Sensemaking and corporate social responsibility: Implications for stakeholder communication amid the COVID-19 pandemic

ambiguity induced by the pandemic. From the literature, we gleaned that sensemaking happens in response to circumstances that present complex and ambiguous issues that organisations need to understand (Ancona, 2012; Weick, 1995; Weick, Sutcliffe, & Obstfeld, 2005). Regarding the corporate social responsibility (CSR), we found resonance in the work of Basu and Palazzo (2008) who regard CSR as a dynamic and unfolding process of sensemaking that shapes the CSR orientation of an organisation in a stakeholder context. Based on the extant literature Purpose: The aim of this research study was to establish via a linguistic analysis of stakeholder communication texts, how organisational sensemaking in a crisis situation, such as the coronavirus disease 2019 (COVID-19) pandemic, might influence the social responsibility orientation of companies. Design/methodology/approach: Data were collected from public social responsibility statements of 10 financial services companies before and immediately after the advent of the COVID-19 pandemic and analysed using the Linguistic Inquiry Word Count (LIWC) method. In this way, qualitative information could be turned into statistical data and made comparable across the sample. Results: The results confirm that the social responsibility sensemaking processes in companies manifest in their public stakeholder communications. While the analysed statements indicate an increase in affiliation with the predicament of stakeholders, it also reveals reduced levels of transparency. This tension between affiliation and transparency holds consequences for authenticity and reputation in corporate social responsibility (CSR) sensemaking and orientation. Practical implications: Corporate decision-makers and communication experts should take special care of the relationship between CSR sensemaking and stakeholder communication as interconnected processes in a purposeful and mindful manner. Originality/value: The research study demonstrates how the analysis of textual statements can be used as a lens into CSR sensemaking in the context of disruption and adversity. This establishes a method for measurement of CSR orientation that can assess an organisation’s current orientation, as well as shifts in orientation because of crisis situations. Organisations can use this information to evaluate their CSR communications so that they are more aligned with intentions.

, we argued that stakeholder communication would mediate the outcomes of 'CSR as sensemaking', and thereby portrays the CSR character and orientation of a company.There is, therefore, a stakeholderminded and linguistic thread that links the corporate sensemaking, social responsibility and communication aspects at stake in this research project.
At the time of the design and data gathering for this research (April 2020 -July 2020), there was a dearth of scholarly literature on the pandemic's meaning and impact for business organisations.Neither was there a favourable prospect for carrying out interviews with corporate decisionmakers, whose attention was absorbed by pandemic-related crisis management responsibilities.Methodologically, we therefore opted for a linguistic analysis of written COVID-19-related CSR statements by making use of the computerbased Linguistic Inquiry Word Count (LIWC) application spearheaded by Chung & Pennebaker (2013).We selected 10 Johannesburg Stock Exchange-listed financial service companies of which we analysed, per organisation, one pre-and one post-lockdown CSR statement.The results of the research study revealed that companies responded with changes in CSR orientation.All the companies analysed shifted to a 'predominant affiliation motivation' during the crisis, meaning that the analysis showed that the language and phrasing of COVID-19-related statements were driven by a need for communal connection with their audience.The analysis also revealed that company communications were predominantly written in legal-toned language and showed low levels of transparency across the board.These results provide valuable insights into how CSR communications might be perceived by readers.The opportunity here is for organisations to re-examine the way information is conveyed via written media and to align language and phrasing with CSR objectives.
In the next section of this article, attention is paid to the literature review with specific reference to sensemaking, CSR and stakeholder communication.This is followed by an explanation of the LIWC method of linguistic analysis as applied to the selected texts.The article concludes with a discussion of the results and the consequences thereof for the CSR orientation of companies.

Literature
The COVID-19 pandemic confronted corporate decisionmakers with the challenge of balancing long-term business continuity with the immediacy of the social needs of their stakeholders.This realisation necessitated for us an exploration of three interrelated concepts in the literature: (1) organisational sensemaking amidst ambiguity and disruption, (2) CSR as an expression of sensemaking and (3) stakeholder communication as a medium for mediating the outcomes of such responsibility-orientated sensemaking to stakeholders.

Organisational sensemaking
Sensemaking as a construct has been primarily ascribed to Weick (1995) who views it as a process in which interdependent people comprehend ambiguous external circumstances to determine action.Weick (1995) understood sensemaking as: [A] process that is (1) grounded in identity construction, (2) retrospective, (3) enactive of sensible environments, (4) social, (5) ongoing, (6) focussed on and by extracted cues, (7) driven by plausibility rather than accuracy.(p.17) In a later publication, Weick & Sutcliff (2005) confirm that the degree of importance of the properties of sensemaking is dependent on the actual scenario and incident that initiates the process.Sensemaking is initiated through circumstances that result in ambiguity or confusion (Caughron et al., 2011;Maitlis & Christianson, 2014;Odden & Russ, 2019;Weick, 1995;Weick et al., 2005).Some studies expanded sensemaking from a micro-level theory to a macro-level neo-institutional context (Helms Mills, Thurlow, & Mills, 2010;Weber & Glynn, 2006), which highlights the interactions between individual sensemaking processes and the actions of organisations.
The literature on sensemaking contains a pertinent focuses on language and communication.Taylor and Van Every (Weick & Sutcliff, 2005, p. 409) refer to sensemaking as a way station at which 'circumstances are turned into a situation that is comprehended explicitly in words as a springboard for action'.According to Gioia et al. (Weick et al., 2005), there are three activities happening at this way station: (1) organisational circumstances are turned into words, (2) organising itself gets embodied into written and spoken texts and (3) these textsin reading, writing, conversing and editing -become the media through which organisations shape conduct.Consequently, Weick et al. (2005, p. 409) state that 'sensemaking is, importantly an issue of language, talk and communication'.When people, whether individually or collectively, find themselves in circumstances of disruptive ambiguity, they become occupied with two questions, 'what is the story?' and 'now what?'.Answers to the first emerge from retrospect, experience and dialogue, while answers to the second emerge from presumptions about the future and the articulation of actions and projects in response (Weick et al., 2005).
Because of its global impact, we argue that the COVID-19 pandemic inaugurated an occasion for sensemaking and confronted leaders and organisations with questions about 'what is going on' and 'how to respond'.Answering these questions will, inevitably, among other things, hold consequences for their decisions and actions regarding both their business and social responsibilities and what and how they communicate consequently to stakeholders.

Corporate social responsibility as a process of sensemaking
As Carroll's well-known pyramid of corporate responsibilities (Carroll, 1979), the concept of CSR evolved towards a more systemic understanding (Elkington, 1998) and onwards to include linkages to strategy (Lamberti & Lettieri, 2009;Porter & Kramer, 2006;Werther & Chandler, 2006), competitiveness (Vilanova, Lozano, & Arenas, 2009), measurement and reporting (Hopkins, 2005), and integrated value creation (Visser & Kymal, 2015).Definitions of CSR abound (Dahlsrud, 2006) and the meaning and scope thereof may differ from one context to another (Matten & Moon, 2008).Some authors also highlight different theories (Garriga & Mele, 2004) and motivations (Galbreath, 2006) for why companies will engage in CSR.Based on these perspectives, we may conclude that companies are socially-embedded institutions, whose overall well-being is simultaneously interconnected with that of their stakeholders and intertwined with the economic, social and ecological conditions of the context within which they operate.Corporate social responsibility, therefore, implies that companies, in their decisions and actions, should be mindful of their relationship with and the wellbeing of their stakeholders (The King Committee, 2016).Such a mindful responsibility orientation, we argue, demands a process of sensemaking from corporate decisionmakers.Cramer, Van der Heijden and Jonker (2004) conducted research on how companies make sense of CSR.Grounded in Weick's (1995) work on sensemaking in organisations, they argue that 'CSR offers first and for all a framework in which people can construct meaning' (Cramer et al., 2004, p. 216).For this process, they observed that organisations use linguistic artefacts (words and notions), and they carry out practical projects, as a result of which CSR becomes a process of sense-making in action (Cramer et al., 2004, p. 218).
A similar, but much more pertinent, connection between sensemaking and CSR is to be found in the approach of Basu and Palazzo (2008).Whereas Cramer et al. (2004) enquire about how companies make sense of CSR, Basu and Palazzo propose that CSR is by its very nature a process of sensemaking.On this basis, CSR does not result from responsivity to external demands, but 'from organizationally embedded cognitive and linguistic processes' (Basu & Palazzo, 2008).Their process model of sensemaking proposes an explanation of how decision-makers think, discuss and act with regard to their key stakeholders.Consequently, they define CSR as: [T]he process by which managers within an organisation think about and discuss relationships to stakeholders, as well as their roles in relation to the common good, along with their behavioural disposition with respect to the fulfilment and achievement of these.(Basu & Palazzo, 2008) Based on their choice for a process model of CSR sensemaking, Basu and Palazzo (2008) posit a set of cognitive, linguistic and connotative processes to better understand a firm's CSR impact.In this tripartite view, each process consists of the following sub-dimensions (Basu & Palazzo, 2008, p. 125): • Cognitive: Referring to how an organisation thinks about its relation to stakeholders and the broader world, this process reveals an organisation's identity orientation (individualistic, relational or collectivistic) and its legitimacy orientation (pragmatic, cognitive or moral).• Linguistic: Referring to how an organisation explains and shares it reasons for engaging in specific activities, this process reveals its basis for justification (legal, scientific, economic or ethical) and its orientation towards transparency (balanced or biased).• Conative: Referring to how an organisation expresses its behaviour in terms of posture (defensive, tentative or open), its consistency (both externally and internally) and its commitment (whether instrumental or normative).
As the postulation of Basu and Palazzo's model, it has been utilised in diverse research contexts.Peters, Waples, and Golden (2014) used it in a conceptual article to connect CSR, in cognitive, linguistic and conative sensemaking terms to strategic decision-making in organisations.Cosenza, Ribeiro, Levy and Dios (2018) used it as a methodological framework for analysing the CSR orientation of Samarco, the Brazilian mining company, the tailings dam of which collapsed in 2015, leaving 21 people dead and 700 homeless.By studying both textual (a sustainability report and news articles) and visual material (videos), Cosenza et al. (2018) were able to characterise Samarco's CSR orientation across all three of the sensemaking processes, namely cognitive, linguistic and conative.Richter and Arndt ( 2018) followed a more selective approach by utilising only the cognitive (identity and legitimacy) dimensions to study and make conclusions about the CSR orientation of British Tobacco Switzerland.Following this pathway, Richter and Arndt pursued perspectives on how organisations in contentious industries (e.g.tobacco) might strive to achieve legitimacy.
In a similar vein than the abovementioned examples of the use of Basu and Palazzo's (2008) model, we also had to consider which of the processes relate best to our research interest.In view of the mediating role that we ascribe to stakeholder communication as a channel for CSR as sensemaking, we opted to focus on the linguistic process, with its two sub-dimensions of justification and transparency.
In the case of the former, the focus is on how an organisation justifies its actions to stakeholders, based on their CSR orientation and expresses in either legal, economic, scientific, or ethical terms.In the case of the latter, CSR communication might be either balanced, irrespective of whether the contents thereof will be positively or negatively perceived, or biased, by communicating only that which is positive and omitting the negative.We will return to this again in the Methodology section.

Corporate social responsibility orientation mediated through stakeholder communication
The stakeholder concept has its origin in strategic management (Freeman, 1984) and over time found its way into management theory (Donaldson & Preston, 1995), CSR (Freeman & Velamuri, 2006) and corporate governance (Freeman & Evan, 1990;The King Committee, 2016).It is commonly defined as 'any group or individual who can affect or is affected by the achievement of the organisation's objectives' (Freeman, 1984) or, more broadly, 'groups united by a sense of shared interests, values and purpose' (Dunham, Freeman, & Liedtka, 2006).Because of the complexity inherent to stakeholder diversity, theories and frameworks for stratification and segmentation have been developed (Freeman, 1984;Mitchell, Agle, & Wood, 1997), as well as standards against which stakeholder management can be benchmarked (Accountability, 2015).
Organisations provide information to stakeholders through primary and secondary communications.According to Balmer (2017), primary communications involve latent interactions that inform company activities, behaviours and strategic initiatives, while secondary communications are the formal, structured interactions that form part of public relations, marketing, advertising and branding.Both forms of interaction are crucial to an organisation's identity and the way this is perceived by the recipient of the interaction.
From a stakeholder management perspective, engaging with stakeholders in a material way is essential for long-term value creation as well as fostering a shared understanding between organisations and stakeholders (Morsing & Schultz, 2006).For stakeholder communication to be effective, it needs to be interactive, dialogue-based and open to feedback.Morsing and Schultz (2006) identify three stakeholder communication strategies in relation to CSR.Firstly, there is the stakeholder information strategy, which is linear, one way, information driven and serves as a method of sensegiving to recipients (e.g.brochures, pamphlets and publicly available factual information).Secondly, there is the stakeholder response strategy, which is a two-way model that allows for feedback from recipients that can be incorporated into the organisation's sensemaking process.Finally, there is the stakeholder involvement strategy that is both two-way and dialogue-based and allows for influence and persuasion, sense-giving and sensemaking, from both the organisation and the recipients.
According to Colleoni (2013), citizens are increasingly demanding that corporations justify their economic, social and environmental actions to the general public.Furthermore, communication needs to be customised, appropriate and clear to avoid mismatches of intent and interpretation (Kinawy, Nik Bakht, & El-Diraby, 2017).In crisis situations, the choice of communications strategies becomes a critical factor in the success or failure of stakeholder outcomes.Consideration of all stakeholders and ethical behaviour in crisis can contribute to the effective handling of crisis situations (Xu & Li, 2013).
In summary, the literature review thus points towards a connection between CSR and sensemaking in organisations mediated to stakeholders via communication.Should the external conditions change, such as with the unforeseen advent of a pandemic or similar crisis, the sensemaking of what corporate responsibilities attend to and what to communicate to stakeholders about it, will be affected as well.Based on this assumption, the main interest of this research is how stakeholder communication might reflect changes in CSR orientation before and during a crisis.The establishment of this connection provides an avenue for the assessment and measurement of CSR orientation.

Methodology
The study required the identification of specific linguistic content that relates to the sensemaking perspective and CSR orientation.A summative approach to content analysis was appropriate for this requirement.The summative approach identifies specific keywords that are determined before and during the analysis of the data.The relevance of the keywords, or codes, is informed by the literature review of this study (Hsieh & Shannon, 2005).Originally, we considered data gathering through in-depth interviews with key decisionmakers in a sample number of organisations.However, given that this research was commenced in April 2020, in the immediate aftermath and disruptive impact of strict lockdown conditions, we assumed that corporate decisionmakers would be occupied by the demands of COVID-19related crises and opted for a study of written stakeholder communiques.
The LIWC method and software is a suitable instrument for standardised analysis and comparison purposes (Chung & Pennebaker, 2013;Tausczik & Pennebaker, 2010) of corporate communication artefacts.The LIWC reflects thinking patterns, social concerns, speech patterns and emotions for specific use by researchers interested in cognitive psychology.Having been validated as a viable mechanism for research (Tausczik & Pennebaker, 2010), the LIWC has been used to measure, among other things, the emotional expression of text documents (Kahn, Tobin, Massey, & Anderson, 2007), to predict the course performance of students (Robinson, Navea, & Ickes, 2012), to analyse the psychological state of participants in media interviews (Pennebaker, Slatcher, & Chung, 2005) and as language-based measure of personality in a digital world (Boyd & Pennebaker, 2017).The literature search on LIWC application produced no evidence of it being previously used for the analysis of CSR-related texts as proposed in this research.
Financial service companies were targeted as ideal for the purposes of this research.They provide tangible and intangible benefits to clients and require a large degree of trust from their stakeholders, especially during times of disruption and adversity.Ten financial service companies listed on the Johannesburg Stock Exchange and all with significant market share and brand recognition were selected for this study.Among those chosen were two banks, four insurance companies and four investment companies.For each company, two publicly available text statements were selected, the criteria being that they were (1) directed at the public, including their more direct stakeholders; (2) official and written by, or on behalf of, the highest decision-making authority of the organisation and (3) related to matters of relevance for the company-stakeholder relationship.Furthermore, the search function for each company site was used to obtain the first official correspondence that mentioned COVID-19 with organisational specific comments and/or action.These statements formed the primary data set for the study.
To provide a suitable comparison reference point, 10 statements issued by these companies that did not make mention of the pandemic were selected.These statements were consistent with the three criteria mentioned above and expressive of CSR orientation in relation to products, services and initiatives.Topics in these statements ranged from personal finance education to updates on community benefit initiatives.These 10 statements form the 'Control group' of the study and provide an appropriate benchmark to which the COVID-19 statements could be compared.Contrary to the control statements, the COVID-19 ones reflected circumstances representing ambiguity and represent a rather dramatic shift in both tone and intent and are focused on the same reality.These can broadly be characterised by expressions of corporate identity or citizenship: an assurance to stakeholders of business continuity, expressions of empathy and pledges of different forms of assistance to absorb the impact of the pandemic.
Having established the validity for inclusion in the research sample, the LIWC was used to analyse both sets of statements.The process model of sensemaking (Basu & Palazzo, 2008) provides a basis for evaluation of how organisations think, discuss and act in relation to CSR.Of the three dimensions of the process model, the linguistic dimension is relevant to the data that were analysed.
The linguistic dimension provides specific categories that can determine orientation to CSR sensemaking, with reference to stakeholder communication.Analysis using this dimension requires consideration on the communications on justification and transparency.Justification can be assessed by which of the four language types (economic, legal, scientific and ethical) dominates the communication.
Transparency is assessed by determining whether the communication is balanced or biased.
On its standard setting, the LIWC uses pre-set dictionaries to make an assessment on five variables that can be used as dimensions relevant to linguistic, psychological and social processes.The program assesses text documents digitally and assigns a score to each dimension.The score assigned takes the form of a 100-point scale.The five dimensions are analytical thinking, clout, authenticity, emotional tone and drive.Descriptors on each of the dimensions are as follows (Chung & Pennebaker, 2013).
The analysis of the LIWC results requires the conversion of a quantitative data value (the score per dimension) into a qualitative description of the relevant dimension (Pennebaker et al., 2016).To promote consistency in analysis, a formal method for the qualitative descriptors of strength was adopted (Rosenthal, 1996).The scores were separated into ranges, with an applicable strength level linked to the primary descriptor of the dimension.

Ethical considerations
The Departmental Ethical Screening Committee of Stellenbosch Business School reviewed the research proposal and confirmed it as exempt from ethical clearance (USB-2020-17232, 17 August 2020).

Results
The results were averaged, and the change in average between the scores of the COVID statements compared with the average scores of the control statements is reflected in Figure 1.An explanation of each dimension, the relevance of the LIWC score and the CSR link is presented in Table 1.
The drive score is a percentage measurement of the presence of motivation or needs exhibited in written communications.The total score is made up of individual scores for the five different classifications of motivation, namely, affiliation, achievement, power, rewards and risks (Pennebaker et al., 2016).As presented in Figure 2, the classifications that made up the main drive score were ranked per COVID-19 statement and control statement.The ranking provided a scale for each of the five drive dimensions.The ranking is in the order of

Analytical thinking
The higher the score, the more the document is presented as formal, logical and implies hierarchal thinking.The lower scores imply informality, personal orientation, here-and-now and narrative thinking.
The analytical thinking dimension relates to the justification dimension of the sensemaking process model (Basu & Palazzo, 2008).

Authenticity
A higher score implies honesty, disclosure, and transparency.A lower score indicates distancing and guardedness.
The authenticity dimension relates directly to the transparency dimension of the sensemaking process model (Basu & Palazzo, 2008).

Drive
The motivations and intentions behind communication can be categorised as affiliation (collaboration, being a part of a group and belonging); achievement (success and avoidance of failure); power (authority and dominance); reward (incentives, goal achievement and rewards) or risk (avoidance of risk and uncertainty).
The drive dimension was included because one of the main uses of a sensemaking perspective is the analysis of motivations and intents in organisational decisions and behaviours (Caughron et al., 2011).

Emotional tone
The higher the score, the more association with positivity and an upbeat tone.A score close to 50 implies indifference or an unemotional tone.
Lower scores indicate pessimism, anxiety and sadness.A very low score can imply hostility.
The emotional tone dimension was included because of the relevance and impact of emotional cues in crisis period decision-making has been established (Sayegh, Anthony, & Perrewé, 2004).

Clout
This dimension relates to the perspective of authority and confidence in the subject matter.A high score relates to higher levels of certainty, confidence and an authoritative orientation.A low score implies humility, uncertainty and a very low score implies anxiousness.
The clout dimension was included because of organisational confidence that is projected to stakeholders presenting as a sensemaking cue that affects the stakeholders' perception of the capability of the organisation (Ancona, 2012).
LIWC, Linguistic Inquiry Word Count; CSR, corporate social responsibility.
prevalence of that specific dimension according to the results.Therefore, a rank of 1 in the power dimension and 5 in the risk dimension means that power was the most prevalent motivation and risk was the least.
The averaged results in Figure 1 revealed that the analytical, clout, tone and authenticity dimensions did not change by more than 10%.The largest change was that of the authenticity dimension that decreased by 9.1% in the COVID-19 statements.This implies a more guarded and less transparent positioning of the COVID-19 statements.The average of the clout dimension increased by 7.9%, implying companies positioning with higher levels of confidence and certainty.
There was a slight increase of 4.9% in the analytical tone, which manifests in the formality of prose during the COVID-19 period.There was a slight decrease in the emotional tone of the COVID-19 statements of 3.0% in the tone dimension.Overall, these dimensions did not change significantly enough from control statements to COVID-19 statements to make significant conclusions for the context of the research.
The results displayed in Figure 2 reveal that all the COVID-19 statements show evidence that affiliation is the main motivational factor.The shift was significant, with the average percentage scores for affiliation being 48% of the total drive dimension.In the control statements, the highest percentage score of any of the other drive descriptors was 36%.The implication being that affiliation becomes the most important motivational factor by a considerable margin.The majority of the statements (70%) had power as the second most prevalent motivation, even though the relative percentage scores were low.From the control group, 60% of the statements had power as the most prevalent motivation.This shift in drive from power to affiliation is noteworthy and will receive further attention in the next section.The other motivational factors were mixed in prevalence, with significantly less focus on affiliation relative to the COVID-19 statements.

Discussion
According to Richter and Arndt (2018) 'CSR character is not static, but dynamic and changing, particularly when firms face increasing pressure to legitimate themselves in changing environments'.The linguistic analysis used in this research reveals how the companies under consideration showed a change in CSR orientation from the period before the pandemic to the immediate aftermath of the lockdown announcement.All the companies shifted to a predominant affiliation motivation, meaning that the language and phrasing of COVID-19-related statements were driven by a need for communicative connection with stakeholders.Organisations, therefore, seem to focus on stakeholder affiliation during periods of crisis.
Viewed through the lens of the LIWC analysis, the shift towards affiliation deserves a closer look.Although the results of the investigation showed little change between the control and COVID-19 statements in terms of the analytical, clout, tone and authenticity dimensions, the change in motivation and intention measurements were noteworthy enough to conclude that the national crisis shifted the drive focus behind those organisations.Whereas in a 'normal market', companies can focus on what makes them distinctive and competitive -even in terms of expressions of social responsibility -a 'crisis market' calls on all of them to focus on one overarching reality, namely the impact of the external conditions on both themselves and their stakeholders and the consequences thereof for maintaining important relationships.especially valid in cases where companies, in the long run, fail to make good on CSR pledges such as those expressed in the COVID-19 statements.
The LIWC analysis of the analytical dimension of the communication revealed a strong orientation towards formal and logical language present in both the control and COVID statements.This supports the conclusion that the linguistic dimension of financial services companies has an orientation towards legal and scientific language.This observation is important for financial services companies because of the regulated nature of the industry on the one hand and, on the other, the exposure to risk that these companies need to bear.While taking a mitigating stance in the interest of both regulation and risk, financial service companies can also not afford to appear without empathy and solidarity when both their customers and society, in general, are faced with crisis and adversity.How companies manage and communicate about this tension provides valuable insight into how their CSR character may be perceived by stakeholders.The tension can be breached through the use of ethical justification, which involves high levels of transparency in terms of organisational actions (Basu & Palazzo, 2008).The opportunity here is for companies to re-examine the relationship between their internal sensemaking processes and their external communication, as conveyed via written media, ensuring they are aligned between formal expressions of social responsibility and the real-life experience of the same by their stakeholders (Colleoni, 2013).
Finally, the authenticity dimensions of the COVID-19 statements deserve further interpretation.The change in the average score of transparency for the statements from control to COVID was small.There were, however, individual firms that made significant shifts in their perceived transparency because of the national crisis.The changes did not show consistency, as some firms increased transparency, while others decreased it.While insufficient evidence exists for a general conclusion on this dimension, the evidence supports the assumption that financial services companies' CSR communications to stakeholders may tend to be guarded and therefore also low on transparency.While from a CSR perspective, high levels of transparency are required for a positive CSR orientation (Basu & Palazzo, 2008;Harrison & Wicks, 2013;The King Committee, 2016), this observation also points towards the complexities involved in the sensemaking of a crisis itself when decisionmakers and communication experts need to deal with the 'what is the story?' and 'now what?' questions at stake (Weick et al., 2005).

Conclusion
This research study confirms the connection between sensemaking, CSR and stakeholder communication, or in the words of Cramer et al. (2004, p. 216) 'CSR offers first and for all a framework in which people can construct meaning'.If we accept that contextual changes of a disruptive nature nudge individuals and organisations towards sensemaking (Caughron et al., 2011;Maitlis & Christianson, 2014;Odden & Russ, 2019;Weick, 1995;Weick et al., 2005), it goes without saying that the outcomes of the sensemaking process itself will have consequences for organisations internally and how they relate and communicate with their stakeholders externally.
The research, furthermore, validates Basu and Palazzo's (2008) sensemaking approach to CSR based on linguistic processes among decision-makers with consequences for both stakeholder relationships and the common good.The statements upon which this research is based provide textual evidence of sensemaking language combined with shifts in CSR orientation in the context of changing circumstances and fortunes.While this research project narrowed Basu and Palazzo's (2008) model down to the linguistic dimension, evidence of both its sub-categories, namely justification and transparency, could be identified in the textual statements via the LIWC analysis.
One of the main uses of a sensemaking perspective is the analysis of motivations and intents in organisational decisions and behaviour (Caughron et al., 2011).The results of the LIWC analysis reveal a distinct shift in the prevalent motivation factors to affiliation in the COVID statements, compared with the control statements.This evidence provides confirmation of the change in intentions because of the COVID crisis.The specific intention of affiliation in this context refers to the need of the organisation to present as part of the community of which the reader (stakeholder) of the statement is a part (Chapman & Schwartz, 2010).
From a methodological perspective, the LIWC has been found to be a useful instrument for realising the objective of this research project, namely to investigate whether and how stakeholder communication might reflect changes in corporate responsibility commitments before and during a crisis.The LIWC provided a methodological bridge between the definitive theoretical constructs of CSR, sensemaking and stakeholder communication on the one hand and the in-context statements of companies on the other.The LIWC may therefore also be useful in other scenarios of text-based analysis in areas such as marketing, communication, public relations and social media management.
The results of this research are of practical relevance for corporate decision-makers and communication experts, especially in a context of crisis such as with the COVID-19 pandemic.Firstly, the research confirms the interrelationship between sensemaking, CSR and stakeholder communication.
In practice, it means that organisations' cognitive and emotional processing of ambiguous or disruptive situations will have consequences for their stakeholder relationships and text and media-based communications.Secondly, the research points toward the importance and necessity of sensemaking to be approached purposefully and with full awareness of both the concrete circumstances and the stakeholders involved.A crisis provides opportunities for redefining both CSR priorities and relationships, but it can also expose fault lines in character and inconsistencies in commitments with reputational consequences.The latter is especially true if there is a gap to be found, perceived or real, between the CSR promises and practices of an organisation.
The limitations of this study also serve as opportunities for further research.The first limitation is that the research was only text-based, resulting in no intimate knowledge of the internal sensemaking processes that played out in any of the 10 companies involved, both at the advent of COVID-19 and thereafter.This shortcoming poses an opportunity for in-company follow-up research via other qualitative methods.The second limitation relates to the focus on only one industry, namely financial services.Consequently, there are no conclusions to be made about the CSR sensemaking shifts in other industries.However, using the same methodology should make it possible to replicate the same project in other industries, also hard hit by the pandemic, for example, retail, tourism and hospitality and manufacturing, and compare the results.Finally, it needs to be mentioned that since the completion of this study there was a proliferation of COVID-19-related research articles in the CSR domain (Carroll, 2021;Crane & Matten, 2021;Giacomini, Martini, Sancino, Zola, & Cavenago, 2021;Manuel & Herron, 2020;Mata, Buil, Gómez-Campillo, & Gomez-Campillo, 2021;Patuelli, Caldarelli, Lattanzi, & Saracco, 2021;Stephens et al., 2020) that will greatly enhance the theoretical foundations and theoretical outcomes of future studies.
This research holds an important message for corporate decision-makers and communicators, especially at a time of disruption and adversity.Sensemaking, whether carried out inadvertently or mindfully, will have consequences for how stakeholders will perceive companies' social responsibility orientation and actions.Authenticity is in the eye of the stakeholder too.

TABLE 1 :
Corporate social responsibility and linguistic sensemaking dimensions.