Rhetorical Strategies Employed by Big Oil in the Context of IPCC Reports of Climate Change

Abstract

Despite long-standing evidence linking fossil fuel combustion to greenhouse gas and climate change effects, and the growing advocacy for reductions and regulatory limits on their use, fossil fuel corporations remain hugely profitable and influential. In response to scientific evidence linking Big Oil’s corporate activities directly to climate change impacts, tactics favoured by Big Tobacco to medical evidence linking smoking to cancer appear to have also been adopted by Big Oil in responding to the Intergovernmental Panel on Climate Change (IPCC) findings. To examine some of these response strategies, a bespoke corpus was compiled from sustainability reports by a sample of three Big Oil corporations over a twenty-year period corresponding to the IPCC’s publication of the third through sixth Assessment Reports. This corpus is statistically and linguistically analyzed for representations and accounts by Big Oil for its activities and how, if at all, scientific evidence is addressed linking fossil fuel extraction and use to the findings of the IPCC. By highlighting corporate response strategies and preferred narrative accounts to the IPCC evidence, the aim is to equip policy- and decision-makers with key insights to develop more effective counter-narratives to facilitate scientific communications in this critical policy space.

1. Introduction

The Paris Agreement of 2015 was a multi-lateral commitment to limit global temperatures to 1.5 °C above pre-industrial levels. However, since then both greenhouse gas emissions and global temperatures have continued to rise, as confirmed by the Intergovernmental Panel on Climate Change (IPCC) in the 6th Assessment Report (AR6), which has stressed the urgent need to intensify global efforts towards decarbonization [1]. Many countries have set ambitious targets for reaching net-zero emissions by 2050, an ambition which will require a whole systems paradigm shift [2]. In practice, these targets translate to a steep reduction in the total energy supply percentage share of oil from 29% in 2020 to 8% in 2050 (with non-energy use of oil accounting for 5% of the supply). The share in renewable energy, on the other hand, will have to rapidly increase from 12% in 2020 to 67% in 2050 [2]. These are clearly seismic-scale ambitions necessitating a complete rethinking of how we generate and use energy. Although countries have been pursuing decarbonisation efforts, decoupling of carbon emissions from economic growth has not been achieved globally. Most countries in Europe and Oceania are showing a reversing trend [3]. However, these econometric studies only focus on direct emissions, and the conclusion will change if indirect emissions are considered.

Decarbonization will be particularly challenging for oil companies as they face the prospect of a rapid declining share in the energy market. However, international oil companies (IOCs) such as BP, ExxonMobil, Chevron, Shell and Total are more vulnerable compared to their national oil company counterparts because of two factors: (1) their costly resource base compared to national oil companies (who hold low-cost resources) and (2) the stakeholder pressure for climate actions, as they are publicly traded companies [4]. These companies have also seen mounting pressure from asset managers who are asking IOCs to recognize the risks of climate change in terms of their operations and to reflect on the financial impacts on the long-term value of assets [5]. IOCs are responding to these demands through progressive plans for repositioning their business activities and investment commitments. For example, BP committed to invest in 50 GW of renewable energy capacity by 2030, while Shell has set a target of investing US$3 billion by 2030 [5].

But the petroleum industry has generally been a very lucrative business for IOCs despite facing a volatile oil market in recent times, particularly during the economic downturn of the COVID pandemic. A recent study estimated that the pure rent from the oil and gas industry averaged US$1.03 trillion per year (in constant 2020 prices) between 1970 and 2020 [6]. The return on investment in the petroleum industry remains high compared to alternative investment options and the windfall profit amassed by IOCs since Russia’s attack on Ukraine received global attention. Estimates suggest that in 2022, after increases in gas prices linked to the invasion of Ukraine rose to ~US$28bn, BP had already begun to scale back its commitments to reducing greenhouse gas emissions [7]. From the perspective of profitability and value creation for shareholders, the temptation to continue with the business-as-usual scenario remains high, a tendency exacerbated by the relatively short-term focus of private investors when compared to the long-term perspective required for climate change decisions. At a number of the Conferences of Parties (COPs), specifically the Conferences in Egypt (COP27), Dubai (COP28), and Baku (COP29), there is well-documented evidence of oil and fossil fuel coalitions actively working towards undermining the decarbonization agenda. Additionally, hedge fund managers have progressively pushed IOCs to drop any commitments to green energy in favour of more traditional (and profitable) business models [8].

Evidence shows that, despite various promises and commitments to disinvest from fossil fuels, fossil fuel-linked greenhouse gas emissions continue to rise, with consumption rising by 1.5% in 2023 relative to 2022, leading observers to note that “fossil fuel emissions have increased to an all-time high” [9] (p. 1). Even though investment in renewables reached its highest on record in 2023 [10], fossil fuel consumption is still approximately 14 times “greater than solar and wind energy consumption”, but this is due to demand and not because renewables are “replacing fossil fuels” [9] (pp. 2–3). Despite numerous policy efforts, however, fossil fuel CO2 emissions continue to increase, albeit compensated for by declines in land-use change CO2 emissions, leading the Global Carbon Budget [11] (p. 5) to conclude that “there is still no sign of the rapid and deep decrease in total CO₂ emissions that is needed to tackle climate change”.

These data underscore the hypocrisy on the part of IOCs, which is also evident from a review of their investment plans. Despite IOCs officially agreeing to support the Paris Agreement, there has yet been no restraint on investments in the petroleum industry. British Petroleum (hereafter, BP) plans to invest US$17 billion by 2030 in oil development and another US$13 billion in gas. Royal Dutch Shell (hereafter, Shell) intends to invest US$18 billion in oil and US$28 billion in gas by 2030 [12]. IOCs continue to plumb the future prospects of the petroleum industry at the cost of the global climate. Even as recently as February 2025, BP reversed its commitments to green energy, claiming a fundamental reset to invest more deeply in oil and gas while blaming a misplaced optimism in renewables [13]. This is not a trade-off made in some fog of ignorance. This is a deliberate investment in the drivers of global warming because it is well established that the fossil fuel sector has long been aware of the damages associated with its products. Studies in the late 1950s already linked carbon emission contributions to the ‘greenhouse effect’ [14,15], a phenomenon identified in the late 1800s, and in 1958, the American Petroleum Institute (API) itself funded a study to respond to growing concerns that “air pollution would eventually become a major problem for the petroleum industry” [16] (p. 268). As the six assessment reports published by the IPCC to date already confirm, the impacts of the petroleum industry far exceed mere ‘air pollution’ and instead now position humanity on the brink of an existential catastrophe through an unprecedented disruption of the atmospheric chemistry that supports life on earth.

In 1986, evidence given to the US Congress documented data linkages between greenhouse gas emissions from fossil fuel use with accelerated climate-scale changes [17]. Evidence provided during that hearing modelled significant scales of disruption to quality of life resulting from the continued combustion of fossil fuels. In the same year, the concept of Sustainable Development was introduced to the public discourse and rapidly became the lodestar in planning for a world that enjoyed the benefits of nature, society and the economy without denying future generations an equal right to share in a similar and equal opportunity. In 1990, the Intergovernmental Panel on Climate Change (IPCC) published the first Assessment Report (AR). This inaugurated a non-partisan collective of climate scientists, combining their models and findings under the aegis of the United Nations (UN) and making the findings of these publicly available. In their first report, and in every subsequent IPCC publication, the fossil fuel industry has found to be front and centre as the cause of global warming.

The central argument is that the consumption (i.e., combustion) of Big Oil’s principal product increases the risk of destabilizing planetary-scale social and ecological systems, because the burning of fossil fuels directly changes the global atmospheric chemistry due to the release of massive volumes of carbon dioxide and other greenhouse gases [18,19]. Studies argue that emissions from IOCs contributed approximately 48% of the gradual increase between 1901 and 2021 in vapour pressure deficit, a precursor causally linked to wild fires and dry conditions putting forest biomes at risk of subsequent fire events [20]. These intensified risks to forest ecosystems are reaching catastrophic tipping points and have been identified to be the primary causes for the sudden escalation in wildfires during 2023 and 2024 [21]. With the surfeit of evidence available, it is inconceivable that IOCs are ignorant of the causal relationships between their product and global warming, with the widespread collapse of global ecosystems such heating foments. Rather, this is a clear and incontrovertible case of profits trumping human interest.

There is qualified support for the Environmental Kuznet Curve (EKC) hypothesis, which posits that as an economy develops, its environmental impacts will be high initially but will taper off over time as the economy matures [22]. But as studies repeatedly report, our emissions are not coming down, and our footprint keeps expanding. There is a clear case to be made for decoupling emissions from economic growth and pairing market liberalization with strong safeguards (e.g., carbon pricing) to align economic policies with sustainability necessity [22]. Nevertheless, despite multiple agreements and directives that endeavour to reign in corporate excess and irresponsibility, aligned with UN Agenda 2030, the European Green Deal, and other mechanisms, these have had little success. For example, an ambitious undertaking by the EU, the Directive 2013/34/EU, stipulated a legally binding duty to report on non-financial matters in corporate management reports and included a clearly defined taxonomy of subjects to report on. However, the sanction mechanism for maintaining compliance remains inadequate [23].

There are parallels in the history of corporate culpability for public health damages, for example, the case against Big Tobacco and the medically confirmed causal link between its product and lung disease and cancers [24]. Then, as now, corporations engaged in deliberate strategies of obfuscation, misinformation, and dragging challenges to evidence out in the courts for years. It is therefore instructive to locate the rhetoric of Big Oil within the historical context of similar tactics employed by Big Tobacco. When the links between smoking and negative health impacts emerged in the 1950s, one of the first counter-steps Big Tobacco took was to invest significantly in Corporate Social Responsibility (CSR) campaigns. In the 1960s, the US Surgeon General published the first report summarizing a decade’s worth of data causally linking tobacco consumption (smoking, especially) with cancer, particularly lung cancer. In the late 1960s, a tobacco industry memo revealed the core of the sector’s response strategy: “Doubt is our product since it is the best means of competing with the “body of fact” that exists in the mind of the general public. It is also the means of establishing a controversy.” [25] (np) One of the ways such industries sow doubt is, like Big Oil today, to argue that bodies such as the Environmental Protection Agency (EPA) in the US use their influence to ensure science fulfils political goals. In 1993, for example, Philip Morris claimed that the “EPA’s recent report concerning environmental tobacco smoke allows political objectives to guide scientific research” [26] (p. 4).

As revealed during the tobacco industry investigations, led by Representative Henry Waxman in the 1990s, the sector also defined its own disclosure processes for reporting on environmental impacts while simultaneously establishing sustainability awards on behalf of third parties for the tobacco sector. Far from the benign or even progressive step this may at first appear to be, however, such environmental goals with which they report compliance are those that they themselves have set, many of which are tangential to the actual sustainability or environmental impacts of their products. Both Big Tobacco and Big Oil push for voluntary standards, thereby deflecting external regulation by providing a veneer of accountability despite there being an absence of industry-wide standards by which data disclosure follows. In this way, the sector cultivates the impression of legitimacy, and the sector (and products) is presented as socially and environmentally responsible. The advantage of this tactic is that it helps to off-set and even pre-empt potential criticism, and because these measures are already in place, this makes political calls advocating for external regulation more difficult to support. However, industries that are capital-intensive may even welcome regulation, because it can be used to whittle down competitors while portraying their industry as respectable and a good corporate citizen in the public’s perception.

Finally, because companies are not legally required to take responsibility for all environmental impacts associated with the life cycle of their products, Big Tobacco placed the responsibility for the waste associated with its products (e.g., cigarette filters) on the shoulders of the consumer and local governments, which must clean these up and dispose of them properly, even though they are not biodegradable [27]. As will be shown, a similar tactic is still employed by Big Oil in the general diffusion of responsibility for both use and bringing about change. These deflections of responsibility save the sectors considerable sums of money by avoiding responsibility for the full environmental impact costs associated with acquiring the raw materials, product processing and manufacturing, and post-consumer waste cleaning and recovery.

In summary, the tactics employed by both Big Oil and Big Tobacco to avoid censure of their practices and products, as well as steep financial risks associated with increased scrutiny and accountability, are:

• Buying political inaction, primarily through lobbying.
• Developing globally distributed denial and delay networks of like-minded industrial advocates, e.g., ‘think tanks’, such as the Heartland Industry funded by the Koch brothers and the right-wing neoliberal Institute of Economic Affairs on Tufton Street, London.
• Using these networks and sympathetic media to broadcast messages that sow doubt as to the legitimacy and validity of the scientific backing for claims of harm, thereby fostering public distrust of science, more critically inclined media and government.

In a similar vein, Big Oil’s entire raison d’etre has been shown to be a cause of global threat, and the nature of this sector’s response warrants scrutiny. Unfortunately, recent reports from the US Oversight Committee observe that, despite Big Oil’s rhetoric in support of transitioning to meaningful investments and their advertised commitment to addressing climate change at source, this amounts to little more than disinformation [28].

Big Oil has been linked to multiple efforts to call the science behind climate change into question, to actively fund disinformation campaigns, and to vigorously lobby politicians to dilute the legislative responses that would reign in their corporate interests [29,30,31,32,33,34]. Arguably, there is also a sustained disconnect between Big Oil’s internal communications and their public-facing reports. Shell, for example, noted in its 1991 film ‘A climate of concern’ that waiting for ‘iron-clad’ proof of the causal linkages between the fossil fuel sector and climate change was irresponsible. Despite this, the company has continued to hold out for such proof for several years hence [35]. Research clearly shows that Big Oil is well informed as to its culpability in driving climate change. Nevertheless, how IOCs have elected to respond to the escalating threats from global climate change is by drawing strategically on obfuscation, misdirection, and misinformation. Taken together, IOCs have strategically opted to undermine public and political efforts to hold those responsible to account. This is not corporate negligence but culpability.

We anticipate tensions in how IOCs present themselves in the context of climate change policy, because not only do their share holders profit significantly from fossil fuels and hence lack financial motivation to change their business model, but this model also literally powers the socio-economic enterprise of modern society, and there is significant systemic lock-in, making change difficult and costly. Nevertheless, the evidence is incontrovertible that a business-as-usual scenario will result in an ever-increasing volume of greenhouse gas emissions which drive global warming. To resolve such tension, IOCs either have to change their business model, at significant financial risk to shareholders (albeit potentially triggering an employment and national GDP boost via the ‘green jobs’ economy), or give the impression that they are changing their model while continuing to generate profits for their share holders. This latter option risks them being found to be bad corporate actors and potentially subject to censor, an outcome that occurred when a 2021 Hague court ruled against Shell for its investment in the extraction of shale gas which, the Court determined, breached the terms of the Paris Agreement. Shell responded by claiming that it was “investing billions of dollars in low-carbon energy, including electric vehicle charging, hydrogen, renewables and biofuels” [36]. Responses like these are part of a widespread pattern.

Despite such claims of investment in low-carbon enterprises, a study by Climate Trace observes that “emissions from oil and gas facilities around the world are about three times higher than their producers claim” [37] (p. np. Added emphases). While the public-facing narrative expresses concern for sustainability and global welfare, the actions of Big Oil corporations are characterized by a marked failure, even unwillingness, to address the causes of climate change. Reports that BP are planning to invest double the amount it spends on renewables into so-called ‘resilient hydrocarbons’ (i.e., oil and gas) suggests that, despite their green rhetoric, IOCs have elected to resolve this tension between words and action by continuing to follow a business-as-usual approach [38]. Using a scale of alignment between corporate activities and the Sustainable Development Goals (SDGs) [39] (p. 2396), IOCs fall between (ii) “mixed activities” because they “have moderate/high degrees of both negative/positive impacts, posing a decoupling imperative” and (iii) “opposed activities” which “provide few benefits yet cause significant adverse impacts, implying that companies must transform in order to better align with the SDGs”. It is undeniable that fossil fuel consumption has been the driver for human socio-economic progress, and paradoxically this same dynamic and economic lock-in are going to undo all that progress.

In 2004, BP published one of the first sector reports explicitly detailing the corporate response to growing concerns about global sustainability and climate change, with Shell publishing their own sustainability report in 2005. Since then, both of these IOCs have published such reports annually, while ExxonMobil released its first report only six years later in 2010, and many of their subsequent reports have been ‘highlights’ only. This study linguistically examines the 53 sustainability reports published between 2004 and 2024 by BP, Shell, and ExxonMobil as three IOC representative corporations (see

Table 1. Summary of report data set (the corpus).

Year	Reports
2004	BP i
2005	BP ii, Shell iii
2006	BP iv, Shell v
2007	BP vi, Shell vii
2008	BP viii, Shell ix
2009	BP x, Shell xi
2010	BP xii, Shell xiii, ExxonMobil xiv
2011	BP xv, Shell xvi, ExxonMobil xvii
2012	BP xviii, Shell xix, ExxonMobil xx
2013	BP xxi, Shell xxii, ExxonMobil xxiii
2014	BP xxiv, Shell xxv, ExxonMobil xxvi
2015	BP xxvii, Shell xxviii, ExxonMobil xxix
2016	BP xxx, Shell xxxi, ExxonMobil xxxii
2017	BP xxxiii, Shell xxxiv, ExxonMobil xxxv
2018	BP xxxvi, Shell xxxvii, ExxonMobil xxxviii
2019	BP xxxix, Shell xl
2020	BP xli, Shell xlii
2021	BP xliii, Shell xliv, ExxonMobil xlii
2022	BP xlv, Shell xlvi, ExxonMobil xlvii
2023	BP xlvix, Shell xix, ExxonMobil xlviii
2024	BP xlix, Shell l

for details), a period coinciding with the IPCC releasing their fourth through sixth Assessment Reports. It seems reasonable to assume that IOCs were aware of the three previous ARs, along with significant evidence from other studies, including those by its own scientists, and hence an ignorance defence for their continued actions is implausible.

It is tempting to concur with research that corporate actors representing sectors with high levels of social-environmental impacts (negative externalities) are more likely to implement sustainability measures to demonstrate impact mitigation. But there is also a compelling counter-argument that such corporations are simply under more pressure to give stakeholders the impression that they are mitigating impacts in order to garner more positive stakeholder evaluations. The phenomenon is expressed in the proposition that “[c]ompanies from industries associated with negative externalities will engage with more SDG targets than companies from industries with more positive externalities.” [40] (p. 216)”. In this paper, we consider the way that these representative IOCs use language to assert their preferred narratives, which characterizes them as concerned corporate citizens, doing their best to preserve human life and to support the well-being of global populations. This characterization is at odds with both the sector’s historical obfuscation and muddying of waters when it comes to adjudicating their complicity in the escalating crisis of climate change, as well as with the pattern of Big Oil reneging on its own commitments to take meaningful actions to mitigate climate change [37]. While the content and narratives espoused in their sustainability reports characterize the corporations as responsible actors, Big Oil continues to put profits before the planet and people. By foregrounding the linguistic strategies deployed by these corporations, we hope to contribute to the growing body of evidence exposing IOCs to be manifestly bad actors. Here we highlight how IOCs use public relations and appropriate-sounding terminology to conceal their activities, to win over political and public sympathies, all the while continuing with a business-as-usual model of economic advantage over global ecosystemic impacts.

We contend that studies of this sort contribute to a wider body of critical corporate ethics research holding corporations accountable for practices with far-reaching ecological, social, and economic consequences. This orientation is a basis for an ethically informed framework as potential leverage to bring about meaningful change in how business is undertaken, facilitating opportunities to endorse a positive business contribution to sustaining coupled social-ecological systems. Given space limitations, however, we restricted our comments to exploring a small sample of strategies and tactics employed by the Big Oil sector sample.

Our paper is structured as follows. In the following section, we introduce and briefly discuss the materials and methods used in this study. Here, we introduce the three methods (qualitative review, text-mining, and corpus linguistic methods) used for the analysis and review their contribution to the study. The third section presents results from the computational and corpus linguistic analysis of the sample corpus, interpreted via the framing of strategic reputation management and sustaining existing business practices. Here we take a closer look at the public-facing narratives that present an account of how IOCs are engaged in the attempt to mitigate the effects (and causes) of climate change, despite the corporate commitment to business-as-usual strategies. We conclude by discussing some of the implications and potential applications to which this study contributes.

2. Materials and Methods

This study is concerned with the public-facing narratives expressed in sustainability reports published by a sample of three of the five International Oil Companies (IOCs) (i.e., Shell, BP, and ExxonMobil). The five IOCs are defined as having significant production and refinement centres in fifteen or more countries with markets in 40 or more countries [41]. The sustainability reports provide accounts detailing how each of the three corporations concerned enact their social responsibility requirements expressed as actions taken to mitigate and reduce the sector’s contributions to climate change. Ostensibly, and throughout their public-facing sustainability reports, the narrative promoted is that IOCs are fundamentally good corporate citizens, committed to their social responsibilities with due diligence and accountability, while leading the call for sustainable energy development and provision. In these narratives, IOCs claim to make be making significant changes to the way that they do things. They offer themselves as having expertise in addressing and managing climate change, understanding the science, and proposing transition pathways. For evidence of this, IOCs point to sector-funded research into clean fuels, low-carbon energy, and so on, and describe their investment strategies intended to benefit particular communities, and report on their agreements with extant international treaties and the sustainable development goals (SDGs) to help bring about a better, healthier, and more sustainable world. Despite all of these activities and endorsements, commitments and ambitions, greenhouse gas emissions and IOC profits continue to rise in tandem.

Our study reviews the claims of this sample of IOC sustainability reports but, unlike other studies which tend to adopt a fact-checking approach to highlight the contradictions [30,31,42], here we are interested in the deployment of language in the construction of narratives positioning IOCs as good corporate actors, despite all evidence to the contrary. We examine the sustainability discourses enacted to manage and maintain IOC reputations through the textual and linguistic analysis of 53 public-facing IOC sustainability reports published by BP, Shell, and ExxonMobil over a twenty-year period.

While we employ three methods of analysis, our intent is to present the findings from these methods as thematically integrated. In this paper, we longitudinally examine the changes in the report narratives from the three sampled IOCs over time, year by year, rather than any changes used by the same corporation over time. Our rationale for this is two-fold. First, we can map these relative use frequencies against the time-line of major events in the scientific and political context of climate change responses and can discern alignment between corporate and contextual events. Secondly, by looking at change over time, rather than within a specific corporation, patterns in the overall development of reputation management narratives across the study sample become more apparent.

Our analysis begins with a qualitative exploration of the sampled sustainability reports. Initially, we wish to highlight some of the common rhetorical devices and preferred narrative frames. At the forefront of the analysis is the identification of how IOCs factor in the problems of GHG emissions inherent to their profit models and reconcile this with efforts to portray themselves as change leaders in sustainability. We find, for example, that IOCs accomplish this typically by referring to future technological solutions involving cleaner sources of energy. Yet, in responding to questions of responsibility for change, IOCs tend to position themselves as part of the solution while attributing responsibility for meaningful change to a wide cross-sector. These strategies are discussed in Section 3.1., below.

The second method uses text-mining to examine the relative use frequencies of key words throughout the corpus of 53 reports of the twenty-year sampling period. 53 PDF reports were downloaded and converted into plain text files using a batch conversion script written in bash 5.3 (https://www.gnu.org/software/bash/ (accessed on 18 August 2023)). The text files were read into a bespoke text mining R 4.5.0 [43] script, using the tm [44,45] and tidytext [46] packages. The sampled corpora are transformed into a Document Term Matrix (DTM) which converts any meaningful coherence of text into an unstructured ‘bag-of-words’, so that it can be analyzed statistically [47]. Words are lemmatized, reduced to their root form, e.g., ‘energy’ becomes ‘energi’, a lemma which also includes ‘energize’ and ‘energies’. This is helpful because it enables the analyst to track specific root form uses rather than multiple variations on that root form and permits the generation of term use frequency tables, in this case, sorted by year. We convert the raw count frequency into a percentage based on the total number of words in the corpus for that year to yield a relative weighting. This provides indicative patterns that may support and inform particular hypotheses, even if they do not support drawing any definitive conclusions. This part of the study is discussed in Section 3.2., below.

Finally, the third approach discussed in Section 3.3. uses two corpus linguistic methods, specifically collocation and concordance analyses. Corpus linguistic methods are warranted here because they help uncover the structure and usage patterns of large volumes of language data that will otherwise remain obscure from close manual reading and annotation favoured in more traditional approaches to discourse analysis. Doing so facilitates proposing empirically grounded insights about language use. The LancsBox X corpus linguistic software suite [48] enables the analyst to deploy the MI3 statistical measure of associative strength (using the GraphCol method in the LancsBox suite). Synonyms for the MI3 measure include ‘pointwise mutual information cubed’ and ‘log-frequency biased mutual dependency’ [49], but it is a bidirectional, contingency-based measure of association, favouring rare (i.e., unique) but statistically significant collocations between the key (search) word and associated (target) terms in the corpus. These relationships satisfy a statistical probability of co-occurrence (here, the threshold of significance is customarily set at 9.0). The presence of high-frequency collexemes (co-occurring terms) is associated with an acquired construction of the word and its meaning [50]. Put differently, the meaning of a word will be known by those with which it associates [51], so this analysis equips key terms with a certain semantic currency by means of how it is used in context.

The second corpus linguistic tool used is concordancing, a method that examines the work that key terms do in the context of the sentences within which they occur. This approach assumes that words are not randomly allocated in the construction of sentences, so consequently the ordering of words (key terms) in such a context plays a role in meaning generation. Taken together, these two corpus linguistic analytic methods provide insight into the statistically significant relationships among key words and their collexemes, and the contribution of each key word to the construction of meaning within its context of use.

By combining these two approaches of text mining and corpus linguistics, our analysis helps to highlight latent structures and patterns in the text not readily visible to the analyst, thereby contributing to a deeper analytic dive into the specificity of how each corpus is constructed and the meaning and contributions to the general character of the corpus from the key words of analytic interest.

In the following sections, we present the findings of the text mining and corpus linguistic analyses and discuss these with reference to the construction of public-facing narratives that adopt a motif of concern and due diligence with respect to corporate sustainability commitments.

3. Findings and Discussion

A total of 53 files from the publicly available sustainability reports published by three case study IOC were sampled, over a twenty-year period. These data are referenced in Table 1 (specific URLs are given in the Appendix A). This period converges climate change reporting and corporate claims by the Big Oil sector about their commitments to ensuring sustainability.

From Table 1, one notes that of the three IOCs sampled, BP was the first to publish a sustainability report in 2004. In 2005, Shell also began publishing a sustainability report which was subtitled a “public record of our progress in contributing to sustainable development”. Since then, both companies have published annual reports. In 2010, ExxonMobil published their first sustainability report, and did so annually until 2017, when it began thereafter to publish ‘highlight’ reports only. These are trimmed down versions of a report that, one assumes, has been published elsewhere, but does not appear to be publicly accessible. However, for 2019, 2020, and 2024 even these ‘highlight’ reports were not published. Of the 53 reports sampled, 40% are from BP and 38% are from Shell, while only 23% are by ExxonMobil (These percentage values are rounded up), the majority of which are, as noted, ‘highlight’ summaries only.

Table 2. Key climate change milestones and other world events 1990 to 2022.

Event	Year
First Assessment Report IPCC	1990
Second Assessment Report IPCC	1995
COP 1, Berlin, Germany. United Nations Framework Convention on Climate Change (UNFCCC) which meet formally as the Conference of the Parties (COP) COP 2, Geneva, Switzerland	1995
COP 3, Kyoto, Japan	1996
COP 4, Buenos Aires, Argentina	1997
COP 5, Bonn, Germany	1998
COP 6, The Hague, Netherlands	1999
Third Assessment Report IPCC	2000
COP 6, Bonn, Germany COP 7, Marrakech, Morocco COP 8, New Delhi, India	2001
COP 9, Milan, Italy	2002
COP 10, Buenos Aires, Argentina	2003
COP 11, Montreal, Canada	2004
COP 12, Nairobi, Kenya	2005
Fourth Assessment Report IPCC	2006
Global Financial Crisis begins COP 13, Bali, Indonesia COP 14, Poznań, Poland	2007
Global Financial Crisis peaks COP 15, Copenhagen, Denmark	2008
COP 16, Cancún, Mexico	2009
Disaster at the BP-operated Deepwater Horizon offshore drilling rig off the coast of Texas COP 17, Durban, South Africa	2010
COP 18, Doha, Qatar	2011
COP 19, Warsaw, Poland	2012
Fifth Assessment Report IPCC	2013
COP 20, Lima, Peru COP 21, Paris, France	2014
COP 22, Marrakech, Morocco	2015
COP 23, Bonn, Germany	2016
COP 24, Katowice, Poland	2017
COP 25, Madrid, Spain	2018
Coronavirus disease 2019 (SARS-CoV-2) Pandemic	2019
COP 26, Glasgow, United Kingdom	2020
Coronavirus disease 2019 (SARS-CoV-2) Pandemic Sixth Assessment Report IPCC	2021
COP 27, Sharm El Sheikh, Egypt Coronavirus disease 2019 (SARS-CoV-2) Pandemic ends	2022
IPCC Final AR6 Synthesis report published COP 28, Dubai, United Arab Emirates	2023
COP 29, Baku, Azerbaijan	2024

summarizes some key events since 1990 in the climate change mitigation and adaptation policy space, along with markers for international events, as context to the development of IOC sustainability reporting, and our analysis of those reports.

We now explore how IOCs respond to the challenges of climate change through the public media published in their sustainability reports. We begin with a qualitative review of the rhetorical and narrative frames common to these reports and conclude with a linguistic analysis which quantifies this perspective of how IOC actors manipulate their performance on the public record of sustainability.

3.1. IOC Rhetoric and Preferred Narrative Framing

To be clear, the IOCs sampled here neither explicitly doubt the veracity of climate change nor challenge the science. These functions are out-sourced to allegedly ‘independent’ think tanks, such as the Heartland Institute among others, which are almost entirely funded by IOCs and shareholder subsidiaries [28,29]. The approach adopted by the IOC is more subtle, as evidenced in the following comment, which is fairly representative of how the IOC discuss climate change: “While future temperature changes and the associated impacts are difficult to accurately predict, we believe the risks of climate change are real and warrant thoughtful action” [52] (p. 30. Added emphases).

Although strategies that are aimed to foment doubt in the science (and scientists) and that invest billions in lobbying campaigns are the functions of well resourced allegedly independent think-tanks [28,32,33], the emphasis in the IO corporate sustainability publications rests on a strategy that defers compliance with the accuracy of the science to some future date and to a time when there is 100% certainty. Notable by its relative absence are references to any of the scenarios modelled by the IPCC, including the BAU (business-as-usual) model which is the most devastating but which also appears to be the default position of the IOC sector. While the IOC positions itself as favouring a well-defined rhetoric, acknowledging that the risks are real, this narrative also counters the ‘predictive’ accuracy of impacts and consequences of the IPCC models. That is, despite Shell claiming that waiting for iron-clad proof before acting was irresponsible, this seems to be precisely the default position IOCs have adopted.

From our review of the sampled public-facing sustainability reports, we observe a number of recurring patterns. IOC sustainability narratives tend to take the form of statements and claims about how the particular corporation is becoming increasingly ‘sustainable’. These claims describe how the IOC is lowering (or is planning on lowering) its emissions which reduce associated social-ecological impacts. These changes are described as ways that the corporation is ‘pioneering’ the transformation of energy, generating energy that will be less polluting, less damaging, with reduced carbon emissions (i.e., ‘cleaner’). Another claim is to be developing technology to capture the CO₂ at source, thereby eliminating it from the atmosphere. This statement by Shell [53] (p. 1) typifies such claims that IOCs are “investing now for the future: in being leaders in developing low CO₂ second-generation biofuels; in building our capacity in carbon capture and storage (CCS) technology, and in working to drive down the costs of renewable power.” Six years later, Shell continued with such future-oriented developments. In 2013, Shell identified four strategies for responding to climate change, namely an increased use of natural gas, the development of carbon capture and storage (CCS), investments in biofuels, and enhanced energy efficiency [54].

It is quite common for IOCs to refer to technological solutions to the problems of climate change. For example, in 2013, the prospect of CCS technology loomed large as a way of reducing the emissions from the Scotford Upgrader which processes bitumen recovered from oil sands projects in Alberta, Canada. Shell’s flagship CCS project, Quest, although not yet on-stream, was anticipated in 2013 to “potentially capture more than 1 million tonnes of CO₂ a year and store it 2 km below the surface, safely and permanently, when it starts operations from around 2015” [54] (p. 13). In its 2015 report, Shell provided an update on the operations of Quest which had come on-stream in August 2015 and by the date of the 2015 report had captured some 315 kilotonnes of CO₂. By 2016, Shell’s CEO claimed that Quest had captured “more than 1 million tonnes of carbon dioxide (CO₂) from our oil sands operations” [55] (p. 4).

However, an independent study of Quest by the non-profit Global Witness reported that, for the five-year period during which Shell claims to have captured and stored some 5 million tonnes of CO₂, the plant actually emitted approximately 7.5 million tonnes of CO₂ over the same period [56]. Emitting 7.5 million tonnes to (allegedly) capture 5 million tonnes is not progress. Moreover, contrary to Shell’s claims that the plant captures approximately 90% of emissions, in practice, Quest captured only 48%, a figure that is actually further reduced to 39% of total emissions captured when the methane pollution from the fossil gas supply chain are included in the calculations [56]. While the figures presented in the sustainability reports are not necessarily wrong, they are misleading because they exclude data from the entire industrial cycle when including these would be unfavourable to the corporate image. Here, the narrative strategy employed by Shell concerns repositioning the boundaries of the analytic window, a tactic which shows them in a more favourable light.

Even some of their own solutions pose further problems with respect to the predicted increases in GHG emissions. For example, extracting oil and gas from tar sands was described by Shell as providing “a less costly fuel and raw material for the chemicals industry” [57] (p. 15), with tar sands being described as “one of the world’s most significant energy resources and provide an important source of energy for North America” [54] (p. 26). Shell’s strategy concerns “maintaining the energy efficiency levels of recent years [even though doing so] will be more difficult in the future as existing fields age and new production comes from more energy-intensive sources. This is expected to increase our GHG emissions over time” [54] (p. 32. Added emphasis). Fuel and raw material costs may be reduced through new ventures such as hydraulic fracking, but greenhouse gas emissions will continue to increase. The basic rhetorical pattern suggests an always future focus on work and developments that have yet to happen, technologies that are in progress but which have yet to bear fruit at the scale predicted even when these developments do come on stream. In brief, it is a ‘jam tomorrow’ tactic, even while actual investment in renewables at scale (which would reduce emissions) continues to remain negligible in comparison to other capital and R&D investments.

The onus of responsibility for making changes is an interesting narrative thread throughout the sampled reports. IOCs appear to approach this thorny issue of responsibility by first acknowledging the problem (although do not attribute causality to any agency in the fossil fuel sector, nor indeed to the sector itself) and supporting the need to shift to a low-carbon economy. The second step however is the sleight of hand: first the tactic ensures that the IOC is positively framed as part of the solution, and then the switch shifts responsibility for actual change to everyone, glossing over matters of infrastructure, economic lock-in, and other factors that the fossil fuel sector has invested heavily in to ensure that they benefit from it. The comments by Pratima Rangarajan (Chief executive officer, OGCI Climate Investments) in her foreword to the 2017 BP report are typical of this pattern. She claims that “[n]o one company or sector alone can deliver a low carbon future. Everyone, from consumers to corporations to governments, needs to take responsibility” [58] (p. 2. Added emphases). IOCs, as such, are not mentioned: despite being the architects of climate change and the drivers of the paralysis that has gripped successive governments since the mid-1980s, their culpability is whitewashed out of the causal framing, which is now repositioned to refer to ‘everyone’, specifically consumers (who have limited power to influence global and national infrastructure), corporations (which often benefit directly or indirectly, such as through pension fund and shareholder investments, with portfolios that may include IOCs), and governments (the target of IOCs lobbying companies with large client budgets).

But this is by now a well-established pattern of blame shifting and goes back to BP’s first report. Here, in 2004, the game-plan was set out through the admonition that “Unless people understand issues such as climate change, there will not be popular support for the measures needed to deal with them and neither will individuals be motivated to play their part” [59] (p. 46. Added emphases). In this framing of the matter, IOCs are clearly exonerated as causal agents of climate change; instead, the problem is reframed as individuals failing to properly understand climate change. In brief, climate change persists due to public ignorance and antipathy to ‘play their part’. These are but two examples of a wider repertoire of gas-lighting and blame-shifting tactics employed by IOCs in their sustainability reports.

The strategy appears to take the following dynamic. Through carefully worded statements, IOCs position themselves as willing but hapless actors, unable to do anything until such time as there is a change in market forces, or until individuals are sufficiently motivated to play their part in bringing about change. Such framing does not acknowledge the comparative powerlessness of individuals to pressure IOCs to change at the scale necessary to make a difference and within the limited window of opportunity counted down by successive IPCC Assessment Reports, the annual COPs, and the available evidence that climate heating is already underway. These narratives also exclude mention of the enormous sums of money spent on lobbying legislators to dilute targets and sanctions, the publicity campaigns portraying IOCs as leading the change, and the prominent investment in community activities showing IOCs in a positive role.

While there are a number of similarities in the strategies adopted by the three IOCs studied here, there are also some differences. Primarily, these differences appear to be in the claimed contributions to climate science made, especially by ExxonMobil which maintains that its scientists are experts in this field and are contributing meaningfully to developing an understanding of climate change. BP and Shell tend not to emphasize such claims to the same degree. BP, as illustrated above, adopts a different tack, suggesting that it is a lack of public understanding about climate change, together with a diffused responsibility, that stifles efforts to address the causes of climate change. Such subtle and nuanced differences aside, however, the similarities among IOCs outweigh these through their common emphasis on denying direct culpability for being the drivers of climate change and the recruitment of diversionary tactics which position the IOCs in a favourable light, depicting themselves as leaders in decarbonising the global energy infrastructure.

Similar findings have been reported in earlier studies into corporate ‘greenspeak’ [60] and almost certainly have been found in earlier studies into how corporations, and IOCs in particular, navigate tensions and manipulate their use of language to portray themselves and their activities in a more favourable light [61,62]. In the following pages, we endeavour to extend these studies by exploring the linguistic construction of how IOCs present themselves as sustainable and responsible corporate actors in the face of climate change challenges. We begin by briefly introducing a few patterns of key term relative use frequencies from employing text mining methods, and we then study collocations and concordance lines which enable a deeper examination into how key words are used in the construction of preferred narratives.

3.2. Text Mining and Term Use Frequencies

Relative frequencies of key word usage are an indicator of conceptual salience and have been associated with lower cognitive demands due to pattern recognition on the part of the reader. Terms come into and fall out of favour, either as a reflection of changes in thinking, prioritization, or simply because the basic ideas are replaced by substitute terms. The first point of analysis compares the relative use frequencies of three terms that, on the face of it, reflect the core concerns of IOCs: ‘oil’, ‘gas’, and ‘energi*’ (the lemmatized form of energy and associates). To weigh the use of terms within the context of the number of words for each corpus, the raw frequency was divided by the total number of words in the corpus for each successive year, returning a relative use frequency. These are summarized in Figure 1.

The pattern in Figure 1 suggests that the term ‘energi*’ became more popular earlier on, particularly in 2005 when both Shell and BP published their sustainability reports. Both ‘gas’ and ‘oil’ were used at about the same relative frequency, rising even while ‘energi*’ seems to drop some in usage around 2006 and 2007. The relative usage trajectories tend to show ‘oil’ and ‘gas’ tracking closely together from 2009 to 2014, and all three were used equally frequently in 2013. Thereafter, however, the frequency of references to ‘oil’ noticeably drop off in 2016, even though ‘gas’ and the more neutral term ‘energi*’ continue to track each other, until 2018. At this point, ‘gas’ also becomes less frequently referred to in the reports, until a sudden spike in 2023. The descent across all three terms in 2020 and 2021 may be due to the global pandemic. In 2023, it is interesting that the relative frequency of references to ‘oil’ and ‘gas’ increase at the same time that the more neutral term of ‘energi*’ decreases. It is difficult to account for this shift, and as it is a pattern that can only be determined retrospectively, it is likely something that future research may be able to help interpret contextually.

One hypothesis that may help understand the apparent shifts towards the end of the sampled period in Figure 1 is the use of term substitution. In Figure 2, we map the relative use frequency of the more neutral term ‘product’ and overlay that onto the patterns in Figure 1.

The relative use frequencies of this neutral term begin by closely following that of both ‘gas’ and ‘oil’, but as the use of ‘oil’ in the reports begins to decline after 2016, the use of ‘product’ remains constant until it picks up towards the end of the sample period. To some extent, this finding is supported by the 2022 InfluenceMap study that drew on “3421 items of public communications materials from the five [Big Oil] companies in 2021”, concluding that “60% contained at least one green claim, while only 23% contained claims promoting oil and gas” [63] (np). Consequently, Figure 2 likely reflects a shift in terminology only rather than indicating any genuine decline in the extraction, processing, and distribution of fossil fuels. Perhaps, with the recent shift in international politics and the reissuing of licences to drill, IOCs feel more empowered to refer to specific oil and gas products explicitly.

While it may be tempting to attribute responsibility to the influence of the Paris Agreement for this apparent change in focus, this may not be the case. The 2022 InfluenceMap study concluded that “none of the companies have aligned their climate policy engagement activities with the goals of the Paris Agreement” [63] (np), pivoting instead towards the alternative strategy of claiming their “support of, or involvement with, efforts to transition the energy mix”, a tactic that was “by far the most popular type of green claim.” We explore this observation further by selecting two adjectives (low* and clean*—note the use of the wildcard ‘*’ which extends the scope of how these adjectives are used) and the first 20 nouns most strongly associated with each adjective as per the MI3 measure of association. These findings are summarized in

Table 3. Collocates with the adjective ‘low*’ across whole corpus using the MI3 measure of association (L8-R8).

Collocate	Freq. (Collocation)	Freq. (Subcorpus)	MI3
carbon	1589	4279	29.4
energy	739	10,111	24.8
emissions	380	6643	22.5
transition	201	1415	22
gas	334	7677	21.8
fuels	169	1024	21.7
future	195	1783	21.5
products	161	1946	20.6
biofuels	131	1132	20.5
technologies	122	1015	20.3
power	143	1740	20.2
solutions	115	985	20.1
emission	73	295	19.9
cost	88	549	19.8
electricity	101	854	19.8
fuel	111	1233	19.6
transport	93	946	19.3
sulphur	56	223	19.1
alternatives	43	101	19.1
natural	121	2278	19.1

and

Table 4. Collocates with the adjective ‘clean*’ across whole corpus using the MI3 measure of association (L8-R8).

Collocate	Freq. (Collocation)	Freq. (Subcorpus)	MI3
burning	108	157	23.7
energy	392	10,111	23.3
cooking	47	92	20.9
fuels	100	1024	20.7
natural	128	2278	20.6
gas	188	7677	20.5
power	99	1740	19.9
fuel	78	1233	19.3
transport	70	946	19.2
solutions	65	985	18.9
affordable	41	248	18.9
air	49	686	18.2
coal	36	408	17.6
solar	45	812	17.6
technologies	38	1015	16.5
sustainable	46	1771	16.5
fossil	21	280	15.8
transition	34	1415	15.5
generation	22	495	15.2
efficient	20	366	15.2

.

Table 3 may be interpreted as follows. The adjective ‘low*’ (including the wildcard) collocates with ‘carbon’ 1589 times within the window span between eight words to the right and to the left of the key term ‘low*’, while the term ‘carbon’ appears 4279 times across the whole corpus, leading to a MI3 value of 29.4. This is very high, given that the typical threshold of significance for MI3 is usually 9.0. As can be seen from Table 3, the adjective ‘low*’ associates strongly with ‘carbon’, ‘energy’, ‘emissions’, ‘transition’, ‘gas’ and ‘fuels’, to take the first six, and this means that the key term ‘low*’ modifies these nouns in a pattern such as ‘low carbon’, ‘low energy’, ‘low emissions’, etc.

In Table 4, a similar pattern to that in Table 3 is found, with ‘clean*’ being most strongly associated with ‘burning’ and ‘energy’, and so on. This lends some provisional support to the claim in the InfluenceMap report that the changes adopted by IOCs are primarily cosmetic, rather than substantial, since claims of ‘clean burning’ and ‘clean gas’ are constructions that gloss the fact, for example, that gas and any form of combustion is still not ‘clean’ in the sense of producing no emissions. Nevertheless, adopting a strategy of claiming to support an energy transition towards lower-carbon or cleaner burning fuels is evidently one of the preferred approaches by Big Oil. Arguably, IOCs have invested up to “hundreds of millions of dollars each year on a systematic strategy to portray themselves as positive and proactive on the climate change emergency” [63] (np), a well-established practice spanning a number of years [30].

In 2005, BP acknowledged that “[a]s a global energy company, providing about 2% of the world’s total primary energy, we have a responsibility to help the world meet its increasing demand for energy in a sustainable way, taking precautionary action to address the threat of climate change” [64] (p. 39). Figure 3 explores the extent to which the three sampled IOCs take climate change, sustainability and greenhouse gas emissions into account. It is notable in Figure 3 that the sampled IOCs clearly recognize matters of sustainability and volumes of emissions, even though specific references to climate (change) and greenhouse gas are comparatively low.

One interpretation of Figure 3 is that there is an apparent disconnect in the reports between greenhouse gases and emissions. Emissions may be more neutral in tone and in terms of causal attribution, whereas terms like greenhouse gases and climate change are more likely to elicit an association between the two processes in the mind of the reader. What is of interest however is how the use of ‘sustainabil*’ is used with increasing frequency, particularly post-2015, with a spike in usage from 2018 onward. To explore this spike, we ran a further collocation analysis using the MI3 measure of association, and these results are summarized in

Table 5. Collocates with the adjective ‘sustaina*’ across whole corpus using the MI3 measure of association (L8-R8).

Collocate	Freq. (Collocation)	Freq. (Subcorpus)	MI3
report	3803	6481	30.8
shell	3015	9938	29.2
bp	2244	8338	28.2
values	613	791	25.9
achieving	537	645	25.7
reporting	791	2487	25.4
development	786	3713	24.8
core	442	643	24.8
performance	818	4586	24.7
energy	958	10,111	24.2
review	481	1446	24
embedding	284	295	24
net	487	2320	23.4
zero	450	1860	23.4
business	579	5260	23
approach	446	2305	23
responsible	277	996	22.2
emissions	518	6643	22.1
external	270	957	22.1
progress	308	1719	21.8

.

As with Table 3 and Table 4, Table 5 shows the collocates to the key term, here ‘sustaina*’ (note the wildcard), in decreasing measures of associative strength (MI3). The strongest associations are with nouns such as ‘report’ and the name of two IOCs, Shell and BP, respectively. This is because IOCs often refer to the report directly as in ‘sustainability report’ and ‘sustainability reporting’. Table 5 becomes more interesting from the fourth row downwards. Here, words like ‘values’ and ‘achieving’ become more prominent, and the usage suggests that the IOCs are reporting on what they identify as their core values. When the use of ‘values’ is investigated in terms of collocation, the over-riding association is with the notion of ‘respecting’ (MI3 = 28.3), which in turn associates with ‘nature’ (MI3 = 26.9) and ‘powering’ (MI3 = 26.3). Of interest, however, is that when the concordance lines for these terms are explored, we find that the most typical and representative phrasing is a variation on this exemplar: “values; achieving net-zero emissions; respecting nature; powering lives; generating shareholder” value (e.g., p. 7, [65]). That is, values, respecting and sustainability tend to assume the form of slogans and strap-lines more akin to vision statements or advertising copy, rather than actual specifics. However, the emphasis on sustainability (and GHGs), while underplaying climate impacts, may suggest the prioritization afforded by IOCs to technological solutions, to generate ‘cleaner’ and lower’ future energy supplies that have fewer emissions. This conclusion is consistent with InfluenceMap’s findings, as reported above, described as a ‘jam tomorrow’ rhetorical device.

From a public relations perspective, it is not in the best interests of IOCs to link their practices with climate change, even though in practice, introducing such distance between their product and GHG emissions, and by extension, climate change, is challenging, and necessitates some linguistic and cognitive maneuvering. The strategy for achieving this was considered in Section 3.1., above, whereby the preferred narrative appears to be a strategy of agreeing that while there is a problem, rather than admit culpability, the IOC details its own research and development funding into alternatives, into tackling the problem, thereby portraying itself as a good actor, even a front runner in responding to the issue of climate change. Two citations from ExxonMobil illustrate this strategy in use.

“For more than three decades, we have continuously funded and participated in research to improve understanding of climate science, often in conjunction with government bodies and leading research universities. This has resulted in hundreds of publicly available documents on climate-related topics, including more than 50 peer-reviewed publications. We promote discussion on issues of direct relevance to the company and contribute to a wide range of academic and policy organizations that research and promote dialogue on significant domestic and foreign policy issues.” [52] (p. 12. Added emphases)

In this statement, ExxonMobil positions itself as working tirelessly to promote robust scientifically evidenced technology to reduce emissions and to promote sustainability. This is a tactic common to IOCs to position themselves as contributing to understanding climate science. But IOCs go further than offering an understanding of climate change, as the report continues:

“We are proud of the progress we’ve made, and we recognize the importance of continuing our research and development to help further expand the understanding of climate science. Our efforts have enabled us to take meaningful action to mitigate the risks of climate change, and we will continue to build on this foundation as we work with governments and stakeholders to further address the issue” [52] (p. 13. Added emphases).

As with the other IOC reports sampled here, this narrative locates ExxonMobil as committed to trying to understand, and then to address and respond to climate change. The corporate actor emphasizes its expertise, and reports on how it funds research and technological developments to mitigate global warming, thereby presenting itself as a good corporate actor. On the face of it, ExxonMobil is vigorously pursuing of climate change research science and meaningful engagement policy making to mitigate the impacts. What ExxonMobil does not acknowledge, of course, are the millions of US dollars it spent on funding climate change denial groups [66,67], nor the systematic disinformation it and other IOCs engage in that attempt to question and discredit the well-evidenced climate science that has been accumulated over decades. This is because climate science all attribute causality unerringly to the contributions IOCs have made to the climate crisis. It is unlikely simple coincidence that only one single reference to ppm (parts per million) of GHG emissions can be found in all of the sustainability reports since 2011. Instead, one must conclude that the IOC strategy has been to simply deflect attention away from referencing this robust and standardized accounting metric of volumetric carbon and other GHG emissions in their reports, because to do so would be to acknowledge and assume responsibility for their on-going and systematic failure to meaningfully reduce emissions, regardless of their numerous commitments to do so. Such duplicitous strategies enable IOCs to play two sides against each other.

To summarize, the strategy appears to be as follows: On one side, IOCs invest significant fortunes into discrediting the science, attempt to reclassify GHG emissions as pollutants rather than as drivers of global warming, and slow legislative changes that would affect their business model. These tactics are part of a larger repertoire, as noted by others (e.g., [29,30]). On the other hand, they portray themselves as leaders in technological solutions to reduce future emissions, which helps to identify and position themselves as concerned corporate citizens, as critical parts of the solutions (to the problems they themselves have caused), and so on.

In this section, we identified a sample of words with significantly high relative use frequencies and examined the vicissitudes of usage patterns. We proposed that IOCs have tended to downplay the use of certain terms that may position them in a less than favourable light and examined the ways that some of these high-frequency terms collocate with other terms. We have also explored some of the more nuanced meanings of words that, on the face of things, even suggest values that could be aligned with a transition to a net zero and sustainable future, only to find that, even here, these terms tend to refer primarily to sloganeering that lack detail and suggest a focus on corporate interests rather than any genuine engagement with or commitment to a transition away from producing the causes of climate change.

In the following section, we take a closer look at the use of language, particularly the associations among key terms (collocation analysis) and how these are put to use in the construction of sentences (concordance analysis). Here we explore how language is used in the promotion of discourses that support and maintain particular power dynamics in the favour of IOC business model, one that is predicated on enervating the voices of both the public and of governmental legislative and policy makers to confront and work through the challenges of shifting to a world no longer dependent on fossil fuels.

3.3. Analyzing the Linguistic Patterns of the Sustainability Report Corpus

The MI3 measure of associative strength was deployed to investigate the corpus to determine the words associating most significantly with key search terms. We begin by examining the collexemes for the primary IOC product line—oil and gas. There are almost 6200 unique occurrences of the word ‘oil’ across the twenty years of reports, and the most statistically significant collexeme for ‘oil’ is ‘gas’ (MI3 = 30.4), followed closely by ‘sands’ (MI3 = 28). This refers to ‘oil sands’, particularly those found in Alberta, Canada, and refers to hydraulic fracking from oil sands. However, this is well known to be a highly damaging process to the environment, as acknowledged almost 20 years ago by the then Group Chief Executive of BP, Tony Hayward who, in 2008 wrote “We recognize that oil sands projects raise significant environmental challenges, but we are actively seeking ways to undertake ours in a way that minimizes the environmental footprint” [68] (p. 2). Even Shell admits, “Fuels produced from oil sands typically emit 4 to 18% more CO₂ than those from the average crude oil consumed in the USA, from production through to use as a transport fuel” [69] (p. 26).

While BP does not detail the specific challenges (read ‘impacts’) associated with fracking, in their report five years later, BP claims that fracking is a necessary part of their transition to lower-carbon fuels. That is, despite such impacts, including an increase in greenhouse gas emissions, BP still “believe[s] a diverse mix of fuels and technologies can enhance national and global energy security while supporting the transition to a lower-carbon economy. These are reasons why BP’s portfolio includes oil sands, shale gas, deepwater oil and gas, and biofuels” [70] (p. 13). However, these remain difficult claims to reconcile—on the one hand, oil sands are promoted as playing a part in lower-carbon transitions, but on the other, oil sand-based fuels emit significantly more GHG emissions than crude oil-based fuels do. The notion of ‘oil sands’ reflects what IOCs refer to as ‘unconventional’ fuels, and while renewable sources of energy (especially wind and solar) remain under-represented throughout this corpus, the emphasis remains firmly on extractive resources which are nominated as necessary for the IOCs’ transition portfolio.

Part of the rationale behind this disjunction might be found in how IOCs account for their GHG emissions. In their 2017 report, Shell provides insight into these accounting methods. Again, in relation to oil sands, using concordance analysis we read that “Our nitrogen oxides emissions decreased from 122 thousand tonnes in 2016 to 107 thousand tonnes in 2017. The decrease was mainly due to the change in oil sands mining reporting boundary and changes in calculation methodologies at some of our facilities” [69] (p. 56. Added emphases). The reported decrease is therefore due to changes in accounting methods, rather than to any actual improvements in technology. Because these changes have been adopted by some (but not standardized across all) of Shell’s facilities, it therefore becomes difficult to meaningfully compare facilities due to different accounting methods being employed.

This is blatant obfuscation of accountability and performance auditing, and reflects a strategy previously observed in Shell’s reports on their carbon capture and storage (CCS) facility, Quest. In a clear move that illustrates a shift in the analytic boundary used by Shell to report on the volume of CO₂ it claimed the CCS was capturing, it actually turns out that it was emitting more GHGs than it was capturing [56], as discussed earlier in Section 3.1. By changing the method of analysis (or accounting), different (and more favourable) outcomes can be presented. The rhetorical strategy of reputational management requires that only favourable outcomes are reported, while those that are less favourable are underplayed or dismissed outright.

The second product of IOCs is ‘gas’, with over 2380 collexemes above the MI3 = 9.0 threshold, the most frequent being ‘natural’ (MI3 = 30.4), followed by ‘greenhouse’ (MI3 = 28.9), typically in the form of the bigram ‘greenhouse gas(es)’, which is closely associated with ‘emissions’ (MI3 = 27.5). Mostly, references to ‘gas’ associate with liquefied natural gas (LNG), reputed to contain the least amount of CO₂ and which is offered by IOCs as a low greenhouse gas emitting fossil fuel. It remains controversial, however, because it is far from a clean form of energy. A 2004 study by Greenpeace, for example, found that the natural gas combustion required to produce and transport LNG to the plants adds 20 to 40 percent more carbon dioxide than burning natural gas alone [71]. This is before one factors in damages caused by the processing plants to the environment and to local biodiversity, as found in Russia’s Sakhalin Island, the Peruvian Camisea fields, and elsewhere. Despite claims about LNG’s low carbon and low impact made by IOCs, available evidence does not support these. As with the accounting strategy used in reporting on the removal of CO₂ by the Quest CCS facility noted earlier, the tactic here seems to be little more than repositioning the analytical frame to emphasize the most favourable data while obscuring and excluding a wider and more representative account which does not reflect IOCs in a positive light.

Unquestionably, humanity must change its current trajectory and the current socio-economic reliance on a fossil fuel-based infrastructure. Consequently, words referring to change, such as ‘transition’, become increasingly important to evaluate how such changes are construed and implemented. In the IOC corpus, ‘transition’ collocates significantly with ‘energy’ (MI3 = 27), and closer study reveals that these associations are predominantly aspirational statements about ways that IOCs can be a part of a transition to a lower-carbon energy future. Note that the emphasis remains stubbornly fixated on ‘lower’-carbon futures, not a zero-carbon future. Describing their contribution to this transition, some IOCs point to the added value of carbon capture, use, and storage (CCUS) facilities. However, by 2017 BP was still admitting that this was an under-developed and under-deployed technology, yet to become a reality. This is evidenced through concordance analysis, an example of which is this statement by BP: “Collaboration is required to make CCUS a reality. The technology has been in use for more than 20 years, but needs governmental support—through a carbon price and other policy measures—to accelerate its deployment” [58] (p. 30). Given the profits accrued by IOCs, it is disingenuous for corporations in this sector to blame the lack of implementation on an absence of governmental policy carbon-pricing mechanisms to deploy what is still a largely controversial and unproven technology. This seems to suggest a deflection in responsibility and leadership from IOCs to governments.

In their 2019 report, BP announced their transition strategy referred to as the ‘reduce, improve, create’ (RIC) framework, which involves “reducing greenhouse gas emissions in our own operations, improving our products to help our customers lower their emissions, and creating low carbon businesses that support the energy transition.” [72] (p. 20). In subsequent elaborations on this strategy, BP aligns itself with the Paris Agreement and uses the Paris Agreement Article 4.1. as their operating definition for achieving net zero, stated as achieving “a balance between sources of anthropogenic emissions and removal by sinks of greenhouse gases” [73] (p. 12). However, when one explores the use of the term ‘sink’, aside from the functional process of sinks being a form for removal of emission-based carbon, ‘forests’, ‘peatlands’, and ‘wetlands’ are the most frequently nominated sinks. The ecological impacts of oil and gas extraction on these very ecosystems have, however, been excluded from the potential capacity of these sinks to optimize carbon capture. Once again, these claims evidence a degree of cognitive dissonance, even compartmentalization, in how the IOC sustainability ambitions and reports are expressed. Of greater concern however is the apparent lack of awareness that, increasingly, former carbon sinks are destabilizing and becoming carbon sources [74,75].

Even if this were not in itself a cause of concern, while forests may continue to offer some potential as carbon sinks, carbon offsetting programs require careful auditing to ensure that the trees are maintained and later managed properly to avoid releasing or to no longer be sequestering the volumes of carbon as calculated. European forest biomass may already be reaching saturation point, becoming less viable as sinks, unless forest management policies are revised accordingly. Moreover, peatlands also pose a potential risk of becoming carbon emitters rather than sinks [76], and caution is advised in how climate change impacts, such as wild fires on the tundra, convert these sinks into methane emitters. As a result, an optimistic future reliance on these sinks cannot be assured, yet no contingency plans for this are discussed in the sustainability reports, even though risk assessments are de facto tools deployed by the IOC sector.

We conclude this sub-section with a final collocation analysis of the term ‘reput*’ (a lemma for ‘repute’ and ‘reputation’). Occurring throughout the corpus 74 times, this attracts comparatively few collexemes (n = 394), and ‘tracker’ is the collocate with the most significant score (MI3 = 22.3). The ‘tracker’ is a mechanism that draws on and combines a range of data sources to calculate the brand reputation of major corporations. In 2020, Shell’s reputation value was £92.3bn, and BP’s was £51.1bn, so these are clearly important metrics for attracting shareholder investment. Nevertheless, BP announced an end to their reputation-based advertising campaigns in favour of redirecting funding towards promoting themselves as a ‘green’ fossil fuel company: “We have stopped corporate reputation advertising campaigns and this is enabling us to redirect resources to promote well designed climate policies. In future, any corporate advertising will be to push for progressive climate policy; communicate our net zero ambition; invite ideas; or build collaborations” [73] (p. 25). ExxonMobil [52] (p. 3), for example, has adopted a similar tactic to promote itself as a front-runner in addressing climate change, claiming, despite sparse supporting evidence, to:

“advance research and development, leading to innovation in areas such as advanced biofuels, carbon capture and sequestration, and energy efficiency. Managing our own greenhouse gas emissions is a core element of our operations. We have and will continue to engage relevant stakeholders to further develop climate science and broaden its understanding by society at large”.

Reputational gains made by the IOC are often based on claims made, rather than by independent verification. As a consequence, there appears little to differentiate any of the three IOCs sampled here from one another. With respect to reputational management, the intent appears to try to manage the climate change policy conversations through several tactics. These include how they position themselves as researchers and policy advisors, citing the research undertaken by IOCs, which is seldom independently peer reviewed, and is rarely self-critical of its own culpability in the continued escalation of climate change as an existential threat multiplier. As noted earlier, while ExxonMobil positions itself as a leader in the scientific understanding of climate change, and BP and Shell engage in a diffusion of responsibility for change, none of the IOCs sampled here assume direct culpability for the widespread impacts of their product.

4. Conclusions

In the previous sections, we examined some of the reputational management strategies adopted by the three sampled IOCs and explored how these are manifested and propagated through the text of their sustainability reports. This process brought to light some recurring themes, which include: how IOCs position themselves as on the front foot of climate change research and as advocates for ‘thoughtful’ engagement in policy making all the while discounting, ignoring, or discrediting non-partisan research spanning decades, commonly through arm’s length lobbying firms. We also noted how the sampled corporate reports emphasize their commitment to a low or, more recently, a net-zero transition. The measures selected to demonstrate this commitment involve promoting fossil fuels described as ‘clean(er)’ and ‘low(er)’-carbon, characterized by an uptick in the use of such terminology to described the IOC product line, including the preference for more neutral terminology, such as ‘energy’ and ‘product’. We also noted how IOCs portray themselves as good actors investing in community development schemes, research and education and as willing participants and collaborators in climate change conversations. Finally, we observed how emission accounting methods are subject to redrawing boundaries and even methods used in calculating emissions, including the use of externalities and investments in carbon offsetting schemes, with significant reliance on carbon sinks, some of which may already be at risk of saturation.

For all of these observations, however, what we did not find was any significant investment in zero-carbon renewable energy sources. While it is true that some IOCs are pursuing financial and R&D investments in wind and solar, investigative reports continue to evidence that investment in zero-carbon technologies pales in comparison to the ongoing investment in carbon-intensive technologies [23]. Moreover, if IOCs were really acting faithfully on their various commitments, it is difficult to reconcile how emissions are three times higher than the IOC sustainability reports would have us believe (e.g., [77]). These findings lead us to correct the proposition that “[c]ompanies from industries associated with negative externalities will engage with more SDG targets than companies from industries with more positive externalities” [40] (p. 216). What we found instead is that the SDG targets that IOCs do engage with are adjacent to the core causal chain driving climate change. This apparent contradiction remains a circle that cannot be squared, unless one concludes that, despite their laudable rhetoric and reputational management narratives, IOCs are, by and large, continuing to press on with their tried and tested business-as-usual strategy. Certainly this is the conclusion reached by Supran and Oreskes [78] in their rhetoric and frame analysis of ExxonMobil’s communication strategy. We see no evidence to dispute this conclusion from the documents we examined.

Instead, our findings provide further support to these previous and highly critical studies, and our conclusions support the determinations made during court cases [36] and the findings of the US Oversight Committee [28]. Together, these studies and critical examinations of the IOC business model reach the inescapable and, by now, well-documented conclusion that IOCs are far from being the good corporate citizens they claim to be. On the contrary, rather than being on the forefront of a just transition to a sustainable future, IOCs invest ever greater efforts in disinformation, deflection, distraction, and gaslighting to draw attention away from the ecocidal consequences of their practices. By examining the use of language in the production and maintenance of false narratives that position IOCs in a favourable light, our intent in the current paper was to bring some of these rhetorical strategies to the fore. Unfortunately, given the access to deep financial resources, the degree of influence IOCs exert over national governments and governing bodies and participant nations at COP meetings, this study can but offer a modest contribution towards supporting and advancing a more informed and critical lens through which to interpret the claims made by IOCs. In light of the accumulated evidence, it is difficult to identify any positive impacts that can be associated with the rhetorical strategies of IOCs. Granted, the economic prosperity of the world has advanced as a result of the discovery and refining of oil and related products. However, this progress has come at a substantial cost to all people, especially those who did not benefit from the economic benefits in the first place. One might consider the development of lower-carbon sources of energy as progress, but considering recent actions of divesting in renewables in favour of untested CCS technologies and their drive to lobby governments to grant extraction licences, these possible benefits turn out to be tokenistic gestures to mollify criticism while they continue with their business-as-usual trajectory of profit maximization.

Moreover, the benefits gained from the fossil fuel infrastructure rapidly pale when the true costs associated with this benefit begin to be accounted for. These costs, already steep, may have been mitigated had IOCs admitted their culpability as causal agents of climate change; instead, IOCs doubled down on their refusal to assume responsibility, deflected attributions of causality, diffuse responsibility for blame, engage in deceptive communications, fund think tanks to spread misinformation, and even continue to lobby nation states during COPs, where the very topic of conversation is the reduction in greenhouse gases. These are not accidental events. Rather, they show a strategic and committed course of action to obfuscate and delay meaningful changes to prolong the maximization of profits while doing the barest minimum to address the causes of the current existential crisis. IOCs have known the dire consequences of their actions for the last seventy years, and even while there was still time to avert these predictable consequences, they refused to do so in favour of their profit model. This is both immoral and profoundly unethical.

The findings here underline that should regulatory policies countering ‘economic freedom’ be implemented, such as the Directive 2013/34/EU which calls for more legally binding transparent non-financial reporting standards, then indeed, sanction mechanisms do need to be in place too. By teasing apart some of the rhetoric around reputational management, our intent here was to identify just a few of the ways that IOCs rhetorically position themselves in the discourses of public relations and governance decision-making as prime exemplars of ‘greenwashing’, a practice of obfuscation and false narratives claiming to be green and environmentally friendly (or at least benign). These have real-world impacts, however, and because historically greater economic freedoms correlate with higher emissions, there is a clear need for balanced regulation that includes strong “environmental safeguards like carbon pricing or emissions trading to ensure economic policies support sustainability” [22]. Equipping the legality of economic and regulatory policies (e.g., the EU Directive) using clear definitions and taxonomy and backed by meaningful sanction for non-compliance with the political review of longevity in senior roles for specific governmental sectors to mitigate the effects of lobbying success [79] is a potential two-prong approach. The exploitation of the International Centre for the Settlement of Investment Disputes (ICSID) by multinational corporations, including IOCs, to bring cases against national sovereign states with claims to their mineral reserves must also be brought under review to stop the forced extraction of fossil fuel reserves from nations that want to keep the oil and coal in the ground [80]. All of this requires coordination, within and between governments and institutions. In its 2017 report on tobacco and its effects, the WHO proposed a series of measures to limit greenwashing [81]. These included international and local legislation to compel companies to meet specific disclosure requirements for their material emissions, water usage, waste disposal, chemical use, child labour and other targets. One of the key points made was that these regulations must apply equally across countries. Time must be called on these practices of deceit, as per the 2023 UN Secretary General’s denouncement of ‘greenwashers’, and the progress in decoupling emissions from economic gains must be accelerated [3].

If this paper contributes to the growing call to hold IOCs to account for their unethical practices, then it will have fulfilled its purpose. Our hope is that, along with other studies of this kind, we can inform members of the public and others who are invested in developing meaningful sustainability transition policies and offer critical tools to counter-balance and dismantle the rhetorical strategies employed by well-resourced and powerful lobbyists when the latter attempt to manage the urgently needed conversation about climate change and its drivers by obfuscating responsibility for the global crisis we face. There is much work to be performed in countering this mode of information warfare, of course, and we hope that this paper might be a modest part of that direction of travel and wider defence.