over, and the government had dodged a very controversial bullet. From a discursive viewpoint, however, the question that remained was which of the discourses present in the debate had been dominant.

The conflation of different meanings into one issue has implications on climate-change reduction politics, since the way a problem is defined forms and informs the strategies taken and ultimately shapes the resulting policies. The identification of different discourses is thus helpful when discerning why people argue against each other when they seemingly have similar goals. Moreover, looking at which discourses are dominant is specifically relevant from a political science perspective, since it is when discourses become entrenched and hegemonic that they exert the most power. This study aims to make a theoretical contribution to discourse analytical research by showing a new example of how argumentative discourse theory can be applied and by developing new ways of operationalisation. It also aims to contribute empirically, by looking at a previously unresearched case, and by analysing discursive dominance in Sweden.

1.1 Purpose

The purpose of this study is to understand how the use of language relates to a given context, and to critically analyse how language and practices shape the understanding of environmental issues in different discourses. This is done by analysing the contemporary Swedish political discourses on the role of Preemraff in the context of climate-change. By applying argumentative discourse analysis (ADA) as developed by Maarten Hajer, the study focuses on how different discourse coalitions employ different storylines around the same topic of climate-change mitigation in the Preem refinery case, and how the resulting discourses are shaped by their context. The analysis is done on the discourses through which the concerned actors express their thoughts, opinions and beliefs and thus create meaning. In order to fulfil this purpose, the following research questions are used:

(6)

3

• What discourses were present in the Preem refinery debate?

• Which of these discourses is dominant?

• How can we understand the phenomenon where actors with different interests use the same motives for their arguments?

2. Previous Research

This chapter reviews influential academic contributions to environmental discourse analysis in general and the current state of Swedish environmental discourse as shown in research. The reason for this focus is first to highlight the strengths and weaknesses of the field of environmental discourse analysis in order to motivate the study’s theoretical contribution to the research field overall. Secondly, the inclusion of an overview of Swedish environmental discourses in particular is to provide a broader national context to the case and to emphasise the study’s empiric, as well as theoretical, contributions to the field of discursive research in Swedish environmental politics.

2.1 Environmental discourse analysis and ADA

Discourse analysis has been a useful tool for explaining the continually contested state of environmental politics by highlighting the constitutive nature of language (Hajer & Versteeg, 2005: 178–80). Such research has shown that the meaning of nature itself is disputed and shaped by discourses, and it is partly through common interpretations of environmental problems that these problems can be solved (Dryzek, 2013: 6, 10). Humanity’s view of nature is socially and linguistically constructed based on its contemporary relationship with it and has thus evolved through modern history from terms of wilderness, obstacles and resources, to something sacred in need of protection (Bennett & Chaloupka, 1993: 5–6). Concepts which today are commonplace, such as sustainable development, have been shown to be results of new perspectives on a nature in crisis and the finiteness of resources (Sachs, 1999: 29–30).

(7)

4

Furthermore, environmental discourse research has highlighted the relationship between this knowledge production and power. Those who get to shape the environmental discourse also shape the resulting policies, as indicated by studies on, for example, discourse construction in anti-whaling activism (Epstein, 2005: 60, 64) and universities (Luke, 1999: 104). In his research on the acid rain controversy in the 1970–80´s Great Britain, Hajer introduced the framework which has been used in this study. Similar to other forms of Foucauldian discourse analysis, ADA was developed to explain the government’s approach to problems through discourses in a way that a mere reference to conflict of interests cannot. In an influential study, Hajer (1993) highlights a shift in the dominant environmental discourse from traditional pragmatist to ecomodernist (Hajer, 1993: 48–51). Ecomodernism is the belief that environmental problems can be solved with existing societal arrangements, through concepts such as green growth and sustainable development. As such, saving the climate does not require radical change of existing market-mechanisms due to the compatibility of economic growth and increased environmental protection (Hajer, 1995: 3). This shift changed the meaning making process of defining solutions to environmental issues, from being purely science-based and reactive to being market-driven and proactive (Hajer, 1993: 61–66). Hajer’s popular approach has been used in several other studies, with a particular focus on dominant discourses, strengthening the evidence of ecomodernism’s continued dominance both in Swedish and international environmental discourse (Bäckstrand & Lövbrand, 2006: 70–1; Bäckstrand & Lövbrand, 2019: 528–9;

Zannakis, 2015: 232–3). It has also been applied to other forms of environmental discourse, such as anti-whaling (Epstein, 2008: 112). Furthermore, the concept of dominant discourses within environmental politics has been explored in studies using other theoretical frameworks.

For example, Mert (2019) uses the concepts of ‘political logics’ developed by Ernesto Laclau to explain the shift in hegemonic discourse in Turkey from ecomodernism to what she calls hyper-developmentalism, as well as the opposing discourses of the protesters against the resulting policies (Mert, 2019: 594, 597–601).

However, some authors challenge the predominant view of dominant or hegemonic discourses, claiming that there are in fact multiple discourses in environmental politics, each dominating different forums and forming so called ‘discursive enclaves’ (Stevenson & Dryzek, 2012: 201). Such a plurality of discourses has been shown to often be of a competing nature, although there are examples where discourses complement each other (Dryzek, 2013: 22;

Gustafsson, 2013: 51). The competitiveness between environmental discourses is the reason for

(8)

5

the continued existence of environmental politics, and should therefore, it is argued, not be underplayed (Dryzek, 2013: 12–3).

2.2 Swedish environmental discourse

Sweden is often conceived as an internationally leading country or pioneer in regard to environmental politics and climate policy – a view that continues to be encouraged by Swedish leaders. Consequently, it has been seen as a country where more radical environmental policies could be made possible (Anshelm & Hultman, 2015: 9-10). There are scientifically supported indications that Sweden has adopted and politically institutionalised a discourse of ecological justice, according to which climate-change mitigation is the moral responsibility of developed countries (Zannakis, 2015: 225). However, research also shows that the political environmental discourse in Sweden continues to be dominated by ecomodernism, similar to most other countries in the world (Anshelm & Hultman, 2015: 9-10). This is indicated by an increasing overall trust in market- and growth-based solutions to environmental issues reflected in Swedish environmental policy (Johansson & Henriksson, 2020: 153-4). In Sweden, the morally driven aspiration to be a forerunner in climate-change mitigation has increasingly been coupled with the claim that GHG reduction and economic growth are compatible (Zannakis, 2015: 231). Not only that – climate-friendly investments are increasingly framed as an economic opportunity, overtaking the ecological justice discourse. This has produced a win-win narrative where climate investments drive economic growth, and climate-change mitigation relies on market- driven technological developments (ibid.: 232–3). However, despite this apparent trust in science and technology, research also shows that Swedish policy discourse often relies on poorly supported assumptions about science by constructing discursive stereotypes (Hellström

& Jacob, 2005: 462–4). This indicates that the dominant ecomodernist discourse in Swedish policy, wherein markets and technology are supposed to be the saviours of environment and climate, is perhaps overly simplified. Previous research mainly covers dominant discourses in environmental policy, but these are not the only environmental discourses in Sweden. The way ideas about environmental issues are (re)produced differs between schools, workplaces, media and politics. This study makes an empirical contribution by focusing on the public discourse around a well-known case, providing an interesting addition to Swedish environmental discourse research and an alternative to policy analysis.

(9)

6

2.3 Summary and contributions

Previous research shows the contribution discourse analysis has made in environmental and climate policy research. Its ability to highlight the constitutive properties of language and the relationship between meaning-making and power has been useful in a contested field reliant on the communication of scientific knowledge. However, the focus on dominant discourses has left a research gap on the existence of multiple contesting discourses within environmental politics. Similarly, the studies on Swedish environmental discourse have largely focused on state-level and policy-based discourses, leaving a gap in research on alternative discursive sites.

As such, this study makes three contributions. First, by highlighting multiple contesting discourses instead of one dominant, it makes an empirical contribution to the field of environmental discourse analysis. Second, by focusing on popular discourse outside the traditional political arena, it makes an empirical contribution to the field of Swedish environmental discourse analysis. Third, by developing a novel operationalisation of Hajer’s theoretical concepts, it makes a theoretical contribution to the field of ADA.

3. Theoretical framework

Since the aim of the study refers to how language and actions shape meaning and thus social reality, this precludes the use of a naturalist ontology where the social world exists separate from human interpretation. Instead, this study uses an interpretivist ontology actualised through discourse analysis. In discourse analysis, theory and method are very much intertwined. In order to follow a standard thesis structure, this chapter’s focus lies on the theoretical reasoning which underlies and enables discourse analysis in general and Hajer’s ADA in particular. The choice of ADA over other discourse analyses as a research method is motivated in the methodology chapter.

(10)

7

3.1 Discourse analysis

3.1.1 Poststructuralism

Discourse analysis builds on the teachings of poststructuralism. If summarised, poststructuralism seeks to question essentialist ideas of the world as something static, and instead proposes a more contingent worldview. In regard to language, poststructuralism sees words and expressions as un-fixed and dependent on specific contexts or discourses, where the meaning of a signifier is ever subject to the possibility of change. Similarly, broader societal concepts are portrayed as historically situated, where the institutions and discourses we today take for granted as truth are actually just a product of the times (Wenman, 2018: 126, 130–2).

From this follows that the dominant words and actions in a certain context and the meaning concerned actors ascribe to them – i.e. a dominant discourse – shapes what is considered to be normal or factual (Mert, 2015: 40). The use of such a theoretical base is appropriate when trying to discern how actors ascribing to different discourses make sense of and talk about environmental issues, since it highlights the fact that the meaning of a concept is contingent on these very discourses.

3.1.2 Discourse theory

“…for interpretative environmental policy research, it is not an environmental phenomenon in itself that is important, but the way in which society makes sense of this phenomenon.” (Hajer & Versteeg, 2005: 176)

The core idea behind looking at political issues as discourses is that in order to tackle a problem, we must first identify what the dominant historically and contextually situated understanding of that problem is. It thus builds on the interpretivist epistemological stance that the world is socially constructed and that our understanding of a social phenomenon can never be more than an interpretation (Marsh et al., 2018: 190). In terms of environmental politics, this means that what is politically discussed and what is actually happening to the environment are not necessarily the same thing (Hajer, 1995: 16). Environmental problems are constructed, and the commonly accepted representation of a problem shapes the actions we take to solve it. When environmental policy is made, it is thus based on a socially constructed definition of what the problem is; a definition which is informed by the environmental discourse (ibid.: 2).

Discourse analysis can be applied in different ways. In his own work, Hajer uses discourse analysis to “understand why a particular understanding of the environmental problem at some point gains dominance and is seen as authoritative, while other understandings are

(11)

8

discredited” (ibid.: 44). This is synonymous with the concept of hegemonic discourses, which was famously elaborated in Laclau and Mouffe’s book, ‘Hegemony and Socialist Strategy’

(2001). However, Hajer introduces the additional terms ‘discourse structuration’, ‘discourse institutionalisation’ and ‘dominant discourse’ (Hajer, 1993: 66), in order to conceptualise the steps and processes a discourse undergoes before reaching a hegemonic state (Mert, 2015: 40).

In this study, Hajer’s tools for identifying discourses are applied, and the theoretical framework of ADA is then used to try to understand the phenomenon of actors with different interests using the same motives. The concepts of discourse structuration and institutionalisation are also utilised to define discourse dominance.

3.2 The argumentative turn

Hajer adopts his use of the concept of discourse from Foucault, who saw it not in an instrumentalist way as something to be purposefully used by actors, but rather as something universal from which actors could not detach themselves or their interests. In other words, according to this view, actors constitute discourse and are themselves constituted by it (Hajer, 1995: 51). However, Hajer furthermore seeks to fill what he sees as two gaps in Foucault’s theory, regarding discourse formation and discourse permanence/change. To do this, he introduces the so-called argumentative turn in discourse analysis. According to Hajer, discourses are formed through social interaction as exchange of arguments. When analysing discourses, one should therefore focus both on the positions of the subject as well as its perceived counter-positions. In this view, it is when actors formulate how they see the world, and how they perceive others to see the world, that discourse is formed.

Furthermore, Hajer introduces the concepts of storylines and discourse coalitions.

Discourses are constantly reproduced and maintained through meaning production informed by historical contexts. In other words, discourses are not upheld by themselves, but by actors who reiterate the discourse they have been subjected to by referencing historic speech and acts.

According to ADA, this ‘routinised understanding’ is perpetuated by employing storylines;

condensed narratives which act as simplified and unifying understandings of phenomena (ibid.:

52–6). Groups of actors who share the same storylines, and thus together (re)produce a specific discourse through their shared language and practices, are in turn called discourse coalitions (Hajer, 2006: 70). Together these additions form Hajer’s approach called argumentative discourse analysis, or ADA. Discourse coalitions are useful when analysing the different arguments around the Preem debate. They can be used to locate and differentiate between the

(12)

9

different positions in the debate and the actors subscribing to them, using storylines as indicators of which coalition different actors belong to. As such, the three concepts of discourse, storylines and discourse coalitions are central to this thesis. They constitute the theoretical foundation on which the forthcoming analysis has been done, as well as the building blocks for the operationalisation which has been developed for and used in the study. In order to properly elucidate the definitions according to which they have been applied, the following subcategories explain these crucial concepts in more detail.

3.2.1 Discourse

First off, this study uses Hajer’s definition of discourse. In his own work, Hajer defines discourse as:

“an ensemble of ideas, concepts and categories through which meaning is given to social and physical phenomena, and which is produced and reproduced through an identifiable set of practices” (Hajer, 2006: 67)

This choice of definition is partly motivated by the methodology, but also because it sufficiently covers meaning and its (re)production through practices. The inclusion of practice in the definition is important since the study applies the Wittgensteinian view that linguistic expressions are contingent on their accompanied practices (ibid.: 70). The definition is minimal in that it states only the necessary parts to differentiate the concept, while still allowing the inclusion of all kinds of discourses as expressed through text, speech, art, culture etc.

Simultaneously it specifies the attributes of ideas, concepts, categories, meaning as well as practices (Schaffer, 2016: 124, 135). In the context of the Preem refinery case, it means that verbal (speeches, interviews), textual (news articles, statements) and practical (norms, protests) expressions of ideas and opinions on climate-change can all be regarded as part of a discourse.

Important to point out here is that even with a set definition of the concept of discourse, there are still many levels of discourse. For example, one could choose to study discourse on a national level between political parties in parliament or focus more on the local political discourse. One could also shift focus outside the traditional political sphere. This thesis aims to contribute to the field of political science, meaning that it seeks to study a political phenomenon.

However, in line with poststructuralist thinking, this study applies Hay’s broad definition of the political as a process rather than an arena (Hay, 2002: 72). This definition widens the meaning of the political and allows for the study of power outside the governing sector formally associated with politics and within the social practices of both public and private actors (ibid.:

(13)

10

73). This is crucial since the aim of the study is to understand how the concept of climate- change mitigation is used, not just by politicians in state settings, but also by non-state actors outside of the traditionally political arena. It could be argued that such a minimal definition is too broad, overlapping with other fields of study (Gerring, 2012: 117–21, 127–30). However, the political as a process does not mean that politics exists everywhere, only that it has the potential to (Hay, 2002: 73). Furthermore, it fits the poststructuralist view of power as context- shaping and ubiquitous (ibid.: 185, 188). In this particular study it means that spheres operated by NGOs as well as private individuals are considered possible spaces for political processes of power.

3.2.2 Storylines

Secondly, storylines are complex narratives condensed into simplified versions, which are built around metaphors. They typically consist of a short story about “how problems came to be (or came to be overcome) and what should therefore be done (or not done)” (Stevenson & Dryzek, 2012: 191). Storylines are used in political discourse when multiple actors need to mutually understand a complex topic that requires knowledge from various fields of study. Since party politicians, for example, cannot grasp the entire scientific complexity of topics such as climate change (and vice versa), they instead re-interpret the knowledge outside of their own expertise and fill in the gaps to make sense of the subject as a whole. This creates the opportunity for actors to talk about topics they do not fully understand as if they do. Moreover, Hajer argues that the resulting loss of meaning is what makes political change possible, because the simplified narrative creates a sense of mutual understanding and consensus between actors. The result is a phenomenon where actors fill in knowledge gaps to facilitate cross-disciplinary communication (Hajer, 1995: 60–5). An example of this is when politicians adapt difficult concepts such as carbon emissions trade or carbon capture and storage technology into their own storylines. The logic behind these concepts is complex and requires long, detailed explanation to convey in their entirety but is simplified to be used in political debates. In my analysis I have used the above definition made by Stevenson and Dryzek to facilitate a practical identification of storylines in the material. In practice, this means that I have looked for sets of problems and solutions as framed by different discourses, which together constitute a simplified story of something more complex.

(14)

11 3.2.3 Discourse coalitions

Lastly, discourse coalitions are an “ensemble of (1) a set of storylines; (2) the actors who utter those storylines; and (3) the practices in which this discursive activity is based” (Hajer, 1995:

65). Discourse coalitions can thus, in the most simplified way, be seen as a group of actors who

“develop and sustain a particular discourse, a particular way of talking and thinking about environmental politics” (ibid.: 13). However, despite this focus on which actors belong to which discourse, the concept of discourse coalitions is not actor-centric as such. Rather than being based on actors with similar interests, discourse coalitions are centred around storylines.

Furthermore, since discourse coalitions contain a set of multiple storylines, the same storyline can appear in more than one discourse coalition. Consequently, actors with different interests can gather around the same storyline, the storylines can in turn affect the interests of these actors, and the same actor can belong to more than one discourse coalition. The reasons why actors adhere to certain storylines differ, but as a structural explanation Hajer points to the concept of discursive affinity. Basically, this means that actors choose a certain storyline because it ‘sounds right’ based on internalised logic. Despite not knowing specific details about the subject, actors choose them because they ‘make sense’. Or as Hajer puts it; “Separate elements might have similar cognitive or discursive structure which suggests that they belong together” (Hajer, 1995: 66–8). Actors thus latch on to these storylines for different reasons and using different logics, but in reiterating them they together promote the messages of the discourse to which they belong.

3.2.3 Summary

As described in the introduction, the Preem refinery debate holds several conflicting discourses which contain ideas about problems, solutions, assumptions etc. The groups of actors that subscribe to and (re)produce each of these discourses can thus be called discourse coalitions.

The actors in these coalitions are not necessarily brought together by shared interests or common goals. Rather, they simply use the same storylines to roughly communicate their ideas on a complicated issue. The continued use of these storylines is what creates and sustains the discourse.

(15)

12

4. Methodology, research design &

material

4.1 Argumentative discourse analysis

The main method of analysis used in this study has been ADA as explained by Hajer (2006:

72), which is a form of interpretive policy analysis (Glynos et al., 2009: 21–2). Although this study does not focus on policy texts, the method has been shown to apply to discourses “geared toward the realm of action (rather than merely debate) and indeed policy change” (Epstein, 2008: 95). In light of the research questions and the theoretical framework, I have chosen a qualitative method. The aim of the study is to understand how the use of language relates to a given context, and to critically analyse how language and practices shape the understanding of environmental issues in different discourses. As such, the main focus lies not on frequency of certain variables in a large collection of material, which would have favoured a quantitative method (Esaiasson et al., 2017: 198–9). A qualitative method makes it easier to catch underlying or unspoken meanings in texts that are hard to find by fitting variables into coding schemes (ibid.: 211). Since the goal is to know what different actors think and say about the Preem refinery debate, a thorough reading of the text is needed to catch unpredictable content.

Moreover, out of numerous qualitative methods I have specifically chosen discourse analysis.

The purpose of this study is not only to describe the content of the studied texts which could be done with a content analysis. Neither is the focus on the causal mechanisms which led to the trial, or to the development of different discourses, to necessitate the use of methods such as process tracing (Beach & Pedersen, 2013: 1–2). Unlike these methods, discourse analysis enables the study of meaning in language and practice within particular contexts (Gee &

Handford, 2012: 5). For the political scientist, discourse analysis is a method for both describing and critically analysing the constitutive properties of language and practice on power (Bäckstrand & Lövbrand, 2006: 51). Thus, for analysing the meanings produced by different actors in the Preem refinery debate and their impact on policy, I deem discourse analysis to be an appropriate and useful tool.

In this study the methods and theoretical framework are very closely connected. The chosen conceptual framework of Hajer’s discourse coalitions and storylines build on ADA,

(16)

13

which will also be used as method. The choice of ADA is motivated by the purpose of the study.

Importantly, the ‘argumentation’ part in ADA signifies the method’s central focus on discursive interplay between actors, who “react to one another and thus produce meaning interactively”

(Hajer, 2006: 72). This emphasis fits this case due to the argumentative nature of the Preem refinery debate; in the debate, actors produce meaning in reaction to the ideas and actions of other actors. In other words, since the purpose here is to study the discourses around the Preem refinery debate, and since the material consists of societal and political debate around a contested issue, i.e. actors’ reactions to other actors’ speech and practices, the argumentative approach is a suitable choice. Additionally, this focus on interactions between actors makes the approach empirically grounded without losing the structural aspect of discourse theory. By looking at actors’ role in discourse formation through argumentation, the approach lies somewhere between an individualistic and structural view on discourse, which according to Stevenson and Dryzek is the ‘most defensible position’ to take (2012: 191). This focus allows the analysis to be empirically driven, by deriving discourses from storylines expressed by actors in the material (Zannakis, 2015: 222–3).

4.2 Case & research design

This study utilises a case study design to analyse a particular event in time, namely the debate around the Preem refinery expansion between 2019 and 2020. This case in particular is interesting due to the puzzling phenomenon where actors on different sides of the debate very clearly emphasise the same need to combat climate change in their arguments. Furthermore, the Preem refinery case is well known in Sweden as it sparked much controversy. As such, there was plenty of research material to pick from, as well as many different actors who had strong and conflicting opinions, providing suitable interview subjects. The case is well documented with utterances in newspapers, and websites of NGOs and government agencies, making primary data sources easily accessible. This also facilitated the identification of key actors to contact for interviewing. Since the example of the Preem refinery is already well delimited, it serves as the single case for this study. Not only are the different arguments presented in the Preem debate as solutions to climate-change interesting examples of how discourses form the meaning of a concept; the case is also representative of how the concept of climate-change is used by different actors for different purposes in a broader national climate debate. It thus follows Foucault’s theory of ‘microphysics of power’ by analysing micro level phenomena in order to explain macro level processes (Glynos et al., 2009: 35). To capture a representative

(17)

14

part of the debate the timespan is set from when Preem’s initial permission to expand their refinery was appealed in 2019, to just before Preem cancelled their plans and withdrew their application.

4.2.2 Preem refinery case timeline

In December 2016, Preem applied for permission to continue and expand operations in Lysekil according to the so called ROCC-project. The declared purpose of the ROCC – or Residue Oil Conversion Complex – project was to reduce the amount of sulphur and metals in bunker fuel, and to transform the fuel into petrol and diesel (Preem, 11: 2019a). This came as a response to new rules from the UN’s International Maritime Organisation (IMO) which significantly reduced the allowed amount of sulphur oxide emissions from ships starting in January 2020 (IMO, 2020). The application was approved in 2018 by the Land and Environment Court (LEC) (SEPA, 2020). However, this decision was later appealed by several actors, including the Swedish Society for Nature Conservation (SSNC), due to its expected GHG emissions (SSNC, 2018), which led to the LEC announcing that it would review the case in June 2019. During the same month, the Swedish Environmental Protection Agency (SEPA) sent a statement of opinion about the case to the government, leading to the government in August 2019 officially taking over the right to decide on the permissibility of the expansion. In March 2020, the SLEC held a legal hearing in order to provide the government with a statement about the permissibility of the expansion. In June 2020, the SLEC announced that they deemed the expansion to be permissible, leaving the final decision to the government (SEPA, 2020). However, Preem withdrew their application in late September 2020 before the government had reached an agreement (Preem, 2020a).

Fig. 1 – Preem refinery case timeline

4.2.2 Legal background & climate goals

According to Sweden’s environmental code, any project concerning increased GHG emissions is part of the EU Emissions Trading System (ETS). This means that such a project must be

December 2016

Application sent November 2018

LEC approval 2018-2019

Appeals June 2019

Review announced

August 2019

Government takeover

June 2020

SLEC approval

June- September

Government deliberation

September 2020

Preem withdraws

(18)

15

authorised and cannot be placed under conditions of restrictions on emissions (Miljöbalken:

chap. 16, 2§). However, it also states that the Swedish government may review the permissibility of a specific case if deemed to be of significant magnitude or be of an interfering kind against the goals of the legal code (ibid: chap. 17, 3§). These goals include the preservation and protection of human health, the environment, nature and cultural environments and biodiversity (ibid.: chap. 1, 1§). Moreover, a government agency that learns about such a case must inform the government about it (ibid.: chap. 17, 5§) ‘in good time’ before the case enters the main hearing or a decision on permissibility is taken (Förordning om miljöfarlig verksamhet och hälsoskydd: 4§). The meaning of ‘in good time’ has been a point of much contestation in the Preem refinery debate. Sweden also introduced a new climate-political framework in 2018 which legally binds the government to work toward the national long-term goal of zero net- emissions by 2045, with stage goals in 2020 and 2030 (SEPA, 2020). This further incentivises the government to intervene in cases with expected large emissions increase.

4.2.3 Research design

As indicated by the research questions, the study is descriptive in that it seeks to identify the different discourses present in the case, as well as the dominant discourse (Esaiasson et al., 2017: 37). However, it also aims to reach a theoretical understanding of the aforementioned puzzling phenomenon. The study has a case-driven approach, meaning that the adopted theoretical framework and methodology are used as tools to explain the case (ibid.: 42). This corresponds with the general problem-driven approach of interpretive policy analysis (Glynos et al., 2006: 24). In line with a discourse analytical approach in general and interpretive policy analysis in particular, focus lies on meaning construction and apprehension in discourses, and a hermeneutic rather than causal approach to policy understanding (ibid.: 6, 21). The empirical focus lies on the language and practices used by the different actors in the debate to ascribe, produce and reproduce meaning, following the aforementioned definition of discourse. Hajer’s emphasis on interplay as argumentation, as well as the context and setting of discourses, are heeded (Hajer, 2006: 72). In practice this means that different forms of expressions of and arguments about ideas and opinions regarding the Preem refinery and climate-change as a political issue, as well as the context in which these are expressed, constitute the empirical basis for the study.

(19)

16

4.3 Material

4.3.1 Data selection and delimitations

Due to the given definition of discourse the scope of possible data types is vast, including everything from court documents to speeches to smartphone recordings. To make the data manageable and an analysis temporally feasible, I have delimited the use to three types of data:

newspaper articles, official statements, and expert interviews. The data must be a ‘closed universe’ and cover the expressions of all relevant actors in the debate (Del Rosso, 2018: 6), which in this case can roughly be categorised as NGOs, Preem representatives, government officials, politicians and civilians/protesters. I argue that the three aforementioned data types collected during the given time span are delimited enough while still fulfilling this criterion.

Firstly, as Hajer argues, newspapers provide a good starting ground for desk research, i.e. initial reading to get familiar with the subject (Hajer, 2006: 73). Furthermore, doing a proper analysis of newspaper material is a good way of studying the ideas and opinions of private individuals who otherwise lack a platform. Newspapers often conduct interviews with ordinary people during protests to measure the public consensus, and provide platforms for politicians with debate articles. Newspapers also have the benefit of documenting actions besides speech, which is useful for the contextualisation of discourses. Secondly, official statements done by NGOs, Preem and government agencies provide primary sources of the standpoints of more formal actors. For NGOs and Preem, this includes press releases and other public information on websites. For government agencies, it means notification reports and statements of opinion to the government. Thirdly, expert interviews allow for a more in-depth analysis of the ideas and opinions of selected key actors. The context of an interview often reveals parts of the discourse not expressed in the strictly revised and scrutinised official statements (Brinkmann & Kvale, 2015: 260). Interviews with experts furthermore provide an overview of the development of the issue over time, as well as academically informed opinions on the debate (Hajer, 2006: 73).

For the newspaper material, a narrow timeframe was chosen after initial readings. The goal was to find articles about the Preem refinery expansion specifically, which reflected all sides of the debate. Too wide a timeframe – say two years – would risk catching too many articles unrelated to the expansion when the debate was not ongoing. Since it is temporally unfeasible to analyse every article about Preem in two years, they would have to be chosen randomly, increasing the risk of missing some sides of the debate. To mitigate these problems, I chose to focus on a timeframe when the debate was particularly heated, and analyse all articles containing the word ‘Preem’ published between February 20 and March 10. On March 10, the

(20)

17

SLEC started its trial of the case. During the weeks leading up to the trial, a myriad of actors saw their chance to express their opinions on the case in the media, including politicians, Preem representatives, townspeople and activists. The odds of covering all sides were therefore favourable. I chose five newspapers, two local and three nation-wide, to get a varied coverage.

The official statements were delimited from 2019 to when Preem withdrew their application. This was to cover all existing utterings from when the refinery expansion’s initial approval first got appealed to the end of the case with Preem’s cancellation. Four key institutional actors were chosen: Greenpeace, the SSNC, the SEPA, and Preem. The inclusion of Preem was given. As for the others, the included actors were the most active and visible in the chosen material types. Greenpeace and the SSNC have been the most prolific and vocal environmental NGOs in the debate, and the SEPA played a crucial role in alerting the government about the expansion. All press releases and official statements about the expansion from both Preem and the NGOs, as well as the two referrals written by the SEPA, were included.

The coverage is admittedly skewed toward actors who could be expected to be against expansion, but this reflects the reality of the debate. There simply were no other large institutional actors with official statements supporting expansion. Statements from other actors on Preem’s side were instead covered in the newspaper and interview material.

Interviews were delimited by focusing on a few select academics as well as representatives of chosen key players in the debate. The academics consisted of three university professors specialising in environmental politics, environmental law, and climate science respectively. The representatives consisted of a legal expert from the SSNC and a campaign leader from Greenpeace. Unfortunately, Preem did not answer my invitations to do an interview. Neither did the labour union representing their workers. To represent the business/industry side, I instead interviewed a legal expert from the Confederation of Swedish Enterprise (CSE).

4.3.2 Data collection

The newspaper data was collected via a database of all Swedish editorial media outlets, including local and national newspapers. To filter the material, only articles containing the search word ‘Preem*’ were used. This resulted in 54 articles – mainly debate articles and opinion pieces, but also reports of protests with interviews. Similarly, official statements concerning the refinery expansion were collected from the homepages of the previously described actors.

(21)

18

The interviews were conducted via video meetings which were recorded for audio and later transcribed. Each interview took just over 30 minutes. The goal of these interviews was not to capture the innermost thoughts and feelings of the individual, but to get a picture of the kind of arguments used by the organisation the interviewee represented. I therefore chose a formal written style, ignoring things like frequent repetitions of phrases, conversational expressions or emotional signals (Brinkmann & Kvale, 2015: 207). This was to put focus on what arguments the interviewee was actually presenting, and not social cues revealing what the individual was feeling in that moment. For the format I used a semi structured interview, preparing a set of suggested questions (Ibid.: 156), which can be found in appendix 1 and 2.

During the interview, I used these questions as a general guide rather than a script, letting the flow of conversation guide the direction of the interview and asking follow-up questions when needed. This gave the interviewee a chance to respond spontaneously without being thematically steered by direct questions (ibid.: 161).

4.4 Operationalisation

The application of ADA on the material was operationalised in three steps. First, a set of guiding questions to ask the text were formulated, tested and tweaked until satisfaction. Next, the texts were thoroughly read and colour-coded based on these questions, and then sorted into a matrix based on discourse coalitions and storylines. Lastly, the findings were analysed using ADA.

Although Hajer provides excellent theoretical justifications for his approach, he does not offer much in terms of practical application and operationalisation of his theories and key concepts.

In order to develop a structured and transparent operationalisation, I thus had to turn to other sources for inspiration. The result outlined in fig. 2 was inspired by a combination of different articles and academic literature on discourse theory and qualitative social science methods in general.

(22)

19 4.4.1 Guiding questions

When doing any textual analysis, it is important to develop concrete analytical tools which help guide the reading and sorting of the material. One way of doing this is by formulating specific questions, informed by the research question, to ask the text while reading and sort the answers into analysable categories (Esaiasson et al., 2018: 216). Hajer’s ADA puts significant emphasis on argumentation. This analysis does the same by focusing on debate articles and official statements where actors express their positions and arguments. Not only the actors’ own views, but also what they consider their opponents views to be, has been taken into account. The underlying research question here is ‘which storylines and discourse coalitions were present in the Preem refinery debate?’. As previously written, storylines can be summarised as a short story about how problems came to be and how they should be overcome. Hence, problem representation is key here. When reading each text, the following four questions have thus been asked:

1. How is the problem represented?

2. What is the proposed solution?

3. What are the underlying assumptions?

Step 3 Analysis

Analyse findings using ADA

Step 2

Coding & categorising

Colour-code text

Sort and record findings

Find themes

Sort themes into storylines

& discourses

Fit findings into a matrix

Step 1 Preparation

Establish research questions Formulate guiding

questions

Test reading

Review guiding questions

Fig. 2 – Operationalisation steps

(23)

20

4. How does the text perceive the position of its opponents?

The first question comes from the theoretical position that problems are socially constructed.

Hence, how a problem is represented in text, speech and action has significant consequences on politics. The second question is connected to the first and is included by the logic that what is proposed as a solution indicates what the perceived problem is. The answers to these two initial questions make up the storylines, in accordance with our previous definition. The third question fills the function of including important clues as to how and why actors hold certain opinions. If an actor assumes that the expansion will lead to increased GHG emissions, it explains why they want to prohibit it. The fourth question is a function of ADA and the importance it puts on the perception of counter arguments. Together, these four questions constitute the discourses, as the (re)production of ideas (of problems and solutions) through arguments and the perception of counter arguments within a certain practical context. The texts were colour-coded based on these questions, and all problem representations, solutions, assumptions and opponent-perceptions were then recorded in an excel sheet.

4.4.2 Quantification

Due to the nature of a qualitative discourse analysis, it is difficult to accurately quantify results.

When looking for answers to the above guiding questions, I have not searched for individual words but rather paragraphs and sentences that convey similar meanings. As such, no two answers are alike, which makes it impossible to count them using word-search. Nevertheless, it can be valuable to get a sense of how frequently a type of expression appears even in a discourse analysis (Esaiasson et al., 2017: 232). It is especially useful when looking for the dominant discourse. When reading the texts and recording the answers, one gets a sense of how many types of expressions there are in each discourse, and how many instances there are of each type. I have thus kept track of the approximate number of times certain types of expression show up in the text, in order to measure the frequency of each discourse. The results have been categorised in the following way:

Table 1 – Quantification chart

Frequency Definition

Low - Few different types of expressions, and;

- Few examples of each type of expression.

Medium - Few types but many examples, or;

- Many types but few examples, or;

- Medium amount of both types and examples.

(24)

21

High - Many different types of expressions, and;

- Many examples of each type of expression.

4.4.3 Matrix

To sort the findings, a matrix in which to categorise the text was developed abductively. First, a provisional matrix was deductively constructed based on preliminary readings and the concepts used in Hajer’s argumentative discourse analysis. It was also inspired by operationalisations done by Zannakis (2015: 227), Stevenson and Dryzek (2012: 193), and Bäckstrand and Lövbrand (2006: 68) in particular, due to the similarities between their research- designs and mine. The matrix was then modified inductively after getting more acquainted with the material. The resulting matrix included four discourses categorised after the following seven variables:

• Position: Indicates whether the discourse was for or against the expansion of the Preem refinery in Lysekil.

• Core themes/assumption: Summarises what core themes the discourse promotes when arguing its case, and which assumptions it rests its storylines upon. This is what the logic of the storylines is built upon.

• Key storylines: Lists the key storylines that are used and promoted by the actors in the discourse coalitions.

• Basis of legitimacy: States the basis of the main claim that the discourse officially used to legitimise its taken position, i.e. moral, economic etc. Note that the basis of legitimacy does not necessarily correspond with the motives of the agents who promote the discourse.

• Function: Indicates the function of the discourse in the Preem refinery debate, i.e. what is it meant to accomplish?

• Key agents: Lists the key agents in the discourse coalition.

• Frequency: Indicates how frequent – or visible – the discourse was in the text, sorted into low, medium and high.

(25)

22

4.5 Validity

In terms of validity, i.e. whether the proposed method will actually answer the research questions asked (Esaiasson et al., 2017: 59), the difficulties are hermeneutic in nature. When identifying different discourse coalitions and discerning what their motivations behind climate change mitigation are, there is always a degree of interpretation done by the researcher. This is because one can never be sure that an actor’s words or actions always correspond with what they feel, think or indeed understand. In order to increase validity, it is therefore important to be transparent with how one has interpreted the material. As with all interpretive studies, the risk with discourse analysis lies in the interpreter’s subjectivity. Whenever one interprets discourse, one does so a human being with existing prejudices (Brinkmann et al., 2014: 21–2).

From a positivist viewpoint such a focus on interpretation can be criticised as subjective, vague or arbitrary, and in the way of the search for ‘truth’ (Schaffer, 2016: 5). Instead of shying away from interpretation, however, when doing any analysis based on hermeneutics and understanding, it is important to be reflexive and open about one's pre-judices (Brinkmann et al., 2014: 22). In other words, it is vital to be explicit with how one interprets the meaning of the studied discourses to mitigate subjectivity (Blau, 2017: 257). In terms of reliability, i.e. the absence of human errors (Esaiasson et al., 2017: 64), the challenge lies in the collection and coding of data. Specific care has been taken in preparation and execution of data collection so to avoid mistakes, as well as during coding so not to miss any vital data points.

5. Empirical analysis & results

After analysing the material, I identified four distinct discourses in the Preem refinery debate.

There is some overlap between them, as will be made evident in the following sections, but all four have a unique core message relayed through a set of storylines that differentiates them from the others. Before discussing my results, it is important to note that there were a few positions and arguments found in the texts that did not fit into any of these four. For example, there was one debate article arguing for the positive effects of increased GHG emissions that

(26)

23

contained clear elements of climate-change scepticism. However, since these cases were so scarce and supported by so few people, they did not warrant the inclusion of a separate discourse. If we follow Hajer’s definition, it becomes clear that one or two people saying and doing something once does not make a discourse.

5.1 Reformist discourse

Table 2 – Reformist discourse

Position Against

Core themes/

assumptions

Climate urgency, assumption of increased emissions following expansion, growth cautious, EU ETS critical, distrust in Preem

Key storylines Fossil fuel investment, dying industry, government responsibility

Opponent Jobs, due process, trust in Preem, trust in EU ETS, ecomodernist, Swedish superiority, interest driven

Legitimacy Moral/idealistic

Function To stop climate change out of moral obligation

Key agents Greenpeace, SSNC, academics, Green Party, Lysekil Party, protesters, citizens

Frequency High

The first discourse had a clear anti-expansion message, representing the problem as one of increased emissions, arguing that an expansion would be bad for both climate and environment.

This discourse had the strongest environmentalist message of all, being sceptical toward purely growth-driven climate change mitigation and existing systems. It thus promoted a reform of existing institutions in order to keep up with rapidly increased global warming. However, it was not fully system critical or radical, as it still acknowledged the importance of growth and innovation in fighting climate change. The reformist discourse centred around three main storylines which can be said to comprise the core of its arguments against the Preem expansion.

5.1.1 Preem’s expansion is a fossil fuel investment. We must stop it to reduce emissions.

The main argument of the reformist discourse was that the refinery expansion would be bad for the climate. This rested on a set of assumptions, the first being that the expansion was mainly done in order to produce more petrol and diesel, making it a fossil fuel investment. The discourse did contain critique against the expected local environmental pollution, but the main focus undoubtedly was on carbon dioxide. Thus, the first storyline represented the problem as

(27)

24

one of increased GHG emissions – both during fuel production and as a consequence of making more fuel to be consumed – and proposed a cancellation of the expansion plans as the solution.

“Despite the severe state of affairs Preem has gotten permission at the first instance to expand the oil refinery Preemraff in Lysekil, an expansion which means that their carbon dioxide emissions will increase by 1 million tons per year. The emissions increase would make Preemraff the single largest emitting facility of carbon dioxide in Sweden, and the permit would in addition apply for an indefinite period of time.” SSNC, in: Naturskyddsföreningen, 2020a

The second assumption was that Preem’s intention of increasing their production of biofuels and using carbon capture storage (CCS) technology if allowed to expand was problematic.

Different claims about this were made, including that CCS-technology was unreliable and that it, combined with biofuel production, was not enough to reduce emissions. The main claim, however, was that Preem’s intentions were empty promises. This was supported by the fact that these stated intentions were not part of the official application and therefore not legally binding.

Instead, Preem had said that they would go through with it only if financially viable.

“Preem’s conditional goal of reduced carbon dioxide emissions includes capture and storage of carbon dioxide using so called CCS/BECCS-technology. This is very costly and still untested on a large scale. […] carbon dioxide capture-and-storage should not be used to motivate new carbon dioxide emissions, as for example with a fossil expansion.” Wennergren – Lysekil Party, 2020

“In Preem’s application there is however no legally binding text on reduced emissions or transition to biofuel. Preem themselves have written that their climate promises will only be fulfilled ‘on the precondition that an environmental permit is acquired, and that it is economically feasible’. Preem first want permission to emit and extra 1 million tons carbon dioxide per year before they consider the profitability of keeping their promises.”

Sweden’s environmental NGOs, 2020a

It was also clear that this argument presupposed that those who were for expansion wrongfully assumed it was a pure climate investment initiated by Preem, and that the company had actively lied about this in their advertising campaign. This assumed that the real reason behind expansion was the new regulation on bunker oil which had forced the company to find new uses for this product.

(28)

25

“There has been a lot of greenwashing; they say they will do one thing, showing Gunde Svan among pine branches in the forest saying ‘look, this will become your biofuel’, but it’s only a small percentage that comes from Swedish forests in their energy-mix. […]

you see that they very clearly try to pretend that they are doing a biofuel investment, when it’s actually a fossil fuel expansion. But even if it were true that they would invest in biofuels it wouldn’t have helped. It is a false climate solution.”

Schlyter – Campaigner Greenpeace, 2020

The given reason why increased GHG emissions were bad was a moral obligation to prevent global warming due to its expected effects on people and environment. The urgency of the climate crisis was often stressed, claiming that it is too late to push solutions into the future.

Furthermore, the Paris Agreement’s (PA) goal of a maximum temperature increase of 1.5°C, as well as national climate goals, were often cited as reasons to stop expansion.

“In this situation Preem is trying to increase their emissions by 60 percent compared with today. That would make Preem the largest emitter in the entire country. The project would in reality make it impossible for Sweden to reach its climate goals and the Paris Agreement.” Sweden’s environmental NGOs, 2020a

Finally, the storyline contained both the argument that an expansion would lead to increased national GHG emissions, which was bad enough in and of itself, but also that a failure to expand would not lead to increased global emissions. This was a response to the claim that the EU ETS would make sure that the same emission rights would go to someone else, and that Sweden was the best country in which to build a refinery.

“Preem means that since the facility is part of the EU’s emission trade, increased emissions in Sweden don’t mean that emissions will increase globally since there is a common ceiling in the EU. That is only true in theory, however. In practice, the surplus of emission rights is so big that you today can increase emission without someone else having to reduce theirs.” SSNC, in: Naturskyddsföreningen, 2020b

5.1.2 Oil refining is a dying industry. Better to invest in sustainable jobs.

The second storyline was an answer to the common argument from the opposing side that an expansion would lead to much needed jobs for people living in Lysekil. The reformist counter- argument to this was that oil-dependent jobs are going to disappear sooner rather than later due to our societies gradually shifting toward renewable energy sources. The solution, therefore, was to invest in jobs of the future.

(29)

26

“We understand that there is a fear for the jobs if this project doesn’t happen. But this whole question is about climate adaption and reduction of carbon dioxide emissions and to put society to the test regarding the creation of other alternative jobs.” Protester, in:

Andersson, 2020a

The promotion of jobs in renewable energy also indicated a trust in technological solutions to climate-change. Growth and investment ought thus to be steered toward climate friendly alternatives, but not disregarded as a solution altogether.

“…we also believe that there is technological potential to solve the climate problems, it has existed for a long time. It’s just a question of making sure it is used.” Alarik – Legal expert SSNC, 2020

5.1.3 Sweden’s climate law appoints the Government as responsible. They should decide the case.

The third storyline centred around the fact that the government took over the case after a request from the SEPA. This had been criticised by those in favour of expansion, and so this was the reformist discourse’s response. The problem was again represented as one of increased GHG emissions, and the argument was that since Sweden’s new climate law places the government as responsible for limiting emissions, this was a case for them.

“The new climate law means that the government is forced to take responsibility for emission sources of this magnitude.” SSNC, in: Naturskyddsföreningen, 2020b

”The government is ultimately responsible for us living up to our climate goals and an expansion of Preemraff would significantly reduce our chances of reaching those goals.”

Greenpeace Sweden, 2019

Furthermore, this storyline included the assumption that since Sweden is part of the EU ETS, Swedish courts are not allowed to restrict or place conditions on a company based on its GHG emissions. This was framed as a flaw in the EU ETS and used to argue for increased involvement of governments in national climate cases. Because of this perceived limitation, the government was deemed better equipped to reach a proper decision, i.e. to take emissions into due consideration.

“The court is not allowed to take into consideration carbon dioxide emissions due to a rule in the ETS-legal code. But then, the court overinterpreted it. In other courts they have, among other things, interpreted that rule more beneficially and the court has actually been able to take emissions into consideration. […] It’s definitely the case that

The discursive battle for Lysekil