Business in Society or Business and Society: The Construction of Business–Society Relations in Responsibility Reports from a Critical Discursive Perspective

In this article, we analyse the discursive construction of business–society relations in Finnish businesses’ social and environmental responsibility reports. Drawing on critical discourse analysis, we examine how these discursive constructions maintain and reproduce various interests and societal conditions as a precondition of corporate social responsibility (CSR). Our study contributes to the recent discussion on discursive struggles in business–society relations and the role various interests play in this struggle. We find that not only are power asymmetries between actors veiled through the universalization of interests, but reporting can also be seen as a communicative action that provides a right to define the role of societal actors for the achievement of CSR. We suggest that the discursive struggle over whose interests dominate, and how they dominate, shapes the role of social and environmental reporting as a social practice.


Introduction
How can we increase social welfare, enjoy the benefits of technical and economic development, and at the same time preserve and care for the fragile natural environment? These questions have very recently been at the heart of the discussion of the relationship between business and society. The business-society relationship has been examined and theorized upon for decades by scholars aiming to understand and answer a question: what is the role of a firm and what are the associated responsibilities? The role of businesses and their impact on and interest in environmental, social and economic issues, and responsibilities has further been conceptualized as corporate social responsibility (CSR), which has been proposed as a theory of the corporation (e.g. Buchholz 1991, Carroll 1993. However, CSR originates in and is related to several theoretical fields, all of which have a slightly different view of what the phenomenon is about and what is or should be the focus of analysis in CSR-oriented studies (for reviews, see e.g. Gray et al. 1995, Garriga & Melé 2004, Banerjee 2007, Secchi 2007, Maon et al. 2010.
CSR and its conceptualization are today often based on how the institutional environment and government policies affect CSR and thus reflect business-society relations (see e.g. Maignan & Ralston 2002, Doh & Guay 2006, Albareda et al. 2007, Matten & Moon 2008, Steurer et al. 2012. In order to understand how CSR reflects cultural conditions and arrangements, many empirical studies have focused on corporate social and environmental reporting (SER) and the contents of the reports and businesses' definitions of their responsibility (e.g. Adams et al. 1998, Hartman et al. 2007). Traditionally, reporting has been viewed as a communication channel through which corporations disseminate information about relevant social and environmental issues, which may also be understood as a way of seeking legitimacy (see Gray et al. 1995, Deegan 2002. However, according to some scholars, corporate SER not only reflects cultural conditions, but actively promotes their (re)construction. Critically oriented researchers in particular have paid attention to the language used in SER and how it can reflect societal conditions, struggles and interests (see e.g. Livesey 2001, Laine 2009, Spence 2009). Previous literature has acknowledged the role of SER as a channel available to businesses to address social and global challenges, such as sustainable development (Tregidga & Milne 2006, Tregidga et al. 2008. The response to these challenges further depends on how the phenomenon is constituted.
Our aim in this study is to offer a better understanding of the discursive strategies that construct business-society relations in the SER of Finnish businesses. More specifically, we examine how these discursive constructions maintain and reproduce various interests and societal conditions as a precursor of CSR. In order to accomplish our task, we first draw on the relational perspective in CSR theories (Secchi 2007). Second, we adopt a critical perspective to understand the ways in which businesses are also political actors (Mitchell 1989; see also Matten & Crane 2005, Palazzo & Scherer 2006, Scherer & Palazzo 2007, Matten & Moon 2008, whose own interests affect their definition of CSR and the relationship between business and society. Our research aim demands that we pay attention to how reporting language is involved in shaping the roles and responsibilities of businesses because businesses can take action to achieve power. We argue for the importance of discursive approaches and critical discourse analysis (CDA) in particular in order to understand how certain discursive expressions associate with dis-cursive struggles in business-society relations. We apply a CDA approach using the ideas of Fairclough (1992Fairclough ( , 2001Fairclough ( , 2003 concerning discourse as a social practice. For our empirical data, we selected 24 Finnish businesses' social or environmental reports published in either 2005 or 2007. What makes the sociocultural context interesting is that Nordic countries have traditionally been taken as an example of countries that not only rank well in global competitiveness rankings but are greatly concerned about the natural environment (Bernes 1993). According to Grennes (2003: 13), the 'Scandinavian model promotes long-term ties between owners, managers, workers, and society, where the role of the company includes promotion of goals of society at large'. We view the Scandinavian model as having traditionally emphasized a 'business in society' view according to which societies need to impose regulation on businesses (Wood 1991). For example, the competitive conditions in Finland used to be highly regulated (Tainio 2006).
The current research contributes to the contemporary discussion on discursive struggles in businesssociety relations and the role various interests play in that struggle (see Levy & Egan 2003, Banerjee 2007, Spence 2007, Tregidga et al. 2008. Through the use of the two main discursive strategies in the data, centralizing and decentralizing strategies, businesses are conceptualized either as a part of society (business in society) or as two separate environments (business and society). However, these constructions are not mutually exclusive. We further show how the discursive strategies play an important role in determining whose interests constitute CSR. Not only is reporting practice a societal legitimacy quest in which power asymmetries are veiled by universalizing interests using 'cooperative' and 'balancing' language, but it can also be seen as a communicative action that provides a right to define the role of other societal actors for the achievement of CSR. We therefore suggest that SER is a social practice shaped by discursive struggle over whose interests dominate and how they dominate.
The article proceeds as follows: in the next section, we will discuss the literature of CSR, focusing particularly on the importance of a critical approach to the business-society relationship. We follow our critical discursive approach with a description of the data and analysis. The study concludes with a discussion of its contribution and limitations.

Theoretical framework
Theoretical approaches to CSR The growth in CSR literature seems to have encouraged several researchers to pay attention to the variety of theories used to describe and explain the phenomenon. One of the most widely cited has been developed by Garriga and Melé (2004) who divide CSR-related theories into four different groups. The first group is called instrumental theories that emphasize CSR merely as a means to the end of profits. CSR as understood in this way has been quite widely criticized on the grounds that it would promote just business benefits rather than social well-being (Banerjee 2007, Gond et al. 2009). The second group is called political theories in which the social power of the corporation is emphasized. There has been increased interest in these theories recently because of the research on the politicization of corporations (Matten & Crane 2005, Scherer & Palazzo 2007. The third group, called integrative theories, includes theories that argue that business depends on society for its continuity and growth and even for the existence of business itself. Integrative theories include approaches that are quite widely used to illustrate the management of CSR and related phenomena. The fourth group, ethical theories, emphasizes the ethical values embedded in the business-society relationship. CSR in this view is primarily a question of ethics and the ethical management of business. Secchi (2007) criticizes Garriga and Mele's approach as unclear, saying that there is no clear reason why, for example, a political theory might not also be an instrumental one. In addition, the groups contain only theories that directly address social responsibility, therefore limiting the depth of analysis. By providing an alternative view, Secchi (2007) proposes that CSR should be examined in terms of how certain theories and approaches define relations between corporations and society and, more specifically, where responsibility is allocated. This proposition is concretized in utilitarian, managerial and relational theories of CSR. The utilitarian stream of thought sees business as part of a wider mechanism, the economic system, and businesses' self-interest as the economic system's driving force. The managerial stream approaches responsibility from inside the firm (internal perspective) and is a kind of counterproposal to the first stream. Relational approaches consider the relationships between the firm and society and direct attention to businesses interacting with the economic system (Secchi 2007). One of the most critical ideas in this third approach is the subgroup of business/society theories. Studies placed in this field often try to analyse society and the role organizations play in it. In these approaches, the constitution of CSR can be associated with power relations: the interests that define CSR. We elaborate on this view next.

Business in society
This term has been used to describe the responsibility of business to society. The terminology has been drawn from functionalism and views of conflict. These emphasize the regulation that societies need to impose on businesses. According to Wood (1991), functionalists view business as a subsidiary of society, created to serve economic functions in an institutional division of labour that, left to operate unimpeded, would result in societal stability and well-being. Functionalism favours regulation, guaranteeing competitive conditions rather than regulation designed to achieve specific social objectives. The traditional idea of business legitimacy is that business exists and acts by permission of society and society's expectations, as defined through public policy processes, emphasizes this view (Wood 1991). Indeed, in many institutional environments, particularly in Scandinavia, there has not been a strong confrontation between the interests in businesssociety relations (see Näsi 1995, Matten & Moon 2008. Businesses and other organizations have been understood to interact with society because they are part of it and are in partnership with other focal actors -emphasizing the view we call business in society. We suggest the business in society approach can be linked to the idea of implicit CSR (Matten & Moon 2008), which is a particularly European and Scandinavian phenomenon: CSR describes all formal and informal institutions of a society which assign and define the extent of corporate responsibility for the interests of an entire society (Matten & Moon 2008). More specifically, in the implicit system, many of the responsibilities are legal requirements. This precondition contradicts the US-driven CSR literature from where the majority of the CSR literature has originated. The distinctive elements of this kind of 'European CSR' are the inclusion of regulated industrial relations, labour law and corporate governance, for example (Preuss et al. 2009). Wood (1991) describes the term 'business and society' as a relatively new phrase in which the figurative chain of command is not clear. According to this view, Wood (1991) suggests that businesses are seen as a collection of independent actors, each with certain interests in abiding by the expectations of society and their stakeholders. Regulation can be seen either as illegitimate interference in corporate autonomy or as a strategic tool for obtaining corporate objectives, or both, depending on the context. In addition to this, we suggest that the business and society view is very much present in explicit CSR (Matten & Moon 2008). It is a reflection of the United States having been more reliant on marketbased forms of ownership, and is more embedded in institutions and traditions that consist of voluntary corporate policies, programmes and strategies (Matten & Moon 2008). Wood (1991) criticizes that 'neither the old term business in society nor the newer term business and society properly expresses modern business-society relationships' (Wood 1991: 66) and argues for corporate social performance as a more comprehensive and descriptive term, for it focuses on the outcomes of performance. However, we see that when these views about business-society relations are combined into recent CSR literature (Matten & Moon 2008), they help to understand the differences between institutional environments. A starting point in this research is that the language used in SERs provides a sense of the business-society relationship, which can be studied by examining conditions constructed as prerequisites for CSR. More importantly, as Matten & Moon (2008) have suggested, we have seen the spread of explicit CSR in Europe (see also Preuss et al. 2009). The choice of language therefore contributes to how we come to understand firms' roles in delivering social good or solving social problems. We now move on to examine the contributions of critical approaches to CSR during the past decade and to specifying our critical aim.

Contemporary critically oriented approaches to CSR
Theoretically and methodologically speaking, some earlier critically oriented studies in the field of CSR have drawn on, for example, post-colonial perspectives (Banerjee 2006, Khan et al. 2010 and Habermasian deliberative democracy (Scherer & Palazzo 2007). Scholars have also favoured the CDA and related approaches. CDA is as much a theory as it is a method, as it includes several different, overlapping and also distinct approaches (Wodak 2001). In CSR studies, CDA is often, but not always, reduced to the use of its specific approaches, for example, the Foucauldian approach (e.g. Livesey 2001, Banerjee 2003, the Laclanian approach (Spence 2007) or the Faircloughian approach (Burchell & Cook 2006, Joutsenvirta & Vaara 2009). Livesey (2001Livesey ( , 2002 made important contributions in the critical field of environmental and social reporting, emphasizing the power asymmetries in CSR and showing how larger socio-political struggles are connected to CSR reporting. Texts can both influence and reflect ongoing socio-political struggles over the meaning of sustainable development (Livesey & Kearins 2002) and other topics relevant in businesssociety relations. For example, Joutsenvirta and Vaara (2009) show how CSR is related to the struggle over the legitimacy of a controversial restructuring decision, which in turn has implications for businesssociety relations in the era of globalization. Tregidga & Milne (2006) analysed the development and discursive change of sustainability discourse from 1993 to 2003 in the language practices in one company. Their analysis showed how the discursive change in the business-environment relationship took place from sustainability management as compliance with regulation towards sustainable development as a business case. In this discursive change, the position of the business was discursively extended from partnership into leadership (2006). 1 Although Tregidga and Milne drew on interpretative discourse analysis, their work shows the change in language patterns used in SER and how a CSR/sustainability discourse can be employed to legitimize a right to speak with an authoritative voice (see also Driscoll 2006, Tregidga et al. 2008). According to Tregidga et al. (2008), social/environmental reporting can be seen as a policy in which companies maintain their right to speak about what constitutes sustainable development.
Some critical researchers see businesses as political actors with their own specific interests who gain more from CSR than society does and are therefore inappropriate agents of social change (Banerjee 2007, Hanlon 2008, Kuhn & Deetz 2008. For example, Spence (2007) explored the ideological implications of the motivations for SER and suggests that SER includes a hegemonic discourse that uses the metaphor of balance and aims to present the interests of business as akin to the well-being of society and the natural environment. Accordingly, hegemonic discourse aims to synthesize the interests of several groups that are basically constructed on the continuity and interests of businesses. Levy & Kaplan (2008) conclude that from a Gramscian perspective, it is not surprising that CSR, as a hegemonic accommodation, largely reflects the dominant cultural, economic and political role of business in society, and the permeation of the discourse of competitiveness and the free market into state and social structures. Spence (2009) has also concluded that reporting practices can run counter to accountability and democracy. As critically oriented management studies have pointed out, the business-society relationship is very much laden with issues of power and with struggles between various groups with competing interests in what CSR is about (Livesey 2001).
Although critical studies are often interested in challenging the status quo and have analysed different forms of domination and discrimination, critical approaches cannot be limited only to those, nor can it be claimed that they share similar specific and unified presuppositions (see e.g. Kincheloe & McLaren 2002). One must also make a distinction between criticism and critical. For example, previous research is often the object of criticism because of the way it has been performed or because of its possible neglect of certain topics. Yet a study that is positioned as being critical should include a specific critical aim. When positioning ourselves in the field of relational approaches to CSR from a critical perspective, we build on the argument that the business-society relationship is laden with various asymmetries of power, those asymmetries being present in various interests that contribute to defining the relationship. Because power asymmetry is associated with discursive activity, we shall now move on to describe our critical discursive approach.

Critical discursive approach
CDA explores the tensions between language as socially shaped and language as socially shaping; it sees language as a form of social practice. Social life is a network of diverse social practices, including for example, economic, political, cultural and familial practices. These social practices are more or less stable forms of social activity, which most commonly involve discourse (Fairclough 2003). Thus, discursive practice is a structure and also structures actions (Weiss & Wodak 2003: 10). According to this view, discourse is socially conditioned and constituted in a reciprocal sense. It is constitutive in the sense that it helps to maintain and reproduce the status quo and may also contribute to its transformation; human agency produces structures, which simultaneously serve as the conditions for reproducing human agency in a continuing process (Weiss & Wodak 2003).
The tensions inherent in language use are often presented in terms of a discursive struggle. Drawing on Fairclough (2003), we view organizations and their environments as sites of power and struggle that are reflected and constructed through the use of language. Organizations are sites of struggle in terms of their conflicts and interests as well as in terms of the blurring of the boundaries of the organization. According to Bourdieu (2000), political (and therefore discursive) struggle is a struggle to impose a legitimate vision of a social world and therefore this legitimate vision strives for the speaker's own wellbeing (Blackledge 2005: 181). The discursive struggle not only constitutes but is also socially conditioned by various groups of interests.
Two types of principles are identified in order to approach forms of power: the ideology critique and communicative action (see Alvesson & Deetz 2000). The theory of hegemony presents a form of ideology critique. Hegemony is power to present things in a certain way within a network of multiple other choices (Hall 1999). Brown (2004: 96, drawing on Clegg 1989) notes about hegemony that it 'is a form of cleverly masked, taken-for-granted domination, most often articulated as what is "common sense" or "natural", and which thus "involves the successful mobilization and reproduction of the active consent" of those subject to it'. For example, discursive frameworks that actively constitute perceptions of mutual interest in communicative actions can maintain business hegemony (Levy & Egan 2003: 653).
We apply the approach of CDA developed by Fairclough that focuses on textual, discursive and social practices. 2 Social practice is an articulation of diverse social elements within a relatively stable configuration, like SER, that always includes discourse. According to Fairclough (2001), every social practice includes, but is not limited to, the elements of activities, subjects and their social relations, instrument, objects, values and discourse. These elements are different but not totally separate. Fairclough (2001) further explains the role of discourse in social practice: discourse figures as a part of the social activity within a practice. For example, a part of doing some job or activity is using language in a particular way. Discourse is also used to make representations and the way representations are made also positions actors within the social practice. Thus, subjects may use whichever order of discourse they can more easily position themselves into and have access to the discourse. That choice is being affected by cultural and social order through hegemony. More importantly, social practices do not rely on one type of discourse, but various discourses are networked together through 'the order of discourse' (Fairclough 2001(Fairclough , 2003. That is to say that some ways of making meaning and representing things are more dominant than others. Hegemony is therefore power over the order of discourse. Discourse and discursive practice is constructed through textual choices, such as the use of vocabularies and other textual tactics (Fairclough 2001(Fairclough , 2003.
We particularly want to analyse the discursive strategies that emerge in CSR reports. By discursive strategies, we mean those discursive and textual patterns and practices used to construct a sense of business and society relations. We see SER as a recently introduced yet legitimized social practice. These social and environmental reports include certain discursive and textual practices through which they contribute to the creation of our social reality. In reporting language, some representations of business-society relations are likely to become dominant but may also incorporate a struggle over how the relationships are understood to be conducted.

Description of the empirical data and the process of analysis
Finland is a European Union country with a population of approximately 5.3 million. Transparency International research (e.g. TI GCR 2009) has presented Finland as one of the least corrupt countries in the world, and Finland's global competitiveness has been praised in global comparisons. Furthermore, it has not been thought necessary in Finland to debate extensively whether businesses should address social or environmental responsibilities (Kourula 2010). The Scandinavian approach has emphasized the stakeholder view: the acknowledgement of groups related to the existence and survival of the firm (Näsi 1995). However, the idea of business responsibilities is in constant flux because of a variety of interests and appreciations that can be in conflict (Joutsenvirta 2009). Environmental reporting gained a foothold in 1990, and the social dimension was established a decade later. In the 2007 State of Responsible Competitiveness evaluation by AccountAbility (Zadek & MacGillivray 2007), Finland was placed third globally. Indeed, it is often suggested that Finnish companies have been relatively progressive in CSR, and they generally perceive it as a potential competitive advantage (Kourula 2010). In particular, Finnish companies have been represented as stakeholder-oriented businesses and proactive CSR agents (Juholin 2004). CSR has traditionally been largely implicit in Finland (see Matten & Moon 2008), which means that the state is assumed to be responsible for social issues, and there has not been the strong philanthropic tradition evident in many other countries. However, a more explicit form of CSR has started to emerge (Kourula 2010).
Public ranking lists can also naturally contribute to the creation of the image of a nation with high competence in matters of CSR, which further emphasizes the importance of examining how a certain sense of the terms and conditions of CSR in practice is constructed in the social and environmental reports of Finnish businesses. We selected the responsibility reports [or sustainable development, environmental or Eco-Management and Audit Scheme (EMAS) reports] of 24 Finnish businesses published in either 2005 or 2007. We based our selection on three criteria: the results of a national CSR/environmental reporting competition, the list of Finnish corporations in the Dow Jones Sustainability Index and the list of EMAS-registered organizations in Finland (regulation number 761/ 2001). We are not suggesting that we have in this way developed objective criteria with which to define firms that are actively participating in SER but rather that these three criteria provided us with a framework for finding interesting data. We went through the national reporting competition starting from the beginning of the 21st century and selected those corporations that had won awards. We also chose all the corporations listed in the Dow Jones Sustainability Index and the corporations whose sites were EMAS-registered at the time. To find EMAS-registered corporations in Finland, we went through the list of EMAS-registered sites in Finland. We disregarded the subsidiaries of larger corporations that were already on the list and the one corporation that was part of an international group. The first part of the data was collected in 2005, when the latest published report was selected. After a preliminary analysis, the selection criteria were expanded in 2007, when EMAS registration was added to the existing criteria. As EMAS reports are published only every 3 years, the reports that we analysed included information from the years 2004-2006. On the basis of our selection criteria, we chose a group of 24 corporations (Table 1). These corporations operate in 14 different business sectors: shipping, food, energy, machinery, metal, waste management, retail, packaging, forestry, communications, information technology, real estate services, transportation and construction.
We first familiarized ourselves with the reports and focused specifically on those parts of the reports in which textual descriptions and definitions of responsibility were documented. This phase eventually led us to pick out 156 descriptive passages. We applied Fairclough's (1992Fairclough's ( , 2003 framework and concentrated on the relationship between textual, discursive and social practice. In terms of textual practices, we paid specific attention to the vocabulary used: how a position for the business in relation to other actors was constructed and what vocabulary signalled this position. We then focused on the construction of discursive strategies; in other words, how the textual choices signified the relationship and interests between the business and society. We noticed how some texts used collaborative and participative expressions in relation to other societal actors. However, the idea of business being in a pivotal position in relation to other actors dominated other texts. We distinguished two main discursive strategies; centralizing and decentralizing strategies based on their mobilization of CSR discourse. Centralizing strategies emphasize instrumental means and decentralizing strategies, integrative means. Society was thus constructed as having either integrated interests with the business (business in society) or was seen as instrumental (business and society) in delivering business interests. We analysed 156 passages and centralizing strategies emerged in 63 of them and conversely, 93 revealed decentralizing strategies. We then positioned these discursive strategies into earlier literature and through the order of interests in the CSR discourse, considered their significance for SER as a social practice. However, this identification and analysis were by no means a straightforward process. For example, in order to comprehend how the discursive strategies participate to position business/society and to show how they link into broader literature, our analysis and writing process entailed constant circling and iteration (Wodak 2001) -going back and forth between the data and the theoretical framework. The construction of a more coherent narrative and reader-friendly version further led us to eliminate some analytical elements at the end of the process.

Discursive strategies
In this section, we will present the discursive moves in the two strategies -centralizing and decentralizing ones -and show their significance and implications for business-society relations.

Centralizing strategies
Centralizing strategies reflect the business-society relation as one between business and society. The discursively produced relationship emphasizes businesses as independent actors with certain interests in abiding by the expectations of society and stakeholders (see Wood 1991). More specifically, through the use of subject positions that imply authority (such as a leader role), centralizing strategies strengthen businesses' right to define responsibility and, more importantly, the limits of that responsibility. We present the analysis by first discussing the textual practices that construct this strategy, followed by the analysis of the importance of discursive strategies. Finally, we analyse the significance of centralizing strategies in constructing specific interests and societal conditions as a basis for CSR.
On the textual level, we focused on how textual choices used in reports discursively mark and signify the position of the businesses in relation to society. The textual structure of centralizing strategies draws on terminology such as market leader, forerunner and know-how. Tregidga & Milne (2006) also noticed how the construction of 'a leadership role' affects how an organization positions itself in relation to its stakeholders: it becomes a knowledgeable  provider of information. A well-recognized argument in corporate responsibility literature is that responsibility aims to address and answer environmental and social problems (Margolis & Walsh 2003). Know-how therefore implies the possession of knowledge of issues that can be understood as goals of CSR. Although know-how was evident in many of the quotations, it was rarely made explicit precisely what know-how the business claimed to possess. The importance of this vocabulary is in how it signifies a certain position of businesses in the business-society relationship. For example, the construction of a 'forerunner image', particularly in environmental issues, can strengthen the image of the business whose actions go beyond mere compliance.
1. Our strategy is to gain a competitive advantage by providing professional management of propertyrelated environmental issues and environmental costs of the properties that exceed the client's expectations during the whole life-cycle of the real estate. In environmental issues we are a forerunner with solutions based on specialized knowhow. (Y16, translated) 2. Huhtamäki is a leading, worldwide package producer, whose products and services fulfil high quality standards. The corporation acknowledges the responsibility that comes with its position in promoting sustainable development and has already for some time incorporated into the principles of central management, objectives related to quality, hygiene, safety, health and environment. (Y10, translated) 3. We are continuously developing our activities in order to better protect the environment and improve safety by applying the know-how we can access through our international contacts and partners. (Y2, translated) At the discursive level, we show the significance of centralization as a discursive strategy that mobilizes CSR discourse in an instrumental sense. In this case, it means how this strategy reflects the business and society view. The businesses are constructed to be capable of regulating themselves beyond the demands of legal compliance by producing benefits for the society and minimizing (the possible) harm they cause. Furthermore, the discursive elements emphasize goals businesses set for themselves instead of those set to meet the expectations of others. A position is thus constructed with the right to define the direction of societal development. This kind of language use emphasizes a right to define and determine business-society relations and what is understood as responsible business. The use of the terms 'review' and 'monitors' draws attention to the power of business to control its own actions and thus, rejects the alternative -the need for external and objective control or auditing. 4. We endeavor to fulfil set targets for environment, quality and occupational safety in our planning and execution of process services, whose fulfilment we regularly review. (Y23, translated) 5. We wish to continuously improve our environmental management, promote development and assess our progress. Wherever possible we do more than the law requires of us. (Y5, translated) 6. TVO takes responsibility for the environment by minimizing the damage that might be caused by its activities and by taking care of the waste it causes; it monitors the state of the environment and undertakes reconstructive action as soon as it becomes necessary. (Y19, translated) In centralizing strategies, the role of businesses is portrayed as akin to actors who can also steer the behaviour of others. The self-constructed expertise in acknowledging the need to address responsibility through investments and control is used to legitimize autonomy over matters: the organizations have 'knowledge capital' to address problematic and challenging issues. For example, in excerpt no. 8, the language practices used describe the business as taking advantage of other actors in order to develop its actions. It is further described how the implementing CSR policy requires other actors and stakeholders to participate. Although this might seem obvious, it is rarely acknowledged that meeting CSR obligations requires the action and participation of numerous other actors in addition to the action of the business itself. Centralizing strategies veil the role of stakeholders as a resource necessary for business to meet its responsibilities. However, when stakeholders are defined as a resource required for a business to perform its responsibilities, it raises an important question of how the interests and expectations of groups constructed as a resource are to be achieved.
7. The goal of Outokumpu is that its business partners, subcontractors and suppliers become familiar with its corporate responsibility principles, and that they follow similar standards. (Y21, original) 8. We wish to influence societal development by offering our clients augmentative solutions that are environmentally friendly and promote wellbeing. Through our own actions we also seek to guide our partners and clients towards making responsible solutions. (Y16, translated) The instrumental view constructs a position for businesses that the business and society view emphasizes. In so doing, it contributes to the order of interests in CSR discourse: the role and interests of business over stakeholders and societal expectations become naturalized, an uncontested phenomenon. Responsibilities are prioritized rather than balanced: economic activity and growth enable the achievement of social and environmental aims.
9. Saarioinen's business is guided by a controlled growth strategy. The even development of production, sales and outcome is required for controlled growth. On the basis of a stable economy we create the prerequisites for environmental work and its development. (Y7, translated) 10. Wärtsilä's target is to improve its financial performance and create added value for its stakeholders and society. Strong financial performance forms a basis for corporate environmental and social responsibility. (Y1, original) To conclude, centralizing strategies convey a sense of expertise for businesses in business-society relations as actors who are able to define what constitutes responsibility. These strategies involve constructing a position for businesses in which they would legitimately promote their own interests and goals rather than being actors subject to the expectations of society. The strategies construct businesssociety relations more from the business and society viewpoint, emphasizing their interactive yet separate natures.

Decentralizing strategies
Decentralizing strategies are constructed more on the basis of the business in society relationship, meaning that the corporation is a part of society and that stakeholders have legitimate interests in its actions. The vocabulary used here draws from openness, interaction and cooperation, and in decentralizing strategies, business-society cooperation emerges in the reciprocal sense. In addition, the strategy applies integrated means in mobilizing CSR discourse and uses the metaphor of balance (Spence 2007). We first discuss the textual practices followed by the discursive construction of the strategy and finally, analyse the significance of decentralizing strategies for interests in business-society relations.
One argument behind demands for business responsibility is often the need to increase the openness and transparency of firms. This is sometimes referred to as a spur for the writing of responsibility reports, although reporting does not always increase transparency (Quaak et al. 2007). The vocabulary in this strategy draws on 'openness' and 'interaction', implying a business that is a part of a wider network of relations. The interaction is strengthened by providing a context for interaction that is associated more with cooperative negotiation than guidance. Interestingly, in centralizing strategies, the leadership position was already taken ('we are forerunners') whereas in this strategy, the forerunner position is portrayed as something the organization would aspire to achieve through cooperative relationships. More specifically, this strategy emphasizes stakeholder consultation facilitated by the interaction and is more about 'partnership' (see also Tregidga & Milne 2006).

We want to act in open and equal interaction
with all the people and organizations around us by discussing, by taking a stance and listening. (Y5, translated) 12. Fortum's values dictate that our operations aim to balance economic, environmental and social aspects by having a positive effect on the communities in which we operate. Our long-term goal is to be the leading sustainable development energy corporation in the Nordic countries. We want to be forerunners and to build the future of sustainable energy in an open and active interaction with our clients and other stakeholders. (Y4, translated) CSR literature often discusses the contribution of business to society in one way or another; the idea of contribution is a premise in any discussion of corporate responsibility (Banerjee 2008). Yet the nature of the contractual circumstances that govern the conditions of this contribution is sometimes ignored. Use of this discursive strategy emphasizes an interactive relationship with other societal stakeholders and acknowledges the demands of those stakeholders the business should contribute to. Furthermore, the use of the strategy acknowledges the impact stakeholders can have on the achievement of goals. However, examples in which the corporation is explicitly represented as subordinate to societal expectations were rare in our data. Although there are more decentralizing strategies identified than centralizing ones, very few feature direct expressions indicating subordination to societal demands.
13. Metso anticipates its clients and society's expectations regarding environmental protection. Processes and best practice is developed with clients and partners, which will save on the environment as well as use energy and raw materials efficiently and on a lasting foundation. (Y6, translated) 14. Ekokem aims towards openly cooperating with its clients, the authorities and other stakeholders in order to be able to develop the corporation in such a manner that its operations may even better answer clients' expectations and the changing demands of society. (Y3, translated) 15. The corporation is committed to active environmental communication and from an environmental perspective works towards carrying out positive solutions in close co-operation with clients, owners, authorities and other stakeholders. (Y18, translated) Equality between, and balance of, various priorities is also a key theme in CSR and SER discussions (Elkington 1997, Spence 2007). This was not, however, something that would have marked the strategy in a significant way. The idea of balance and equality is, however, extended to the treatment of various stakeholder groups.
16. In all of our operations, we acknowledge our customers' needs, expectations and principles of sustainable development, good economic results, continuous improvement of environmental issues and our social responsibility. These elements support each other. (Y17, translated) 17. The principles of sustainable development guide the continuous improvement of our actions and the development of issues related to the environment and society. In order to accomplish this task, we consider it extremely important that an open dialogue and interaction with stakeholders as well as with authorities and NGOs takes place. (Y12, translated) Many of the theoretical approaches used to describe CSR feature the explicit claim that CSR means answering societal demands, and this is particularly common in legitimacy theory approaches (see Deegan 2002). The societal legitimation argument (Coupland 2005) describes the needs of society and presents business as responding to those needs. However, the interactive production of openness is not a direct translation of adaptation to societal expectations and norms. The strategy differs from the previous centralizing one in the sense that the attitude to stakeholders focuses on universalization (see e.g. Thompson 1990); businesses appear open and available to all and address the needs and claims of everyone. More specifically, this strategy presents stakeholder interaction in a rather idealistic manner in which conflicts of interest do not take place, preferring to emphasize win-win solutions. The order of interests in CSR discourse is thus different from that in centralizing strategies: a decentralizing strategy mobilizes CSR discourse using integrative means towards the idea of balance in CSR priorities and balance in stakeholder relations. For example, the reference to neutrality (excerpt no. 18) as a guiding principle in stakeholder relations implies that such activity is possible. However, in previous literature, the interests are often understood to be competing or confrontational (Mitchell et al. 1997, Banerjee 2007 To conclude, contrary to centralizing strategies, decentralizing strategies emphasize the idea of business in society. Furthermore, the discursive means used contribute to attempts to seek legitimacy by presenting business as an actor that can openly address the interests of stakeholders in a neutral and balanced manner. However, universalization of interests also dispels the power asymmetries in business-society relations.

Discussion and conclusions
Our aim in this article was to examine the ways in which business-society relations are constructed in the SER of 24 Finnish companies. Using a critical discursive approach (Fairclough 2003), we examined the asymmetry of interests in constructions concerning the business-society relationship. The aim was to understand whose interests dominate in the relationship, how it affects the construction of the relation-ship and what its significance is for CSR discourse. We elaborate on our findings and conclusions and discuss our limitations more broadly here.
Our process of analysis leads us to distinguish two main discursive strategies through which businesssociety relations are constructed in SER: centralizing and decentralizing strategies. In centralizing strategies, through the construction of subject positions that imply authority (such as a leader role), the role of a business and the right to define not only its responsibility but the limits of that responsibility is strengthened. Centralizing strategies construct businesses as strategic forerunners rather than actors merely responding to expectations or regulation. Centralizing strategies simultaneously veil the role of stakeholders as an instrumental resource to achieve responsibility. Decentralizing strategies were more prevalent and provided constructions in which businesses would meet the expectations of stakeholders. The discursive means used thus construct business as an actor that can openly address the interests of stakeholders in a neutral and balanced manner. Interests are constructed as jointly negotiated and shared with stakeholders in an integrated sense. However, these kinds of discursive means, when applied, can neutralize the fact that businesses are pursuing their own interests in seeking cooperation with other stakeholder groups (Spence 2007). Although decentralizing strategies emphasize integrated interests, regulative procedures and societal legitimacy have long been acknowledged as a precondition for business-society relations in Nordic countries. The emergence of 'stakeholder talk' is therefore no surprise, but SER also provides a communicative channel for the rediscovery of the value of stakeholders: something that has generally been acknowledged for decades in Finland as a way of positioning the business. In international CSR literature, the focus has been more on whether CSR pays off (see Margolis & Walsh 2003).
We contribute to the contemporary discussion of discursive struggles in business-society relations and the role various interests play in that struggle (see Levy & Egan 2003, Banerjee 2007, Spence 2007, Tregidga et al. 2008. CSR discourse as such can be considered powerful because it describes how companies would give something back to society or develop its well-being, but at the same time, it hides tensions and is vague about how such well-being could be accomplished -through what sort of relationship and on whose terms (Banerjee 2007). Discursive processes have implications for how we come to understand not only the responsibilities of businesses but also the limits and terms of taking such responsibility.
We therefore suggest that SER is a social practice shaped by discursive struggle over dominant interests. SER also strengthens the hegemonic assumptions that businesses can solve social and environmental issues and naturalizes stakeholders' input as a part of this activity. Not only is reporting practice a societal legitimacy quest in which power asymmetries are veiled using 'cooperative' and 'balancing' language, but it can also be seen as a communicative action that provides a right to define the role of societal actors for the achievement of CSR. Furthermore, in each of the 24 corporate responsibility reports we analysed, both of the discursive strategies were used, whereas in the current literature, the business-society relations are often treated as an either/or phenomenon (Wood 1991). This means that social practice such as SER would then reflect either the business and society or the business in society view. Based on our discursive analysis, this is not the case. In some extreme cases, a passage from a report could indicate both business and society and business in society orientations. We see that because businesses are positioned in both ways, the analysis shows the role SER can play in business hegemony. More specifically, not all societal actors can be constructed in a position that would contribute to their interests, nor can all actors have access to a certain discourse. Yet in SER, these two views were constantly mixed, which indicates the power over the order of discourse and more specifically, the power over the order of interest. The distribution of the passages across the companies is illustrated in Table 2.   The current study raises issues that future studies might tackle. For example, it would be interesting to have comparative studies on business-society relations between larger and smaller businesses or between businesses with their origins in various national contexts. For example, as Table 2 shows, there was no clear difference in reporting language between companies operating internationally and only in the domestic market. In addition, the focus could be more firmly on the cultural conventions on which the limits of responsibility depend, and on how such constructions might become more global. Detailed analysis of the argumentation that constructs the two major strategies would greatly contribute to our knowledge of the micro-processes through which phenomena such as business-society relations come to be constructed. Furthermore, while studying corporate responsibility from a critical point of view, it seems the critical field of CSR still remains rather fragmented. Despite the fact that CSR literature has attracted a number of theoretical reviews in recent years, the field would greatly benefit from an explicit review of the role of critical theories and the contributions of a number of critical scholars, including, but not limited to, Habermas (Scherer & Palazzo 2007), Gramsci (see Spence 2007) and Foucault (see e.g. Banerjee 2007) to business-society relations.

Limitations
Although our study makes some interesting contributions to the field, it inevitably has a number of limitations. First, the research method chosen for the study is ideological in itself, a fact that CDA scholars acknowledge, in that it provides a particular way of reading the texts and thus, constructs a particular picture of a phenomenon. Other readings are possible with other theoretical and methodological approaches. The analysis of discursive strategies can also be seen as limited. Providing a rich picture of the variety of discursive practices used is not possible in one short study. Second, the data included companies operating both nationally and internationally, and the cultural richness of SER could be better explored using more limited and structured data. One of the most important issues would be to adopt a longitudinal approach and concentrate on discur-sive changes in SER. The nature of SER is continually expanding and developing, and this calls for more research.
1. Although Tregidga & Milne (2006) do not make the distinction between partnership and leadership, they emphasize only the leadership position. However, we suggest that such a distinction can be examined as a discursive change from the quotes and vocabulary provided in their study. 2. However, we do not aim to analyse all kinds of textual tactics, argumentation strategies, genres and intertextual elements related to the use and production of discourses, this being far too vast a scope for one essay (see Reisigl & Wodak 2001, Fairclough 2003.