Abstract
Festivals and events are becoming an increasingly important part of economies around the world. However, they can result in negative impacts to the environment via waste and carbon emissions. Wonderfruit, a popular annual arts and music festival held in Thailand, espouses eco-friendliness and environmentalism among its key tenets. Perhaps in line with these principles, Wonderfruit is also one of the few festival organizers that publishes corporate social responsibility (CSR) reports. This study utilizes corpus linguistic and discourse analytical frameworks to explore Wonderfruit’s social responsibility reports published between 2019 and 2024 in order to understand how the festival navigates this form of corporate communication. The analysis of key grammatical and semantic tags from the festival’s CSR reports indicates that over this period the festival moved away from sustainability and environmental discourses toward a more promotional form of discourse. Methodological implications for approaching small datasets using corpus linguistic methods are also discussed.
Keywords
Introduction
Founded in 2014, Wonderfruit is an annual large-scale event held in Thailand. Taking place over 4–5 days, the event incorporates live music performances, food and beverage activities and events, art installations and art projects, cultural activities showcasing both local Thai and Asian culture, wellness activities such as yoga classes, and environmental activities, such as planting trees. With the 2024 iteration garnering over 25,000 attendees from around the world, it has rapidly grown into what is now one of the largest festivals in Asia. The event is organized by a Thai production company headed by the two original founders of Wonderfruit. According to Wonderfruit’s website, these founders desired to create a movement that was both fun and engaging, but also importantly, sustainable (Wonderfruit, 2017).
One of the core branding aspects of Wonderfruit is that of environmental sustainability and eco-friendliness (Wonderfruit, 2023), while the mission of the festival is to “encourage, develop and innovate creative solutions for sustainable living and bring together a global community to celebrate them” (Wonderfruit, 2019). Thus, even though this event incorporates activities akin to many other music and art festivals, one relatively unique aspect of Wonderfruit is the apparent centrality of sustainability in both their festival practices and branding.
Perhaps with these eco-awareness beliefs in mind, Wonderfruit began publishing annual reports on their sustainability initiatives in 2019. To date, research on social responsibility reports has mainly focused on large corporations (see Bhatia, 2012; Fuoli, 2012, 2018; Lin, 2021; Zappettini and Unerman, 2016, for examples). Smaller non-Western business entities have received much less attention. At the same time, Nwagbara and Belal (2019) have noted a turn toward discourse analytical approaches to studying social responsibility reports. Moreover, festivals, and more broadly, the tourism industry, do not typically produce social responsibility reports. Thus, it would be interesting to explore how smaller non-Western business entities in this particular industry, especially those with environmentalism at their forefront, navigate this type of corporate discourse. At the time of writing, we had access to four reports (2019, 2022, 2023, and 2024) and during our initial examination of them, we noticed aspects of the layout and language use that had changed across these four periods. Rather than picking out a few obvious examples, we wanted to carry out a more systematic comparison, which would help us to identify subtle and repetitive instances of language use that might otherwise be overlooked. Within the framework we use in this paper (Corpus-Assisted Discourse Analysis, Partington et al., 2013), a prospective and exploratory approach is taken, meaning that aims are purposefully left open, allowing the corpus procedures to drive the analysis – we have to account for what they identify as statistically salient or frequent. Therefore, this study begins with the broad aim of investigating the publicly available sustainability reports published by Wonderfruit in an effort to understand how this organization navigates this form of business reporting, while at the same time managing and representing their company values.
Literature review
Corporate social responsibility reports
Corporate social responsibility reports (henceforth, CSR) are reports released by companies which describe how their business practices impact on society and the environment (Bhatia, 2012; Zappettini and Unerman, 2016). These reports often detail how the business has attempted to minimize any environmental damage their commercial activities may cause, with a focus on current, past, and/or future sustainable actions and initiatives. Originally, CSR reports were mainly published by large, multi-national, publicly-listed corporations (Deegan, 2002; Stubbs et al., 2013). However, they have since spread to a wider variety of businesses, with companies of different sizes and different sectors taking on this practice (Higgins et al., 2015, 2018).
Research on CSR reports has mainly focused on the motivations, trends, and contradictions of CSR reporting. Although usually voluntary (Bhatia, 2012), there are a plethora of reasons why CSR reports are becoming widespread (Yu and Bondi, 2019). From a business perspective, CSR reports can lead to better profitability for companies (Morhardt, 2010; Thompson, 2018), while for stakeholders, it has been claimed that the average consumer now has more environmental awareness and therefore demands more from businesses (Brugman et al., 2024; Frandsen and Johansen, 2001; Fuoli, 2018; Gâţă, 2017). For example, Tregidga et al. (2014) have argued that reporting of sustainability practices helps organizations demonstrate trust, honesty, and knowledgeability – aspects outside of profitability that are also deemed important for a successful business identity. Thus, CSR and sustainability reporting has been claimed as a method of legitimizing a business (Cho and Patten, 2007). Whatever the possible rationales may be, it is clear that sustainability reporting is seen as an increasingly desirable business practice (Andersen, 2003; Douglas, 2007; Higgins et al., 2018).
With the ever-evolving nature of CSR practices and the increasing ubiquity of sustainability reporting, some global organizations (e.g. the Integrated Reporting Initiative) have attempted to provide frameworks for businesses to adopt for their corporate sustainability disclosures. These frameworks, however, often serve as guidelines. Moreover, these guidelines have been criticized as insufficient and problematic, in the sense of prioritizing aspects of financial over impactful sustainability practices (Flower, 2015; Milne and Gray, 2013).
Researchers have also been critical of the nature and manner of CSR and sustainability reporting practices. The concept of sustainability and how CSR texts frame the seemingly opposing values of economic growth, social equity, and environmental protection have been noted in several studies. Scholars examining CSR reports from this perspective (e.g. Higgins and Coffey, 2016; Higgins and Walker, 2012; Milne et al., 2009; Tregidga et al., 2014; van Bommel et al., 2023) have claimed that a great degree of sustainability reporting attempts to persuade readers that sustainability and economic growth are harmonious. In other words, a business can achieve economic growth and protect the environment without impeding either goal. How these seemingly contradictory objectives can be achieved, however, are often reported in uncertain and ambiguous terms (Tregidga et al., 2014), lack transparency, (Higgins et al., 2020), contain omissions, and lack contextual information (Tang and Higgins, 2022). This seems to go hand in hand with “green-washing”, that is, the use of environmentalist ideologies to legitimatize business practices (Brugman et al., 2024; Martin et al., 2024; Sterbenk et al., 2022).
Cho et al. (2010) have claimed that businesses manipulate the rhetorical and linguistic features of CSR disclosures in order to manage stakeholder impressions. This underscores the importance of how language is utilized in these reports. Although the CSR phenomenon has been extensively studied in business-related academic disciplines, research using linguistic frameworks is still relatively limited (Fuoli, 2018). Within linguistics, studies identifying the features of CSR as a genre have received particular attention (e.g. Catenaccio, 2011; Lin, 2020; Rajandran, 2018; Skulstad, 2008; Yu and Bondi, 2017). Another trajectory of CSR linguistic research is the exploration of the discursive strategies that companies use in CSR reports to promote a positive image of themselves whilst engaging in accountability to the public and their stakeholders (e.g. Bhatia, 2012; Ghadiri et al., 2015; Jaworska and Nanda, 2016; Rajandran, 2018; Zappettini and Unerman, 2016).
The language in CSR reports can influence our attitudes, perceptions and beliefs towards the qualities of the company in question (Nwagbara and Belal, 2019). Thus, the notion of CSR reports as being strategic (Bondi, 2016), self-serving (Fuoli, 2018), and perhaps even a commodifying practice (Ghadiri et al., 2015) where sustainability is only mentioned by businesses to promote a positive image of themselves, their products, services, features, and ethos, are worth further investigation. Linguistic studies of promotionality within CSR reports have focused on various levels of language use. Some scholars have examined the lexico-grammatical and rhetorical aspects of these documents, highlighting how these linguistic features are used to enhance the image of companies. For instance, Bondi (2016) explored depictions of the future in CSR reports through verb form and use (e.g. will and other futurity markers) noting that futurity is often used in a promotional manner, through discussing a corporation’s forthcoming commitments and promises. Moreover, Bondi and Turnbull (2025) examined how concessives (words and phrases that indicate a contrarian thought, e.g. in spite of) are used strategically in CSR reports to downplay a company’s shortcomings while emphasizing their positive aspects, while Fuoli (2018) revealed the common use of grammatical stance markers (linguistic forms which express a speaker’s subjectivity, e.g. strive, intend, understand) in CSR reports, linking these forms to corporate promotionality and trust building. Finally, CSR reports frequently highlight continual progress, improvement, and learning with no clearly defined sustainable end goal. Thus sustainability is often depicted as a metaphorical journey that both promotes a company’s socially responsible image, while at the same time allowing it to avoid fundamental changes to practices and policies (Milne et al., 2006).
“The style and length of CSR reports will probably vary with industry and the sociocultural and political constraints of the country in question” (Bhatia, 2012: 224). Social responsibility reporting is still a relatively Western phenomenon and research on developed countries has outlined a range of different reporting practices (Kolk, 2010). In contrast, research on how CSR reporting operates in developing regions has been under-examined but is believed to be of increasing importance (Carroll, 2021; Lin, 2021). This underscores the need to investigate CSR reporting across different parts of the world, especially the Global South. For instance, due to historical exploitation, Nigerian CSR reporting is viewed by many of the local population as illegitimate and untrustworthy (Nwagbara and Belal, 2019). In contrast, CSR is argued to be positively received in Thailand, which according to Padungsaksawasdi and Treepongkaruna (2023), is due to the country’s Buddhist tradition and values espoused by the king and royalty, all of which pay heed to socially responsible business practices. Turning back to promotionality, Lin’s (2021) study comparing CSR reporting in the UK and China revealed that UK CSR reports contained a larger amount of promotional discourse, arguing that while UK companies might view CSR reports as a means to promote and build their image amongst various stakeholders, Chinese companies may view CSR reporting as a necessity to comply with the law.
Ecotourism, events and festivals
Events and festivals have become an important economic sector within the global tourism industry (Collins and Cooper, 2017). Numerous studies have also discussed the large negative environmental impacts of festivals (e.g. Ashdown, 2010; Brennan et al., 2019; Collins and Cooper, 2017; Mair and Laing, 2012). Perhaps because of this, the sector is also paying increasing attention to their environmental impact (Brennan et al., 2019). While CSR reports from a range of different industries have been examined (see the overview in Sengur, 2021), work on event and festival CSRs is notably absent.
Stamou and Paraskevopoulos (2004) have argued that “ecotourist” discourses contain both tourism discourses, which centers on hedonism and pleasure, and environmental discourses, which involve our relationships with the environment. These discourses overlap and vie for influence over one another during certain discursive events and influence consumer behaviors. Attending Wonderfruit can be seen as an example of ecotourism, whereby many tourists that visit this festival, both from Thailand and abroad, are looking to pursue hedonistic activities. At the same time, Wonderfruit markets itself as more than just a “typical” festival, but one in which Wonderers (the term used for Wonderfruit attendees) can “move mindfully, conscious of the impact on the land, the greater environment and our community” (Wonderfruit, 2023). This is further made explicit by one of the co-founders of the festival, where he claims that (emphasis ours): We wish to create a platform for sustainability, where Wonderfruit is just a part. It’s heading in a direction where there will be a convergence of multiple things, but we’re focused on building measurable impact through creativity. And of course we will have fun along the way!
Notwithstanding the rapidly growing popularity of this festival, it would therefore be interesting to study Wonderfruit, as they are not only an organization situated in the Global South that prominently carries an environmental ethos, but they are also pioneers of publishing CSRs in the festival and tourist industry, where there are large environmental impacts.
Methods
Case studies have been deemed a suitable and reliable way of conducting research on corporate social responsibility reports (Cañizares, 2022). At the time of writing (as of 2025), Wonderfruit has published four CSR reports; for the 2019, 2022, 2023, and 2024 festivals (the festival did not take place between 2020 and 2021 due to the Covid-19 pandemic). The reports were downloaded from Wonderfruit’s website (wonderfruit.co) and their wordcounts are shown in Table 1.
Wonderfruit CSR reports.
There are clear differences in the names of these reports and their lengths, with the 2024 report being five times longer than the 2019 report. The 2019 report contains sections which relate to the report’s structure like “Introduction” as well as those which are clearly about sustainability: “Waste Management” and “Carbon Offset”. On the other hand, there are a large range of intriguingly-named sections in the 2024 report, including “The Latent Power of Sound”, “Championing Emerging Talent” and “Food as Medicine”. Despite these differences, the texts purportedly serve the same function for Wonderfruit, as evidenced in the founder’s letter within the 2023 report, where he states: “we first shared a Sustainability Report in 2019, which evolved into an Impact Report in 2022 and has now culminated in the Wonder Report” (p. 2). Therefore, they can be directly comparable with each other.
Corpus linguistic techniques are useful in exploring business and corporate discourses (Jaworska et al., 2024). Thus, we take a corpus-assisted discourse approach to the analysis (Baker, 2006; Partington et al., 2013), using computer software to identify non-obvious patterns in texts, which are then analyzed through the scrutiny of concordance lines in order to interpret and explain their meanings. Initially, it was decided to identify keywords across the four reports. A keyword is a word which occurs statistically significantly more often than expected in a corpus or text, compared against a second “reference” corpus (Scott, 1997). However, due to the small size of the corpora, only a small number of keywords with reasonably high frequencies were retrieved. For example, using a cut-off score of 3.84 for log likelihood and a minimum frequency of 50, only two keywords (waste, of) were found for the 2019 report, three keywords (we, our, to) for the 2022 report, three keywords (the, as, a) for the 2023 report and one keyword (with) for the 2024 report. Instead, it was decided to carry out automatic annotation on the four texts and derive key part of speech and semantic categories. This process combines similar words into larger categories and can therefore be a good option when working with low frequency data. This approach also increases the likelihood that linguistic patterns will be identified which may not be otherwise noticed by analysts if they had only considered words as separate entities.
The four reports were uploaded to the online corpus analysis tagger Wmatrix (Rayson, 2008) which can automatically carry out part of speech and semantic tagging of texts. The part of speech tagging is based on the CLAWS7 (Constituent Likelihood Automatic Word-tagging System) tagset, which comprises 18 main tags (e.g. A article, N noun, R adverb) which are further subdivided into 137 tags (e.g. NN1 singular common noun, NN2 plural common noun, NP1 proper noun). Some tags only relate to a single word (e.g. EX refers to existential there) while others, like NN1 can cover tens of thousands of words. The tags are assigned automatically, using a combination of rules and probabilities (see Garside, 1987) with reports of accuracy being between 95% and 98% (UCREL, 1996). The semantic tagging requires texts to be initially part-of-speech tagged, which can help to differentiate word meaning. It uses a tagset called USAS (The University Centre for Computer Corpus Research on Language Semantic Analysis System), and like the CLAWS7 tagset, is hierarchical, consisting of 21 higher-order tags (e.g. F Food and Farming) subdivided into over 600 more specific tags (e.g. F1 Food, F2 Drinks and Alcohol, F3 Smoking and non-medical dugs). The system is reported to have an accuracy of around 91% (Rayson et al., 2004). During analysis, we were mindful to check tagging accuracy by examining concordances of tags, making adjustments to frequencies where necessary.
Key part of speech and semantic tags were derived for the each of the four reports by combining the other three reports as a reference corpus. For example, to obtain a reference corpus for the 2019 report, the 2022, 2023, and 2024 reports were combined together. A score of 3.84 for log-likelihood was used as the cut-off point for keyness (which indicates p < 0.05). Additionally, to ensure against making claims about low frequency data, we only considered key tags which occurred with a minimum frequency of 50.
Concordance analyses were then carried out on the key tags for each report, noting the contexts that the tags occurred in while considering the potential effects on the reader of the particular grammatical patterns and semantic meanings in the discourse. During this process, we began to group tags together, noticing that they tended to be used in ways that contributed towards a particular theme or discourse pattern, both within a single report and across all four reports.
Results
Before we analyze the part of speech and semantic tags which were distinctive to a particular report when compared to the other reports, it is useful to consider the baseline – what kinds of words are common to all of the reports? This was achieved by comparing each report against the American English 2006 Corpus (Potts and Baker, 2012), a 1 million word general reference corpus of written American English containing 500 text excerpts of published writing. The key part of speech and semantic tags for each report were noted, and Table 2 shows only the tags that were key for all four reports.
Key part of speech and semantic tags for all four reports when compared to AmE06.
In the first few rows of Table 2 we see a general set of shared part of speech tags referring to singular and plural common nouns, as well as conjunctions and prepositions. It is perhaps notable that we do not see verbs or adjectives as key across all four reports, which may indicate that these part of speech classes are either used differently in the reports, or are not a distinctive feature of any of these reports. After that are a set of semantic tags which indicate the kinds of words and associated topics which we would perhaps expect to see in sustainability reports: references to waste and materials, food and drink, farming and housing, collaboration, and cause and effect. There is one tag which relates to quantification – N5 (quantities: little) but it is noteworthy that other kinds of quantities do not feature as a stable part of the discourse. Table 2 contains few surprises then, apart from the absence of certain types of tags we might have expected to see. Our subsequent analysis, which focusses on tags which are distinctive to each report, will be more revealing.
In total, there were 6 key tags for the 2019 report, 10 for the 2022 report, 9 for the 2023 report, and 12 for the 2024 report. Table 3 shows all key part of speech and semantic tags.
Key tags for each individual report.
The analysis that follows is structured according to the themes and discourse patterns observed within the key tags in these reports, covering Topics, Authorial voice and social actors, Verb processes, Quantification, and Evaluation.
Topics
In the 2019 report, the key semantic tag N5.2+ refers to waste (with 56 out of 57 occurrences of this tag being the word waste), while the key part of speech tag NN1 (singular common noun) also contains words which relate to waste – the most frequent NN1 words are waste, food, management, carbon and cup, and tend to be used in statements which indicate how waste was minimized at the festival, for example, “No wastewater from food waste was discharged at the site and is properly treated”.
In both the 2019 and 2022 reports, the semantic tag O2 (objects generally) is key – indicating that the 2023 and 2024 reports have relatively fewer references to objects. In both 2019 and 2022, the most frequent O2 words relate to packaging – cups, cans, containers, plastic bags, bottles, and these words also tend to be used in contexts related to avoiding waste, for example, “A zero single-use cup policy was implemented”. Therefore, both the 2019 and 2022 reports have a stronger focus on waste management.
In contrast, the 2023 and 2024 reports have a number of key semantic tags that refer to topics that might not immediately bring to mind the concept of sustainability. For example, H1 (architecture) is key in the 2023 report. In this report, H1 refers to the buildings and houses which were used at the festival with descriptions relating to hiring practices around builders, for example, “Seventy people were hired to build the Slow Wonder cabins, with 50% of that workforce coming from Pattaya and the remainder from around the country”. Additionally, descriptions of the accommodation often involved descriptions of sustainable practices, for example, “The cladding was made from mostly recycled materials like pinewood and plywood, stained to enhance aesthetics”. In a small number of cases, there was emphasis on how accommodation was made comfortable for participants at the festival, for example, “Each cabin was further fitted out with air conditioning, supplied by Daikin, and a premium mattress by Omazz made from Talalay”. Words tagged as H1 do relate to sustainability then, although they also indicate a shift away from the earlier focus on waste management.
However, in the 2024 report, it was less easy to find key tags which directly related to sustainability. Consider, for example, the key tag S9 (religion and the supernatural). In the 2024 report, S9 often referred to spiritual self-improvement processes that attendees could take part in, for example, “We turn to rituals as a way to ground ourselves in intention, creating space to reconnect with our minds”.
S5+ (belonging to a group) is another key tag in the 2024 report. Li’s (2022) study highlights how Starbucks utilizes CSR reports to foster a sense of community with their consumers as part of their corporate identity, which seems to also be the case in the 2024 report. S5+ words in the 2024 report tend to relate to connection: community, team, collective, together, groups, and are used in ways which emphasize that everyone involved in the festival is part of a community, for example, “Translating experiences into insights to uncover numbers that reveal the rhythms and habits of the Wonderfruit community”. The 2024 report also describes activities which take place at the festival which facilitate this sense of connection, for example, “In 2024, we continued to expand family-friendly experiences, creating even more spaces for Wonderers of all ages to explore, learn and play together”. Similarly, the key part of speech tag IW (which in 99% of cases refers to the word with) is also suggestive of a process of connection, for example, “deeply connect with the land, with each other, and with the fundamental space of wisdom”.
Finally, another key tag in the 2024 report that indicates a possible discourse shift is X2.2+ (knowledgeable), where 30% of the words within this category were related to the lexeme experience. The prominence of discussing experiences can be directly linked to promotional discourse of the activities of the event itself, for example, “Molam World provides a space for exploration and discovery – a place where bands can share their music and audiences can have a unique musical experience”.
Authorial voice and social actors
The 2019 report suggests a de-emphasis on social actors, with a significantly high use of passive verb forms tagged VVN (e.g. treated, encouraged, reduced). Seventy one out of 85 uses of words tagged as VVN referred to efforts to reduce waste at the festival. However, in 75 out of 85 verbs tagged VVN in this report, the grammatical agent was deleted, for example, “the remaining single-use cups were collected”. This has the effect of foregrounding the process of managing waste as opposed to the people who are carrying out those actions. Moreover, the key tag in the 2019 report A1.1.1 (general actions/making) also further highlights the authorial style adopted in this report. The words in this category tend to be correlated to scientific reporting, in which the agent was often deleted, for example, “The transition process was considered a success”.
Gâţă (2017) reports on how using first-person pronouns in CSR texts have the effect of creating an affectionate relationship with readers. In the 2022 report we begin to see a significantly higher proportion of words tagged with PPIS2 (the plural first person pronoun we) and APPGE (possessive pronouns – 62% of cases are the word our). In 99% of cases, these pronouns indicate authorial voice, for example, “We encourage all vendors to source ingredients locally and from sustainable sources”. Only a couple of cases potentially imply a more general we, which includes both author, reader and possible others, for example, “Since our inception, we have always been exploring how we can live more closely with nature”. This suggests a more personalized tone than in the 2019 report, which also could be interpreted as promotional – the organizers of the festival are more explicitly taking credit for the sustainability practices they are engaging in.
There is a further change with the 2023 report, which diminishes a sense of a collective authorial voice but has higher use of the semantic tag Z3 (Other proper names). The most common Z3 tag is Wonderers, which refers to people involved in the festival (see example below).
The Wonderfruit community, affectionately known as Wonderers, is a diverse collection of creatives, professionals and explorers. Curious about culture, Wonderers have evolved alongside the brand, deepening their connection between mind and nature.
This increasing use of the label “Wonderers” to refer to attendees accentuates the view of festival-goers as being part of a diverse community, although another interpretation could be that this distinct label is also promotionally beneficial to the organization by encouraging behaviors like repeat attendance and brand loyalty along with an increase in merchandise sales.
Verb processes
There is also a shift in verb usage between the four reports. As we have already seen, the 2019 report shows a higher use of verbs tagged VVN (past participle of lexical verb), which typically involves the auxiliary BE followed by a verb describing a physical action like treated, issued, and diverted. The vast majority of VVN verbs in the 2019 report describe actions that have been carried out to reduce or process waste, for example, “single-use trash bags were replaced by baskets”, providing concrete examples of efforts made to manage the environmental impact of the festival. The 2022 report, however, has a higher use of verbs tagged VV0 (base form of verb) and VVI (infinitive form), both which tend to be used in that report to express intentionality, for example, “so we endeavour to look at impact more holistically”, “our goal is to experiment with them to learn more until they become standard practice”. In the 2022 report these verbs tend to relate to cognitive and social processes like learn, explore, educate, encourage, share and allow as opposed to describing physical actions.
In the 2023 and 2024 reports, the verb tag VVG is more common, used on verbs ending in -ing, which denote actions that are implied to be continuous, for example, “Unlocking the mind through ancient practices”, “Designing timeless venues for now”, “nurturing its ecosystems and building a space that reflects our shared values”. Looking more specifically at the 2024 report, we can also observe that many VVG verbs are also related to the key semantic tag A2.1+ (change). Verbs that fall into this semantic category include blending, evolving, and transforming, which clearly highlights this continuous verbal process.
The kinds of actions described in the 2023 and 2024 reports also relate to creative (blending, creating, crafting, designing), social (offering, nurturing, connecting, inspiring, celebrating) and exploratory (discovering, journeying, seeking, roaming) processes. Specifically looking at the 2024 key tags, we see another key semantic tag, S8+ (helping), that seems to also highlight another prominent semantic domain in which many of these verbs seem to be relegated. The verbs in the 2024 report seem to focus on creating a cooperative atmosphere, for example, “supporting those around us, creating opportunities and sharing resources to build a more connected future”.
Moreover, the 2023 report also shows significantly higher use of verbs tagged VVD (past tense forms), with these verbs being similar in nature to the VVG ones: celebrated, created, shared, contemplated, for example, “To better study the meanings of mindfulness and its interrelations, we paused and went inward. What began as a concept evolved into a venue”. The key semantic tag X7+ (wanted) in the 2023 report also highlights intentionality with key words in this tag including verbs like aimed and planning. For example: This year, we will take a larger step in formalizing these practices and processes, aiming to develop Wonderfruit as a space where culture can serve to learn and grow. The project looks to elevate hospitality at Wonderfruit and introduce thoughtful cultural and wellness programming in the years ahead.
Such cases of X7+ words tend to be situated within plans which are implied to have positive outcomes but are perhaps difficult to quantify and provide evidence for, for example, “learn and grow”, “elevate hospitality”, and are also suggestive of a gradual shift at Wonderfruit towards promotional discourse.
Overall, there is a sense then, that over the four reports, verb usage shifted from reporting evidence of concrete action relating to decreasing waste towards describing intended or ongoing actions that were less physically tangible, involving creative, mental or social processes and change. The verb features in the most recent CSR reports provide linguistic evidence to Carroll’s (2021) findings, where they discuss how businesses tend to reference sustainability within their CSR reports as a work in progress. This is also in line with what Milne et al. (2006) claim is often seen in business discourses – a conceptual metaphor framing sustainability as a journey without a clear finish line.
Quantification
Another shift in discourse style involves use of quantification. In 2019, the tag IO is key (referring to the preposition of). Of the 80 occurrences of this tag, 32 of them (40%) occurred within statements which specified quantities, for example, “100% of liquid waste was captured”, “591kg of food containers being composted on site”, “The total amount of money collected was 200,650.65 Thai Baht”. In particular, use of percentages or proportional data in these kinds of statements (e.g. x% of y), help to provide readers with a specific understanding of an achievement.
In the 2022 report, the semantic tag N1 (numbers) is key, also indicating that quantification is seen as important, although in this report, the numbers tend to be used less in terms of providing proportional information, for example, “This year we bought 3000 t- CO2e credits”, “we planted 20,000 trees”. Tang and Higgins (2022) discuss how data in CSR reports are often not easily understandable because they are reported without clear benchmarks. Whether or not, for example, the amount of carbon credits and trees planted is meaningful may be hard for the average person to interpret. Moving to the 2023 and 2024 reports, there are no key tags relating to quantification, which suggests that providing evidence in terms of numbers or proportions has become less important by this point.
Evaluation
The 2019 report had no key part of speech or semantic tags which directly related to evaluation. Instead, as we have seen, the key tags tended to relate to descriptions of activities, alongside quantification or comparisons to previous years, which had been undertaken to prevent waste at the festival, for example, “Recycling rates have also increased from 2% in 2018 to 48% for 2019”. There is no explicit evaluative language in this example, perhaps because readers are expected to understand that what is being reported is good, without needing to be told.
Contrasting to this, in the 2024 report, there is a higher use of general adjectives (tagged JJ), which all indicate positive evaluation, for example, new, diverse, unique, creative, authentic, holistic, fresh, timeless, sustainable. These adjectives tend to be paired with nouns which relate to a range of aspects of Wonderfruit, helping to create a sense of promotional discourse, for example, “This year’s focus was on sustainable and creative crafts, with activities that explored natural dyes, plant-based materials, and upcycled plastics, inspiring new ways to transform waste into art”. Therefore, the 2024 report is more explicit in terms of positive evaluative language about the festival itself, although this language is not always about actual sustainable practices, e.g. “Through unique menus, potent elixirs and distinctive dining traditions, we offer a taste of perpetual and timeless practices”.
Discussion and conclusion
To interpret these findings, we draw on the corporate reporting guidelines provided by the Integrated Reporting Initiative. In these set of global guidelines, businesses are encouraged to report on six metaphorical “capitals” in their CSR reports: financial, manufactured, intellectual, human, social and relationship, and natural capital (IIRC, 2024). Our study indicates that Wonderfruit’s CSR primarily reports on two forms of capital: social and relationship capital, and natural capital. Social and relationship capital is described as relationships within and between communities, stakeholders or networks, including shared norms, common values, behaviors, key stakeholder relationships, trust, brand, reputation, and the organization’s social license to operate, while natural capital are all renewable and nonrenewable environmental resources and processes that provide goods or services that support the past, current or future prosperity of an organization (IIRC, 2024).
As discussed earlier in this paper, businesses have been observed to mask their sustainability disclosures, omit information, and decontextualize the information that they do provide (Higgins et al., 2020; Tang and Higgins, 2022). This seems to echo with our findings about Wonderfruit’s manner of reporting about natural capital, which is surprising, in part because a large part of Wonderfruit’s branding is that of environmental sustainability. The 2019 report is more concerned with reporting concrete evidence used to manage waste at the festival, with more emphasis on actions taken than on those who carried them out, along with attempts to quantify waste management activities. The 2022 report still references waste management, although there is increased emphasis on intent as opposed to what has been done. The 2023 report has much less reference to waste management; actions are often described in this report as not only ongoing but also as intentions, making it difficult to make claims about what has been achieved – resultingly, there is much less quantification in this report. Although the 2023 report presents a positive account of the festival, it also could be viewed as providing a less tangible account of sustainability than the other reports. Finally, there are increased verbal processes of ongoing, and therefore incomplete, change in the 2024 report, linking sustainability practices to a metaphorical journey.
We can also observe the expanding emphasis on social and relationship capital, alongside increased promotionality over time. The 2022 report, although still referring to waste management, makes use of a more openly promotional style of discourse through the shift of authorial voice to more open use of pronouns to make evident that the organizers are the ones performing these sustainable actions. The 2023 report focused on the festival’s attractions, especially the architectural and accommodation aspects. In the 2024 iteration, the report shifts to the artistic, cultural, spiritual, and musical experiences of the festival, utilizing words like with and without and lexical verbs with the -ing participle to create a cooperative community of stakeholders and attendees, alongside a prominent uptake in evaluative language through dense adjective usage. Key semantic domains like S8+ (helping) and S5+ (belonging to a group) in the 2024 report also clearly indicate a thematic shift toward emphasizing social and relationship capital.
These results therefore indicate a clear change in the discourse used across the four reports. Wonderfruit, while initially focusing on the informative and sustainable (i.e. about how the festival is attempting to eliminate or mitigate their negative impacts on the environment), has become increasingly promotional over time (i.e. a greater focus on the activities and community experiences of the festival). This coincides with the manner and the extent in which natural and social and relationship capitals are discussed within these reports.
Wonderfruit’s conception of “hedonistic sustainability” as their ethos, and their business as a festival organizer, may shed light into the possible reasons for this shift. Hedonistic sustainability can be neatly mapped onto how the discourse of these CSR reports are informational or promotional more broadly, along with how the concept fits within the continuum of environmental and touristic discourses more specifically. Firstly, as a touristic experience provider, profitability for Wonderfruit is inevitably tied to bringing in an increasing number of attendees. To increase attendee rates, the festival must promote itself. As an environmentally-conscious festival, this promotion, we believe, can also appear in their CSR reports through how they disclose and discuss their environmental practices and the festival’s attractions. Although Wonderfruit’s CSR reports initially paid greater focus to their environmental practices (i.e. natural capital), the discourse around natural capital has become increasingly obfuscated. Meanwhile, we observe a shift to social and relational capital, indicating that the discourses tied with this form of capital appear to be increasingly important for the festival, perhaps from a financial perspective. Although Wonderfruit has claimed that their festival is hedonistically sustainable, the data here suggests that the hedonistic aspect of this festival is prioritized, possibly at the expense of the sustainable. Underlying Wonderfruit’s ethos of hedonistic sustainability, we can therefore observe the inherent tensions between business profitability and environmentalism. To what extent can profitability, which in the case of festivals involves touristic experiences and the consumption behaviors around it, and environmentalism both be achieved without compromising the other? The discursive shifts within these CSR reports indicate that achieving this balance may not be easy.
To conclude, these findings suggest that Wonderfruit’s CSR reports seem to be moving away from the environmental ethos that they were purportedly found on, as the sustainability disclosures seem to be less foregrounded with each CSR report, while promotional elements are increasing. From a critical perspective, if one were to treat corporate communication in all forms as strategic, at worst this shift might indicate possible greenwashing. In other words, sustainability, environmentalism, and other green discourses are commodified and framed as an experience in which tourists to Wonderfruit can take part in as one of the many features of the festival. Nevertheless, Wonderfruit still remains one of the very few festivals (if not the only one) which publishes annual CSR reports. As more and more events are paying heed to the environmental impact of their operations, more transparent CSR reporting from a wider range of event organizers is needed.
Future directions
The methodology adopted in this study provides a robust approach for scholars interested in critical discourse analysis, where datasets are often small, to still effectively identify salient aspects of their data using corpus linguistic methods. While key tags have been used before in corpus-assisted discourse research (e.g. Baker, 2006), our study considers both key semantic and part of speech tags in concert. This enables the linguistic features of smaller texts to be grouped together in a way in which the text’s topics and themes, and potentially discourses, which may not be easily observable, can be seen more clearly. Future projects can also utilize this same approach to examine the discursive features of smaller corpora.
Our study identified how discourses are realized through repetitive use of language choices, including pronouns, verbs, adjectives, nouns and numerals, which are linked to discursive strategies like quantification, evaluation, and social actor deletion/inclusion. These features could be used to analyze a wider range of CSR texts, both from festivals and other businesses. Another aspect of the analysis could involve consideration of images within the texts. Baker and Collins (2023) have shown how tags can be automatically assigned to images and incorporated into a corpus-assisted analysis of discourse. Such work can show how meaning is achieved through combinations of words and images. It is interesting to note, for example, that the 2019 Wonderfruit CSR report contains no images at all, whereas reports after this contain images on almost every page, many of which show idealized, highly colorful pictures of local indigenous people, working within nature (e.g. farming or planting), which contrast with images of groups of young festival-goers, often seated or engaged in creative or leisure activities. The inclusion of these kinds of images contributes towards the shift towards promotional discourse, making the later reports resemble aspects of holiday brochures as opposed to mere accounts of responsibility. Additionally, it would be interesting to consider reception of CSR reports, analyzing the language used to report or evaluate them in news and social media.
Footnotes
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.