Speaking Up for the Natural Landscape : A Rhetorical Dilemma

This article presents textual evidence which shows some of the ways in which green business corporations and environmental NGOs represent the natural landscape and their relationship with it. It reviews the origin and development of stakeholder dialogue and questions to what extent such dialogue can contribute to a process of corporate change. It shows how the corporations use different language to represent nature than the NGOs and provides evidence suggesting that the green corporations understand their relationship with the natural landscape differently. NGOs that wish to speak up for the natural landscape, face a rhetorical dilemma which has an important implication for their practice. Either they can enter into a stakeholder dialogue with business and risk becoming a party to the exploitive management of nature, or they can refrain from entering into a dialogue and risk becoming marginalised.


Introduction
Business corporations' recognition of stakeholders would appear to be almost universal; although not rigorous, a sampling by the author of 25 websites selected at random from the CNN Global 500 list, revealed that every one of the corporations used the term.Dialogue with stakeholders is often presented by corporations as being a desirable and important factor in effecting change towards more sustainable ways of doing business.The literature review in this article includes research that supports this view; some dialogues can be productive for both corporation and stakeholder.However, nature, or the natural landscape -the term I shall use in this articlesingles itself out from the other potential stakeholders of the corporation by the simple fact that it cannot speak up for its own rights.The task is often undertaken on its behalf by environmental campaigning organisations.This article discusses what the term dialogue can mean, particularly when one party to the dialogue is unable to talk for itself.Further, it makes a contribution to understanding what stakeholder dialogue's possibilities and limitations might be for securing the interests of the natural landscape, when it is cast in the role of a stakeholder of the corporation.

Scope of Study
As the literature review shows, dialogue is the first stage in a corporation's development of stakeholder engagement.Scholars have suggested that dialogue ought to lead to consultative and from that to participative or decisional relationships.This study has identified characteristics in the representation of the natural landscape by green business and environmental NGOs.An obvious area for further study would be to find samples of language in which dialogues and/or interactive processes are attempted.However, this is beyond the scope of the current work.
The study sought to identify language samples produced by corporations that had committed their businesses to the principle of sustainable development.These 'green' corporations would, presumably, have the most enlightened attitude to the rights of the natural landscape.The green businesses were selected on the basis of one or more of three criteria.The first criterion was membership of the World Business Council for Sustainable Development (WBCSD, 2012).The second criterion was membership in the UK organisation which is affiliated with the WBCSD; the BCSD-UK (2012).This provides a membership forum for corporations which are not as large as those that traditionally join the WBCSD.The third criterion was being a signatory to the United Nations Global Compact (2012).The compact did not require member corporations to make an explicit commitment to sustainable development.However, the scope of its three environmental principles was so comprehensive as to make it the equivalent of a commitment to sustainable development (Note 1).NGOs were selected on the basis that they had to be British based, campaigned on behalf of some aspect of the natural landscape, and had no connection with either government agencies or business.Many websites were rejected for a variety of reasons during the process of selection (see Brown, 2008, pp. 336-338).
From these two groups of websites language was downloaded which made representations of the natural landscape as a stakeholder of the corporation.In the next section there is a very brief history of the stakeholder concept followed by a summary of the literature on stakeholder dialogue and examples of research focused on its communicative aspect.Freeman (1984) introduced the concept of a stakeholder as "any group or individual who can affect or is affected by the achievement of the firm's objectives" (p.25).Following his pioneering work, much research was carried out to develop the idea further and then to clarify it as a theory of the corporation, its most notable of many endorsements being delivered by Donaldson & Preston (1995).Examining the theory's descriptive aspects they cited empirical work from as early as the 1960s.Surveying the opinions of upper-level managers, Baumhart (1968) reported that "80 percent regarded it as unethical management behaviour to focus solely in the interests of shareowners" (reported in Donaldson & Preston, 1995).Other later work (Brenner & Molander, 1977;Posner & Schmidt, 1984) is also cited as confirming that the idea of stakeholders was already present -in the consciousness of managers at least -at the time at which Freeman's book was being written.

Literature Review
In the first figures which Freeman used to illustrate the stakeholder model of the firm, he refers simply to "Environmentalists."It is Carroll (1992) who is credited with making the case for the natural landscape to be considered as a stakeholder of the firm.Onkila (2011) has recently provided a useful summary of research work that has advocated the natural landscape as a legitimate stakeholder (p.382).
As business began to implement the stakeholder concept into its operations, a new object of study developed, allowing researchers to examine how the corporation-stakeholder relationship functioned in practice.One of the areas of research focus was to explore how corporations interacted with their stakeholders.Grafé-Buckens and Hinton (1998) collated responses to a survey sent out to 100 European corporations all of which had already undertaken what the authors termed "Environmental Stakeholder Initiatives (ESIs)" (p.125).Out of their findings they created an evolutionary path for ESIs which started at the "Informative" level, progressed through various "Consultative" levels, and could reach its highest form in the "Participative" level (p.129).Oxley Green and Hunton-Clarke (2003) interpreted previous literature, largely focused on experience from the public rather than the private sector.However, they landed on a very similar three-level typology of stakeholder participation: "informative, consultative and decisional" [emphasis in original] (p.295).
Implicit in this and other work is the assumption of a dialogue between corporation and stakeholder.Madsen and Ulhøi (2001), whose primary focus is to develop a model for assessing stakeholder groups and their influence on the corporation, nonetheless conclude their paper with an observation about the communication process between stakeholders and the corporation.They suggest that the building of "learning partnerships with the stakeholder system requires an open communications strategy" (p.86).Coinciding with the 2001 publication of the article, it is roughly since the start of the new millennium that the term stakeholder dialogue has come into usage and research focus.In 2003 Kaptein and Van Tulder reviewed experience to date in an optimistic tone and sketched out "ten preconditions for effective stakeholder dialogue" (pp.210-213).
One important consequence of stakeholder dialogue which is reviewed, is the potential for organisational learning.The idea that corporations can learn through dialogue with their stakeholders is a theme that is explored by other scholars in addition to Kaptein & Van Tulder (see Clark & Roome, 1999;Roome & Wijen, 2006).However, three years after the optimistic tone of the Kaptein & Van Tulder article, Burchell and Cook (2006) expressed the first misgivings about stakeholder dialogue that I have found in the literature.Their article is based on survey results followed up by interviews of personnel in both corporations and NGOs which were actively engaged in stakeholder-dialogue processes.Burchell and Cook report that some corporate and NGO respondents certainly saw the potential of dialogue for expanding personal understanding of competing perspectives and developing a structure which allows participants to move away from traditional confrontational stances which have often characterized existing company/NGO interactions (p.220).
But they observe that this optimism was tempered on both sides by the need to justify the investment of time in the dialogue with "tangible outcomes" (p.220).They return to this dilemma in a subsequent article (Burchell and Cook, 2008).Two years on, they now report that a "significant number of international NGOs have already decided to no longer participate [in stakeholder dialogues] due to what they perceive as the unwillingness of companies to engage in meaningful dialogue" (p.45).The term "meaningful dialogue" [emphasis added] is important in developing my argument.Some conversational participants are satisfied simply to take part in a process of broadcasting information and opinions at other participants.Having delivered their message into the space between themselves and the other participants and ignored the messages that were being broadcast at them, they retire from the conversation; mission accomplished.Some people might consider this to be a dialogue.But in the quote above the NGOs were clear that they had entered into a conversation with green business with the expectation that the dialogue would be meaningful i.e. that they would exchange meaning with business.This is a much more ambitious project than the process of broadcasting information and opinions at each other.In order to take part in a meaningful dialogue, one needs shared meanings in a shared vocabulary.This is a subject to which I will return at the end of this section.
Burchell and Cook are two of the more sceptical voices in a scholarly discourse around stakeholder dialogue which is decidedly mixed.More neutrally, O' Riordan and Fairbrass (2008) present a model for the integration of stakeholder dialogue into a CSR decision-making process in which "management response" consists of CSR "strategy development," followed (axiomatically?) by strategy "implementation and control" [emphasis in original] (p.754).Their research may shed very welcome light on the circumstances in which stakeholder dialogue leads to positive changes in corporate practice.Continuing in this vein, Ayuso et. al. (2006) report on a case study of two Spanish companies "that have successfully learned from stakeholder dialogue and have generated innovations that are beneficial both for the company and for sustainable development" (p.475).
Certainly among the more positive views in the discourse on stakeholder dialogue are Van Huijstee and Glasbergen (2008).They express great optimism in the potential for stakeholder dialogue to evolve and lead to substantive corporate change.Although their research findings lead them to concede that "sustainability considerations may not be at the core of [current] corporate decision-making," they counter that "the evolution of dialogue processes should not be disregarded" (p.307).They explain in greater detail how they anticipate this "evolution of dialogue processes" might play out, invoking Habermas's ideal of communicative rationality (1981).In the final paragraph of their discussion, they sketch out the possible, desirable consequence of ongoing stakeholder dialogue.Encouraged by the possibilities of organizational learning, the goal of sustainable corporations appears to be within reach: Once the corporate participants come to understand the NGO's viewpoints, their mental models might be challenged.Sustainability considerations might become more important in their personal value system and consequently in their decision making.If so, the dialogue participants would have experienced a fundamental learning process of the kind Argyris and Schön (1978), Senge (1990) and Cramer (2005) discuss in their work on organisational learning.The experience might induce corporations to move from the strategic management model towards the sustainability model (p.308).
I should underline that it is not my intention to argue against the practice of stakeholder dialogue.Articles in various journals are building up empirical evidence which supports the view that corporations that are willing to spend some of their profits on community-building projects can and do improve the welfare of different stakeholder groups (e.g.Kepore & Imbun, 2010;Nilsson and Fältholm, 2011).Stakeholder dialogue clearly has some useful role; it is delivering some "tangible results" -to use Burchell and Cook's phrase -for some stakeholders.Endorsing the O' Riordan and Fairbrass (2008) project, more research is needed to find out which stakeholders succeed in winning concessions from business and, crucially, whether the dialogue process has any useful role to play in securing them, rather than the gains being, for example, simply a consequence of the amount of power which the stakeholder is able to leverage over the corporation.Further, we need to see more evidence of the type advanced by Ayuso et. al. (2006), in order to understand in what circumstances stakeholder dialogue can lead to organisational learning.With reference to the earlier point regarding the meaningfulness of communicative exchanges, is there evidence of shared meanings in a 'natural landscape-business' stakeholder dialogue?Some scholars, whose opinions on the potential of stakeholder dialogue to effect change appear to be more sceptical, turned their attention on the dialogue process itself.Rather than studying the 'real world' circumstances of stakeholders and any possible 'real world' improvements, or possible 'real world' changes in corporate practice, they have studied the ways in which the real world and any possible changes in it have been represented in language by the actors involved (Note 2).This linguistic focus was foreshadowed as early as 1997 by Richard Welford.Also frustrated by experience of too much talk and too little action he commented; "[i]t is increasingly clear that when we discuss environmentalism many of us are essentially speaking very different languages" (p.32).
One theoretical approach to conceptualising the relationship between language and reality is to conceive of them as potentially separate from each other; language, although supposed to represent reality, can float free of it.In this socially-constructed sphere, discourses compete for hegemony and the successful one(s) will then shape social practice.This view has its origins in the 'speech-as-action' work of Austin (1955), continued by one of his students (Searle, 1969), and also in the sociological interpretation of knowledge advanced by Berger and Luckmann (1966), although the separation has been carried further than they, perhaps, envisaged.Two articles which are inspired by this approach and which take corporation-stakeholder discourse as their object of study have been selected for discussion here.Onkila (2010) studies the corporate end of stakeholder dialogue in 25 Finnish companies using "New rhetoric" (p.383).In the perspective of this theory, language comes very much to the fore, while the 'real world' of carbon dioxide concentrations and contaminated drinking water fades into the background.The only forms of reality which are of interest are our representations in text; "[n]ew rhetoric contends that no distinction can be made between rhetoric and reality and that rhetoric is part of a socially constructed reality" (p.383).It is important to note here that for new rhetoricians the term reality is the same as the term socially-constructed reality.Onkila is interested in how business conceptualises its relationships with stakeholders, manifestly an example of a social construction of reality, so the theoretical approach is perfectly sound in this context.It would have been interesting to learn how the stakeholders to these 25 Finnish businesses conceptualised their relationships with the corporations.This would have contributed to our understanding of how large the gap is between stakeholders and corporation in their socially-constructed interaction with each other.
Returning to an interest in a 'real reality' of supply chain sweatshops, Ählström's (2010) research is based on a comparative approach between corporate discourse and corporate practice.Using a case study of the Swedish clothes retailer Hennes & Mauritz (H&M), she shows that what actually happens in the supply chain for H&M's fashion garments is different from the CSR goals -what she terms the "responsible business discourse" (p.70)that the corporation espouses.She observes that civil society organisations have entered into a dialogue with H&M in which the corporation has broadcast this "responsible business discourse."Their expectation has been that the discourse would lead to real change in business practice.However, she argues, this "responsible business discourse" is a cynical rhetorical ploy by the corporation; business "practice is decoupled from the responsible business discourse to refrain from threatening profit maximisation" (p.78).These two articles provide a basis for working out the theoretical approach on which the research in this article is based.

Theoretical Approach and Research Questions
My first assumption, in line with Ählström, is an insistence that, in addition to many socially-constructed realities, there is a reality 'out there' of which the natural landscape is a very important example.Natural science provides us with a great many fairly objective techniques for measuring the pollutant output of our activities, the condition of the natural landscape and the effects of interaction of the former on the latter.This is not to deny the importance, for which discourse specialists argue, of studying how this reality is represented.But the condition of the natural landscape as measured by hard science is the reason why we do research and, in contrast to all the other social constructions of reality which one might study, it provides us with a common point of reference for all the social actors; comparisons of how they each choose to represent the natural landscape can tell us something about the differences between them.
The second assumption is that although this nuts-and-bolts, fleshy-and-fibrous reality is, of course, very different from language, the two are not entirely separate, at least not all of the time.Sometimes, I concede, they are different for a very good reason.In the study of CSR discourse, for example, it is common to examine the goals or mission statements of corporations.These are language representations of visions that the corporation has for how it would like to run its business operations.Visions are just that.They are not reality.It is legitimate for a corporation to describe its aspirations about how to be a better corporation, even though it hasn't got there yet.It is legitimate for a corporation to make representations of codes of conduct that will, hopefully, lead to more sustainable modes of business without these necessarily being matched by practice.If research between discourse and practice is to be attempted, then great care must be taken to ensure that practice is compared with corporate representations of that same practice for any apparent differences to be interesting.
Disregarding such 'envisioning language' then, the second assumption, within the context of corporate greening, is that discourse and practice are mutually constitutive.De-coupling of the discourse from practice is not a long term option for an aspiring green corporation.As Ählström (2010) points out, "H&M is put under scrutiny" (p.73).Whatever one's estimate may be of the cynicism of corporate senior managers, running a PR campaign of deliberate misinformation in order to continue with a profit-maximising business-as-usual practice, is professional suicide.In the long term, internal whistle blowers, Wiki leaks, the power of the media and consumer outrage would ensure catastrophic consequences for any corporation whose senior officers were reckless enough to attempt such a project.In contrast to de-coupling, the term mutually constitutive means that corporate practice -what the corporation does every day/week/year -contributes to the shaping of its discourse of CSR practice which, in turn, influences the way in which that practice is carried out.
These assumptions underpinned my research on CSR discourse and shaped its approach.I studied language that makes representations of corporate practice, the activities of business in the natural landscape and the consequences for the latter of those activities.Further, I selected language representations from two different actors; green business and NGOs speaking on behalf of the natural landscape.Further information concerning the creation of this object of study may be found in a previously published article (Brown, 2010).The second assumption in the argument I have just presented, means that my working hypothesis was that since the 'practice' of the two groups was different from each other, their respective discourses of practice must also contain differences.
The first research question was: What are the differences and similarities in the language used to make representations of the natural landscape and the actors' relationship to it by (i) the environmental NGOs and (ii) the green business corporations?
The second research question sought interpretive answers to how these two groups conceptualise their relationship with the natural landscape.In the case of the NGOs, who have taken it upon themselves to speak on its behalf, the language usage ought to provide evidence of how they conceptualise the natural landscape rather than their relationship with it.But for the green corporations, the language ought to provide evidence of how it practices its interaction with, in addition to its conceptualisation of, the natural landscape.
Research question two was: What evidence does the language provide which can indicate how the groups conceptualise the natural landscape and/or their interaction with it?
Clearly, my assumption of a mutually constitutive relationship between discourse and practice ought to lead to evidence in the language that illustrates (i) differences between the two groups' usage of language and (ii) patterns of language usage which are consistent with what we already know about the practice of the two groups.As such this empirical confirmation of already existing knowledge is not a significant contribution.Of greater value, however, are the conclusions that may be drawn about the possibilities for a successful two-way communication of meaning.For the NGOs who speak up for the natural landscape, how can they engage in a meaningful dialogue with corporations in order to best secure the interests of their chosen stakeholder?

Method
Electronic texts written and published on their websites by British green businesses such as British Petroleum (BP), Rio Tinto and Anglo American were downloaded to make a language database that contained over three million words.37 British-based NGOs were identified as having something to say on behalf of the natural landscape.There were major actors such as Friends of the Earth, GreenPeace and Oxfam and many smaller single-issue organisations.This database contained over 12 million words.In both databases the texts concentrated on making representations either of the natural landscape and its condition or of the practice of business in the natural landscape (Note 3).
Using a PC-based language analysis tool called Wordsmith (2012) and using the benchmark English-language database of the British National Corpus (BNC, 2012), I ran statistical reports of the keywords of the two groups.These listings, for one-, two-and three-word keywords, show which words are used by the two groups with the highest frequency when compared with the 'average' English contained in the BNC.From the resulting objects of study, I then extracted words into four different collections which were part of the representation of some aspect of reality.
Given my research interest in representations of the natural landscape, the first collection consisted of representations of what I termed the fleshy and fibrous natural landscape.As the term implies, this was vocabulary which I judged as representing a 'fleshy' or 'fibrous' part of the natural world e.g.crops, forests, people and food.This collection of words was selected because of a very clear distinction between the two groups that was observable when comparing their keyword lists.The green business listings were almost completely devoid of vocabulary that represented the fleshy or fibrous natural landscape.Instead, green business language made representations of what I have termed the socially-constructed natural landscape -the second collection.Examples of such words are areas, biodiversity, health and habitat.These words also have their origin in the natural landscape, but they represent some sort of social construction that has been placed on it.They are definitions of some aspect of the landscape which is of interest to the text writer.The third collection consisted of the very prominent vocabulary of a process of managing used by both groups, though with some internal variation in focus.The fourth collection of words was of representations of risk.This category of words makes representations which express a concern about threats to the natural landscape, the source of the threats and their possible consequences.The collections comprising these four categories of vocabulary provide the empirical response to research question one.
Whereas the response to the first research question required a comparison of the whole databases in order to identify major differences in the representations, the second research question dictated that individual terms be compared, looking for differences in their usage that might point to differences in their conceptualisation by the two groups.This presented an interesting challenge because the green business database was, as already mentioned, almost devoid of vocabulary that made a representation of the fleshy and fibrous natural landscape.The only vocabulary of the natural landscape which was used frequently by both groups was that belonging to the socially-constructed natural landscape.Terms taken from this group of words were, therefore, selected for usage comparison.
The technique used for comparing the usage of particular words by the two groups was to examine their collocation.This is a linguistic term which refers to the words that appear in the close vicinity of some particular node word, in whose usage the researcher is interested.Work within this field of linguistics is revealing that patterns in the usage of words (which are observable in texts) correlate with meanings and with culture (Stubbs, 1997;Sinclair, 2004).The significant collocates of a node word can help us to interpret how the user of the word conceptualises it.The Wordsmith tool was instructed to find all of the occurrences of a selected node word, e.g. the occurrences of the word biodiversity in the green business database.It then ran a listing of all of the words that appeared within a horizon of + or -five words from the biodiversity node word; in the green business database this amounted to 809 different words.It then performed a statistical calculation known as an MI procedure which is intended to indicate the extent to which the appearance of a collocating word in the vicinity of the node word is unusual.In order to calculate this measure of 'unusual' co-occurrence, Wordsmith needed to have a reference database and this was provided by the BNC.A full explanation of the MI procedure is available in Brown 2008 (pp. 219-223).However, a simple way to conceptualise the results of this procedure is to think of it as a view, from somebody who speaks and writes 'average' British English, of the ways in which the word biodiversity is used by British green corporations in markedly different ways from the average user's usage of biodiversity.In the list of the 809 collocates of biodiversity, the MI calculation, using the BNC as the reference, led to 42 collocates being assigned a positive MI and 23 a negative one.But the overwhelming number; 744 out of the total of 809 collocates, were assigned an MI of zero by Wordsmith, indicating that their co-occurrence with biodiversity in the green business corpus was not significantly different when compared with the BNC.
In order to further ensure the rigour of the MI procedure, I followed the more demanding rule of thumb that an MI value had to exceed +/-3.0 in order for its significance to be considered as sound.This further reduced the number of significantly unusual collocates around the node word.Finally, the listings were edited to remove proper nouns such as the names of the corporations or the NGOs, simply because they contributed nothing to the analysis.To illustrate the end-result of the procedure, the report of significant unusual collocates of the term biodiversity when used by green business contained just 16 terms ranging from the most unusual co-occurring collocate, sites with an MI of over 17 down to an with an MI of just over 3.0.A similar report for the NGOs contained 39 significantly unusual collocates, with degradation at the top of their list.These types of results, presented graphically in target diagrams, provide the response to research question two which comes in the next section.

Results
Research question one: What are the differences and similarities in the language used to make representations of the natural landscape and the actors' relationship to it by (i) the environmental NGOs and (ii) the green business corporations?
The keyword listings for one-, two-and three-word keywords for both green business and the NGOs are too large to be included in their entirety in this article but may be found in previously published work (Brown, 2008, pp. 381-389 & 511-516).From an examination of these listings, the following findings are important in terms of responding to research question one.
First, as mentioned in the method section, from an examination of its top 500 one-word keywords, green business's representations of a fleshy and fibrous natural landscape are almost negligible.In contrast, the NGOs have a great many such words in their top 500 one-word keywords.Table 1  The words in the table are one-word keywords found in the top-500 listing for the two groups.The rank number alongside indicates the word's position in the keyness ranking; the smaller the number, the higher it is in the ranking.The keywords for green business have been placed alongside words from the NGO listing with a roughly similar ranking in order to give an impression of their importance.In order to save space, the table has been halved in length and doubled in width so that the higher-ranking keywords for the two groups are presented on the left and the lower-ranking keywords on the right.
The first observation to be made regarding these findings is to concede that there is a certain arbitrariness about the selection procedure.There is, of course, no clear-cut distinction in the real world between my classification of (a) fleshy and fibrous and (b) socially-constructed terms.The attempted delineation has been to look for a reference that speaks of organic things or their lives.Thus, terms such as poor and indigenous, although social constructions that are placed on groups of people, are also included.
These findings do not surprise when one reflects over the respective practices of the agents which have produced the representations.The NGOs speak on behalf of the natural landscape, whether it is inanimate water quality or the lives of indigenous tribes people in the rainforest.Even though the discourse may be composed from an office in London, the NGOs will position themselves in the landscape, alongside whatever is suffering from the negative effects.Part of their discourse must attempt to give a voice to the voiceless by providing, for example, eye-witness accounts of suffering.
Green business, on the other hand, speaks from corporate headquarters, at least in the texts that I have studied.
Clearly, there will be texts in circulation within the corporation that may come closer to the natural landscape than the evidence here suggests.One can imagine that reports describing the situation from managers who are on the spot will be sent in to senior management at headquarters and that these would probably use language representations which are closer to the fleshy and fibrous.However, the texts in the green business language database are statements of each corporation's practice.At this overview level the fleshy and fibrous vocabulary is not suitable for making representations of a structured corporate approach to environmental problem solving.For that, a socially-constructed representation of the natural landscape, as can be seen in table two, is more useful.
The second important finding from the examination of the keywords is that both groups -not just green business -use words that make socially-constructed representations of the natural landscape.The words in the table are one-word keywords found in the top-500 listing for the two groups.The rank number alongside indicates the word's position in the keyness ranking; the smaller the number, the higher it is in the ranking.The keywords have been placed alongside words from the other listing with a roughly similar ranking in order to give an impression of their relative importance.
The findings in table two are of potential interest because of the extent of the overlap in this vocabulary; approximately 80% of the terms listed in table two are used by both groups.From a communications point of view, these words might provide the first requisite for a dialogue between stakeholder and corporation.Here, at least, is a common vocabulary which might be used as a vehicle for sharing common meanings.Whether or not the two groups use these terms with the same intended meanings is a matter to which I return in research question two.
Continuing the theme of a vocabulary used by both groups, the third important finding is the massive presence of terms that represent a process of managing.These terms make few direct reference to the natural landscape which is the focus of research question one.However, their ubiquity in the keyword listings of both groups suggested that they would nonetheless have some relevance to the ways in which the NGOs and green business represent their relationship with the natural landscape, so evidence of this vocabulary has been included.Words that may have a role to play in representing these processes can be found in the one-word listings.However, the intended meaning of these one-word signs can be ambiguous and presenting such a list might be misleading.In the two-and three-word keyword listings the terms used to represent the process of managing stood out more clearly.For that reason, it is these that I have selected for presentation.A complete list is provided in the appendix as table 3.1, but to save space, table three contains selected examples.The base listing is of the top 200 two-word and top 100 three-word keywords.In order to make the discussion easier to follow, the listings are presented differently to tables one and two.In table three there is a list of terms which are common to both the NGOs and to green business.Then there are listings that are unique to the NGOs and then to green business.In order to save on space, they have been presented as lists separated by commas in normal text.They have also been organised alphabetically.In the heading for each of the three subsections is an indication of how many words there are in the full table in the appendix.
If one includes the ten common two-and three-word keywords, this field of representation accounts for 30% (91/300) of the top 300 green business keywords.A few of them refer to standards and objectives that have their origin in agencies external to the green businesses.But the vast majority of the keywords refer to the internal business processes by which the corporation manages itself.The list of keywords used by the NGOs is shorter, and there are two strands of representation.First, there is evidence of an appeal to government and international agencies, to find ways of controlling the activity of business, as well as a moral appeal to the concept of environmental justice.The need for corporate control is also revealed by terms such as accountability and compliance.Second, the NGOs focus more on the damaging effects on the natural landscape, whereas the green corporations represent their actions to report on the effects, analyse the problems and their efforts to rectify them.
In common with the findings in table two, however, the evidence from table three also suggests the possibility of establishing a dialogue of meaning between stakeholder and corporation.
The fourth important finding is also presented for a similar reason to the third.The findings in table four illustrate a vocabulary which makes representations of threats to the natural landscape and risks for its welfare.
The words in the table are one-word keywords found in the top-500 listing for the two groups.The rank number alongside indicates the word's position in the keyness ranking; the smaller the number, the higher it is in the ranking.The keywords have been placed alongside words from the other listing with a roughly similar ranking in order to give an impression of their importance.In order to save space, the table has been halved in length and doubled in width so that the higher-ranking keywords for the two groups are presented on the left and the lower-ranking keywords on the right.The first observation to make about these findings is that all eleven of the terms that are found in the green business keyword listing are also present among the 34 words that may be found in the NGO keyword listing.Once again, there is a common vocabulary.The second observation is the more subjective assessment that the eleven words that are used by green business are the more neutral terms of the 34 whereas the 23 that are not found among their keywords are more emotive.Again, when one considers corporate and NGO practice, this should not really surprise.It is part of the role of the NGOs to act as eye witnesses to events, speaking up for the victims, and such close-up language will be more emotive.Green business, on the other hand, views events from headquarters and sees the injuries and damage at a distance which reduces the unpleasantness.Secondly, green business may simply feel uncomfortable about using the more emotive terms to make representations of effects for which they may well be responsible.
In summary, the four sets of representations from the texts of the NGOs and green business illustrate differences and similarities, all of which are plausibly justified by their respective practice.In terms of establishing a meaningful dialogue, the representations of a fleshy and fibrous natural landscape manifestly do not provide a platform, for the simple reason that the corporations do not use them.Terms within the socially-constructed natural landscape are, on the other hand, used by both actors and might enable meaning to be shared between them.
This leads to research question two: What evidence does the language provide which can indicate how the groups conceptualise the natural landscape and/or their interaction with it?For reasons of space, this section presents the findings for just three terms of the socially-constructed natural landscape: biodiversity, health, and habitat-s.More evidence is presented in earlier work (Brown, 2008, pp. 553-565).As explained in the method section, the unusually significant collocates clustering around these three node words were identified using the MI statistical analysis in Wordsmith.In addition to the representations of the socially-constructed natural landscape, which is the focus of analysis, the three other collections were utilised for the classification of the collocates.First, there are the representations of a process of managing, for which yellow background shading has been adopted.As mentioned previously, the ubiquity of these words points to their having an important role, particularly in the green business contextualisation of the natural landscape.Second, with the contextualisation of the radical NGOs primarily in mind, terms representing risks for the welfare of the natural landscape have been shaded in red.In addition to the words listed in table four, a few others have been shaded red.On a few occasions within the listings for the radical NGOs, there are collocates which describe either the activities or the products of corporations.In some cases, such as transport and trade, where the threat to the natural world was not so obvious, the word has been left without shading.However, in others, where the radical NGOs' representation of a threat is clear, such as mining, logging and oil, a red background shading has been applied.Third, since the three node words are all representations of the socially-constructed natural landscape, it was interesting to see the extent to which the fleshy and fibrous natural landscape is represented among the collocates.
There is a minority of words which, not surprisingly, do not fall into any one of the three fields of representation.This is entirely consistent with our own common sense knowledge of language, and they have simply been left without shading.
The target diagrams which I designed in order to illustrate differences in the usage of a word, are all organised in exactly the same way.There are four concentric circles which delineate the same MI values.In the central section are the most significantly unusual contextualising collocates of the node word.In order to qualify for the bull's eye, they have an MI greater than or equal to 10.0.From the perspective of the BNC benchmark, it is the frequency of occurrence of these words in the vicinity of the node word, which is most remarkable.Moving out from the centre, the next circle contains collocates with an MI between 7.0 and 10.0, the third is between 5.0 and 7.0, and the fourth is between 3.0 and 5.0.The selection of these band widths was made on the basis of a review of the diagrams, the objective being to divide up the lists so that the collocates were spread out over the whole target.The placing of the collocates within each of the three circles outside of the bull's eye, follows a consistent pattern; the collocates with the highest MI coefficient within a band, are placed at twelve o'clock in the target diagram, and the rest of the words follow in a clockwise direction.Apart from the fact that the one immediately before will have a greater MI coefficient, and the one immediately after a smaller MI coefficient, their respective placing around the 'clock face' is not significant.The colour shading in figures one to three communicates well, but is only visible in the online version of this article.The most immediate impression created by the three figures is that it is the NGOs which produce the greatest number of unusually significant collocates around the node word.From the perspective of the 'average English' of the BNC benchmark, it is the NGOs which contextualise these three words of the socially-constructed natural landscape in the most unusual way.One possible explanation is that this contextualisation confirms the role which the NGOs play in being a radical voice speaking up for the natural landscape; they make representations in language which are different from the mainstream.
The second general observation to be made is that the NGOs' diagrams are more colourful.This can be explained by three non-exclusive alternatives.First, although it may not be fleshy or fibrous, the socially-constructed natural landscape does suffer the impacts of business activity just as the former does.One might expect to see representations of concern, therefore, in its contextualisation.Second, concerns about, say, biodiversity may be illustrated by reference to the details of crops, farmland and wildlife.This might explain why there are representations of the fleshy and fibrous natural landscape among the NGOs' significant collocates.Third, the NGOs may also make use of the representations of managing in order to present the degree of damage to the natural landscape, in a language which is by green business.
As for green business, the contextualisation of these three terms displays a strong presence of representations of managing.The corporations need to be able to monitor their own effects on the natural landscape.They also need to formulate the standards of operation to which they should aspire, if they are to take their responsibilities seriously.Finally, they need to set themselves objectives, make implement the plans, and then report back, on the progress that is being made towards the objectives and the, hopefully, reduced impacts on the natural landscape.

The Rhetorical Dilemma
The implication of these findings is that NGOs which speak up for the natural landscape face a rhetorical dilemma.They can choose to continue in their traditional role of providing eye witness accounts of the destruction of the fleshy and fibrous natural landscape by the activities of business.They can broadcast their message with this radical voice and some of their listeners will understand the meaning that the NGOs wish to convey.However, green business will not understand the meanings that the NGOs are attempting to communicate.
This has important implications for NGO practice.If they wish to establish a meaningful stakeholder dialogue with green corporations they will have to engage them with a language that is meaningful to the corporations.Such a language adopts a managerial approach to represent its relationship with the natural landscape.This is qualitatively different from the preservationist stance that NGOs have traditionally taken.Their dilemma, then, is that in establishing a more productive, meaningful dialogue with green business, they become party to the management of natural landscapes rather than to their preservation.Management of the landscape is not necessarily negative; one could envisage a form of management that approached stewardship.Some NGOs may choose, therefore, to establish a meaningful dialogue with business which is based on the managerial discourse, hoping that they can influence business towards a stewardship role.
However, once absorbed into a system of planning, implementation, monitoring and control, the natural landscape is prey to a corporation's instrumental objectives, so one can understand NGO hesitancy to join the dialogue.Others will prefer to maintain the integrity of their traditional discourse of giving a voice to the voiceless.Rather than working through stakeholder dialogue, they will seek other ways to influence the activities of business.But in refraining from participation, they run the risk of being marginalised.Returning to the mutually-constitutive relationship between discourse and practice, their rhetorical dilemma is just as much a strategic question of practice for environmental NGOs which wish to speak up on behalf of the natural landscape.

Figure
Figure one compares the usage of biodiversity between the NGOs -on the left -and green business -on the right.Note that the NGOs' contextualisation is more colourful than that of green business.The strong primary colours of red, representing risks and threats, and bright green, representing the fleshy and fibrous natural landscape, stand out very clearly and draw attention to the paucity of this representation in the target diagram for green business.Note, too, how the only two red-shaded words on the right are the bland WASTE and IMPACTS, whereas the NGOs' target diagram contains degradation and protection in the bull's eye, and other words are damage, pollution, threat and destruction.In the green business diagram the leading colour among the collocates of biodiversity is yellow; representations of the process of managing.In the bull's eye four of the six words have yellow shading.The first of these is BAP, an acronym for Biodiversity Action Plan.

Figure 1 .
Figure 1.The significant collocates of biodiversity presented on target diagrams

Figure 2 .
Figure 2. The significant collocates of health presented on target diagrams

Figure 3 .
Figure 3.The significant collocates of habitat-s presented on target diagrams

Table 1 .
presents the findings.Keywords which represent the fleshy and fibrous natural landscape in the top 500 listing

Table 2 .
Keywords which represent the socially-constructed natural landscape in the top 500 listing

Table 3 .
Selected list of keywords which represent a process of managing Two-and three-word keywords which represent a process of managing Common to both NGOs and green business (6 terms of total 10) ACTION PLAN, CODE OF CONDUCT, CORPORATE SOCIAL RESPONSIBILITY, ENVIRONMENTAL IMPACT, ENVIRONMENTAL STANDARDS, GLOBAL COMPACT

Table 4 .
Keywords which represent risks for the welfare of the natural landscape in the top 500 listing