A corpus-based approach to corporate communication research

Full Text

1. Introduction

Businesses today must deal with numerous requirements designed to ensure their adequate accountability. Corporate Social Responsibility (CSR) reports have recently grown into one of the key instruments of accountability relying on the premise that the focus on profits and sales cannot outshine the companies’ crucially essential commitment to comprehend and regulate their impact on the society for the benefit of the latter. While companies may interpret “impact on the society” with a varying degree of substantive appreciation and consistent with their line of activity which may carry specific traits, broadly speaking it all comes down to acknowledging and managing any applicable impact on the environment, health, local communities, employees, and general public welfare. Simply put, the idea behind CSR reporting suggests that businesses are required to make sure and subsequently prove that the negative socially relevant impact is being reduced, while the positive impact is being scaled up.

This study sets out to analyse the language of CSR reports using a corpus-based approach to single out the most efficient language patterns catering for efficient CSR reporting narratives. By looking into the linguistic environment of the most frequently occurring tokens within the compiled corpus, the study intends to address two essential research questions:

1. What predominant language tokens and patterns are used to ensure efficient CSR reporting?
2. Which priority directions in CSR narrative composition are evidenced in language use?

To offer substantiated answers, the study suggests theoretical observations to define the scope of inquiry, as well as empirical evidence, which is explored to set forward a number of relevant conclusions. The findings of this research will hopefully suggest valuable takeaways regarding the linguistic composition of efficient CSR reports which reflect the underlying principles guiding the linguistic manifestation of modern corporate communication and culture.

2. Theoretical background

The fundamental idea behind CSR as a rather recently established corporate practice is inherently rooted in sustainable development as a globally evolved construct. The very notion of sustainable development emerged on the bedrock of three key frameworks of reference – economic, social, and environmental.

The economic frame of reference making up the conceptual substance of sustainable development relies on the theory of sustainable income essentially alluding to the idea of the maximum flow of benefits possible from a given set of assets. The theory envisages the best possible utilisation of limited resources along with the environmentally sensitive technology catering for eco-friendly extraction and processing of raw materials, delivering environmentally sound commodities, waste reduction and management, etc.

The social frame of reference is human-centric and builds on the preservation of enduring social and cultural systems that entails dialling down toxic conflicts for the benefit of the society (Petrosyan & Grishechko 2019). A major plank under this framework is fair division of goods, preservation of cultural capital and diversity, as well as consistent commitment to employ the best practices for sustainable development. Seeing that a human’s right to choose from multiple options is being increasingly recognised as a key value, the very concept of sustainable development assumes human participation in the matters that shape up their livelihood, which in turn implies their direct involvement in the processes of decision making, implementation and enforcement (Tamimy et al. 2022).

The environmental frame of reference promotes maintenance of sound and intact natural systems, where the survivability of ecosystems as a basic premise for global biosphere stability is put to the forefront.

The three frames of reference have translated into the concept of the so-called Triple Bottom Line Reporting calling for an all-inclusive and accurate disclosure of financial, social, and environmental information, which has grown to be accepted as the best practice by both public and state agencies worldwide. Finding ways to accommodate these three perspectives within the bounds of specific corporate cultures and specific corporate communication practices adopted at individual companies can pose a challenge, for the three bedrocks of sustainable development making up the core of CSR reporting need to be handled in a balanced manner (Malyuga & McCarthy 2021). Whatever mechanisms are engaged in making the three perspectives interplay with one another is of chief concern for those involved in report compilation. Thus, disclosing the nature of the reciprocity of economic and social strands may be linked to attaining intergenerational justice (as in, in relation to allocation of earnings) or offering targeted support to the poor. The mechanisms behind the co-dependent economic and environmental outcomes will certainly have to do with the fair assessment of external environmental impacts and a due account of their budgetary implications. Finally, the interplay of social and environmental perspectives may call for a thorough disclosure of a myriad of issues having to do with respecting the rights of future generations and citizens’ entitlement to weigh in on the decisions made (Popova 2018).

Fairly obviously, a company opting to disclose social responsibility information alongside their standard annual performance reports – and managing to do it right – will most probably manage to harvest additional support from stakeholders and the general public, which is understandably the ultimate goal for corporations engaged in any and all kinds of operations. Research suggests that the merits of CSR reporting are three-fold. First, this is a highly effective way to mould a holistic and consolidated image of a company. Second, disclosing this kind of data contributes to better market performance. And third, by offering non-financial reporting companies can manage the growing public demand for corporate transparency (Romanova & Smirnova 2019).

Thus, a CSR report can be defined as an instrument for informing stakeholders, workforce, affiliates, and communities about the manner and pace of delivering on the company objectives and plans related to economic resilience, social welfare and environmental stability. Interestingly, an older definition interpreted CSR as applying across “the economic, legal, ethical and discretionary expectations that society has of organisations at a given point in time” Carroll (1979: 500), thus leaving out – or at least not stressing specifically – any references to the environmental dimension of CSR, which has by now taken on a much more prominent role.

The structural-thematic composition of a CSR report, the procedures for its preparation and circulation, as well as the modalities for target audience feedback appraisal may vary depending on the specifics of company activity and the social communication strategies put in place. The overreaching goal, however, always presumes providing well-consolidated data on social initiatives and communicating these data to all stakeholders involved, including the general public, to uphold the company’s image.

In a sense, a CSR report mirrors the functional potential of a promotional tool, as it raises the company’s profile as far as the public image is concerned and, even better, has the power to drive the demand for its produce without actually “force feeding” it to the consumer, if handled correctly. Besides, regular social reporting can be useful for the purposes of holding and attracting investors who are much more likely to collaborate and stick around with a corporate partner that values social, economic, and environmental transparency.

Considering the apparent merits associated with CSR reporting, what comes to the fore is the need for an accurate and favourable data presentation, which of course requires certain language strategies to be put in place (Grishechko et al. 2015). Fostering efficient dialogue with the stakeholders and the general public to attain and maintain competitive advantage will require a circumspect approach to the wording of relevant facts, and the various issues associated with the language of CSR reports have naturally been brought up in multidisciplinary research offering a variety of perspectives on the topic.

While available research on the language and narratives of CSR reports is scarcer than that concerned with the more global spheres of marketing, business ethics or sustainability, it has been gaining traction to offer a number of valuable insights. Over the past two decades, some studies have looked into the composition of CSR reports to consider the use of narratives to communicate the firm’s financial position (Yuthas et al. 2002, Boginskaya 2022), outline the specific characteristics of CEO-authored sections of CSR reports influencing their ultimate narrative outcome (Nickerson & de Groot 2005), or single out the genres related to CSR reports as a standalone type of corporate documentation bearing its unique narrative traits (Kolk 2008, Yu & Bondi 2019). Generally, the related research tends to go by a textual analysis of CSR reports investigating, for example, metaphors used in communicating with investors to validate their needs (Livesey & Kearins 2002), or the specific language units helping bring certain arguments home to the stakeholders (Ellerup Nielsen & Thomsen 2007, Ivanova & Larina 2022).

Importantly, literature review rather apparently highlights the tendency to pay heed to the national and cultural context, bringing into the focus the discourse, language and structural composition of CSR reports delivered by Indian (Planken et al. 2010), British and Spanish (Breeze & Fernández-Vallejo 2020), Malaysian (Rajandran 2016), Chinese (Yang et al. 2015), Nigerian (Nwagbara & Belal 2019), Korean (Lee & Parpart 2018), Romanian (Danilet & Mihai 2013), Ghanian (Andrews 2019) and other corporate entities manifesting culture-specific traits of social responsibility reporting.

An industry-specific outlook on CSR discourse is also a popular subject of research with studies offering insights into the language and communication strategies adopted in CSR reports within the oil (Jaworska 2018), gas (O’Connor & Gronewold 2013), mining (Talbot & Barbat 2020), energy (Dahl & Fløttum 2019) and other sectors.

What appears to be lacking in the research agenda with respect to the narratives of CSR reporting is the distribution of priorities evidenced in language use. By looking into the most consistently used patterns within the linguistic environment of the frequently occurring tokens, this study shall further on consider the priority directions in narrative composition evidenced in language use.

3. Material and methods

This study considers the case of Microsoft’s CSR reports obtained from the company’s publicly accessible Reports Hub (Microsoft 2022). The choice of Microsoft as the source of the material is due to its consistent high ranking for CSR reporting recognised as an example of best practices in the corporate field. Being one of the world’s top companies, Microsoft has been very proactive in their commitment to pursue the most efficient social responsibility approaches and release their CSR reports following the guidelines put forward by the Global Reporting Initiative. Microsoft has also been placed at the top of the list of Corporate Responsibility Magazine’s Best Corporate Citizens (Sehgal et al. 2020). This, coupled with the company’s far-reaching stakeholder pool that motivates the executives to keep up with the ever-increasing reporting demands, makes Microsoft one of the most suitable corporate entities in terms of CSR language analysis.

Since Microsoft’s official CSR reports record covers the period from 2016 to 2022, a total of the corresponding seven documents spanning the seven-year timespan has been included as research material subject to evaluation. The customised corpus was compiled using the Prime Machine corpus concordancer tool (Jeaco 2017) and comprised 99,176 aggregate tokens. The auxiliary parts of speech were not considered as analysable tokens and were thus disregarded (Figure 1).

All remaining notional parts of speech grouped as per frequency of their occurrence were further contextually analysed using the cluster instrument that displays the content both to the left and to the right of the node in question. The length of both left and right context included in the examples in the Study and Results section is curtailed or extended on an as-needed basis to provide essentially required contextual information in any individual case that displays enough context to grasp the idea behind the narrative. The linguistic environment along with the keyword frequency of occurrence information were assessed to infer the most efficient “go-to” language patterns of Microsoft’s CSR reports that contributed to having them recognised as exemplary within the communities of corporate culture. This, in turn, proved helpful in deconstructing the hierarchy of priority topics in the analysed texts.

Figure 1. Customised corpus word list

4. Study and results

While the inferences made in this section are grounded on the frequency analysis later on supplemented with the contextual cluster analysis, the immediate results obtained upon corpus compilation called for some pre-emptive clarifications regarding possible groupings of analysable tokens that could be put in place. As shown in Figure 1, the most frequently occurring token encountered in the sample is Microsoft tallying at 8,289 total occurrences across the 99,176-token corpus. Although this would logically prompt its priority handling in the analysis as the most commonly occurring unit in the CSR reports considered in this research, proper allowance must be made for the following three tokens rated no. 2, 3 and 4 in the frequency ranking, since they constitute interlinked pronounal elements (we personal pronoun plural, us personal pronoun plural objective case, our possessive pronoun plural). Considering that for the purposes of this analysis these can be contemplated as a unified aggregate, the total occurrences of Microsoft amount to a whopping 10,801 instances placing it at the top of the list and making it an incontestable leader.

The following tables incorporate selective examples of the encountered linguistic environment for our, we and us found in the sample (Tables 1‒3).

Table 1. Examples of linguistic environment for “our” registered in the sample

Table 2. Examples of linguistic environment for “we” registered in the sample

Table 3. Examples of linguistic environment for “us” registered in the sample

Following on, Microsoft has been found to be a runner-up in terms of the frequency of occurrence throughout the sample with one rather visible pattern of use standing out as predominant – namely, Microsoft + a verb denoting positive action. While multiple other patterns are also present, they are spread thinly in the sample and are not exactly repetitive. Table 4 below has been compiled to include some illustrative examples of the most recurring pattern incorporating the Microsoft token.

The next item in the top-to-bottom frequency list is the token more (4,311 occurrences) that was registered as consistently occurring as part of two repeated patterns, namely, more + than and more + adjective. Having mapped these two recurrent patterns as the most representative for the token in question, the study brought them into the focus with some of the examples of the linguistic environment for more provided in Tables 5 and 6 below.

Table 4. Examples of linguistic environment for “Microsoft” registered in the sample

Table 5. Examples of linguistic environment for “more + than” registered in the sample

Table 6. Examples of linguistic environment for “more + adjective” registered in the sample

With a noticeable drop to 2,754 occurrences, Corporate was the next on the list with the results returned showing the token’s clearly evidenced predominant use as part of compound terminological units related to company’s legal, social, administrative and financial activity. Some of the most representative examples are offered in Table 7 below, displaying Corporate capitalised as part of terms and often supplemented with acronyms showcasing the terms’ abridged usages as formal references mostly associated with formalistic technical documentation language.

The next table incorporates examples of rights, responsible, technology, commitment, people, social represented with a close-to-equal frequency of occurrence in the sample (1,202; 1,120; 1,017; 1,015; 1,009; 1,003; 1,000 occurrences for each of the seven tokens respectively) (see Table 8).

Table 7. Examples of linguistic environment for “corporate” registered in the sample

Table 8. Examples of linguistic environment for “rights”, “responsible”, “technology”, “commitment”, “people”, “social” and “environmental” registered in the sample

5. Discussion

Following from the results obtained for the linguistic environment of registered tokens of relevance, this study was equipped to make a number of inferences relying on both the frequency and contextual data analysis. The ten tokens of relevance for this research in the order of the frequency of their occurrence in the sample are  (1) our/we/us, (2) Microsoft, (3) more, (4) Corporate, (5) rights, (6) responsible, (7) technology, (8) commitment, (9) people, (10) social, and (11) environmental. To suggest credible implications for the purposes of this research, the linguistic environment was considered with the following key conclusions drawn.

If any language pattern can be characterised as the most evident and clearly discernible in Microsoft’s CSR reporting narratives, it is the pattern that aims towards displaying the personal approach, which rather apparently stands out in the overwhelming usage of our, we and us as references to the company and its representative bodies, agencies, departments, affiliates, and partners involved in the company’s business operations at different levels and dimensions of activity. This trend towards personalising corporate entities via consistent deployment of personal/possessive pronouns has been explored in scientific literature and has been proven to serve as a powerful tool of corporate self-representation (see Casan-Pitarch 2016, Puschmann 2010, Wei 2020). This study views the consistent use of personal/possessive pronouns in Microsoft’s CSR reporting narratives as a consolidated and very much intended strategy towards reducing the distance between the corporation and the general public, which works towards instilling the general sense of togetherness and creating the ‘we-the-people’ sensation for the target audience. On a deeper level, in a large number of instances the retrieved examples can be argued to appeal to the reader’s sense of empathy, whereby the us­-related reality is the one that the audience also appears to be a part of.

(1) “This places us at a historic intersection of opportunity and responsibility to the world around us”;

(2) “If we continue to live our mission, there is no limit to what we can achieve together”;

(3) “We understand that meeting our responsibility to address these challenges and drive opportunity requires a principled approach, principles that guide us over the long term while inspiring us to take action today”.

The Microsoft token used overwhelmingly in the Microsoft + verb denoting positive action pattern can be argued to testify to a well-calibrated approach to language choice in the company’s CSR reporting. The positively connoted linguistic environment clearly works towards creating positively charged perception, where no wrongdoing whatsoever could possibly be associated with the company name. The varied examples of text fragments compiled this way largely engage verbs in Present Simple, Present Continuous and Present Perfect to refer to actions and activities presented as facts, processes or completed performative measures. At that, the Present Continuous form appears to stand out as it is used abundantly to underline the consistently undertaken efforts to do something, thus conveying the idea of a purposeful and longstanding commitment to push towards completing the required goals. This is in line with Koller’s (2009) theory suggesting that the focus on continuance has become one of the core policies in brand and corporate communications empowering self-representation grounded in the persistence of invested effort. Below are some of the examples registered in the sample testifying to the above argumentation:

(4) “Microsoft is investing to help customers manage resources more effectively using big data”.

(5) “Microsoft is working to create a trusted, responsible, and inclusive cloud”.

(6) “Microsoft is now offering its employees family caregiver leave, a new global benefit that will allow workers to take up to four weeks of fully paid leave to care for an immediate family member with a serious health condition”.

(7) “Microsoft is always looking for the best and brightest talent”.​

The list of some of the most frequently encountered verbs denoting positive action registered in the linguistic environment of the Microsoft token includes spark, believe, strive, invest, empower, advocate, support, help, contribute, leverage, uphold, enhance, ensure, encourage, launch, aspire, work, enable, collaborate, fund, engage, etc.

The next observation somehow derives from the focus on positive perception, as it is rooted in reinforcing the company’s assertive image with the targeted emphasising of its accomplishments that can be both calculated (quantitative assessment) and descriptively ascertained (qualitative assessment). This is implied by the extensive and sustained exploitation of more as a token of interest in the sample, most repeatedly used in the patterns more + than and more + adjective. As has been noticed in the analysis, in many instances the two patterns and their linguistic environment point not just to an accomplishment, but rather an overaccomplishment. This transfers the idea of an exaggerated progress, especially when encountered repeatedly with an insistent accentuation of doing more than expected or even conceivable. This particular pattern adds a sense of triumphalism to Microsoft’s CSR reports, and yet may not come across as too self-aggrandising to the general public because of the copious amount of figures mentioned to substantiate the covert claim for praise. For example:

(8) “The number of Black-owned businesses added to our supplier pipeline has grown more than 30%”.

(9) “Windows 10 is active on more than 500 million devices around the world”.

(10) “Microsoft grantees have trained more than 180,000 teachers who will go on to teach computer science around the world”.

The list of descriptive adjectives accompanying more in the sample is quite extensive and bears witness to the committed appreciation of using consistent persuasive language within the corporate language and communication policies adopted for CSR reporting. The list includes but is in no way limited to adjectives such as equitable, sustainable, inclusive, secure, variable, productive, innovative, trusted, diverse, respectful, accessible, affordable, accurate, resource-efficient, informed, effective, informed, systemic, agile, responsive, secure, flexible, etc.

The abundant use of the more + than and more + adjective patterns can be argued to be reflective of one of the principles of CSR reporting mentioned in Zsóka and Vajkai (2018), which is the principle of comparability that requires CSR reports to offer verified information on the previous and current state of affairs – obviously, in terms of the growth perspective. The principle implies that social and ethical account, overhaul and accountability serve as the fundamental measuring points to assess and compare the activity of different corporations and look into the dynamics of their development in the sphere of corporate social responsibility.

As mentioned in Study and Results, Corporate was the next token to make the list, and it was this token that posed a particularly interesting case in the framework of this research. Corporate is met widely across the sample as evidenced in the data on the frequency of its occurrence, yet it is chiefly registered to be part of a compound term, rather than a descriptive adjective with a linguistic environment that warrants contextual evaluation. This led us to suggest that the high frequency of its occurrence in the sample testifies to the intentional attempts to, in a way, “blur” its actual semantics and dissociate its “loaded” meaning from the company’s image painted in the CSR reports. In other words, by using corporate in an almost exclusively formulaic context, the texts build a narrative that distances the brand from its corporate background. Not to be repetitive, the study lists some of the most representative examples supporting this argument in Table 7 above.

As Table 8 above has illustrated with a number of examples, the rest of the most frequently used tokens found in the sample correlate with the topics of importance for CSR reports and are used as part of linguistic environment that helps shape effective CSR reporting narratives. Thus, the following intended meanings were correlated with the corresponding tokens:

rights – used to place into focus social entitlement as one of the key prerogatives governing corporate activity and goal setting (e.g., human rights, fundamental rights, basic rights, cultural rights, privacy rights, users’ rights, legal rights, people’s rights, etc.);

responsible – used to lay emphasis on duties and roles both stemming from corporate operations and seemingly outside of corporate activity (e.g., responsible sourcing, responsible governance, responsible practices, responsible supply chain standards, responsible innovation, responsible decision making, etc.);

technology – used to showcase core activity as a virtue (e.g., the power of technology, to advance technology, access to technology, accessibility in technology, benefits of technology, affordable technology, the role of technology, building trust in technology, etc.);

commitment – used as a promise of dedication and investment of consistent efforts stemming from genuine concern (e.g., commitment to customer security and privacy, commitment to ethical business practices, carbon neutrality commitment, commitment to respect human rights, commitment to address social injustice, deepening the commitment to diversity and inclusion, remain resolute in our commitment to make a positive impact around the globe, etc.);

people – used as a recurrent reference to key stakeholders (e.g., to empower people across the planet, protecting people from digital threats, to improve people’s quality of life, talented/young/creative people, to support people, bringing people and organizations together, etc.);

social – used a recurrent reference to the welfare beneficial to the society (e.g., social impact, social benefits, social opportunities, to advance social goals, social matters, social psychology, social capital, etc.).

environmental – used to place into focus the efforts correlated with maintaining sustainable green practices (e.g., environmental sustainability, environmental impact, environmental accountability, environmental innovation initiatives, environmental protection, environmental data, environmental regulations, environmental challenges, environmental science, environmental footprint, etc.).

6. Conclusion

This study set out to investigate the language of CSR reports using a self-made corpus of Microsoft’s social responsibility documentation to ultimately answer two research questions: what predominant language patterns are used to ensure efficient CSR reporting, and which priority directions in CSR narrative composition are evidenced in language use.

The findings relying on the frequency of occurrence of the most recurrent tokens were used to establish and analyse the “go-to” language patterns that included personal/possessive pronouns us, we, our referring to the company; Microsoft + a verb denoting positive action; more + than; more + adjective; Corporate as part of compound terminological units, as well as a set of key thematic tokens – rights, responsible, technology, commitment, people, social, environmental – carrying descriptive linguistic environment with positive connotation.

The corresponding priority directions in CSR narrative composition evidenced in language use were deduced from contextual analysis and can be summarised as follows: (1) displaying the personal approach, (2) creating positively charged perception of the company, (3) emphasising accomplishments through quantitative and qualitative assessment, (4) distancing the brand from its “loaded” corporate background, (5) placing into focus social entitlement as one of the key prerogatives governing corporate activity and goal setting, (6) laying emphasis on duties and roles both stemming from corporate operations and seemingly outside of corporate activity, (7) showcasing core activity as a virtue, (8) offering a promise of dedication and investment of consistent efforts stemming from genuine concern, (9) maintaining consistent reference to key stakeholders, (10) maintaining recurrent reference to the welfare beneficial to the society, and (11) placing into focus the efforts correlated with maintaining sustainable green practices.

The findings have thus suggested some comprehensive takeaways shedding light on the linguistic composition of efficient CSR reports which reflect the essential principles guiding the linguistic manifestation of the modern corporate culture. Some further research into the issue may include studies on the incremental change in CSR reporting practices, where a year-by-year evaluation of priority shifts might be deduced from the changing focus on different linguistic strategies applied in corporate reporting. Another possible direction of research could involve exploring the issue of precedence (Malyuga & Akopova 2021) to characterise the linguistic strategies and patterns of CSR reports that appeal to familiar notions to attract the audience’s attention.

About the authors

Elena N. Malyuga

Author for correspondence.
Email: malyuga-en@rudn.ru
ORCID iD: 0000-0002-6935-0661

Professor of Linguistics, Head of Foreign Languages Department at the Faculty of Economics, Peoples’ Friendship University of Russia, Doctor Habil. of Linguistics, Academician of Russian Academy of Natural Sciences, Editor-in-Chief of the research journals Issues of Applied Linguistics and Training, Language and Culture. Her research interests include theory and practice of intercultural professional and business communication, pragmatics, corpus studies and discourse analysis. She is author and co-author of over 300 publications.

References