A systematic literature review on ecological discourse analysis (2014–2023)

2.3.1 Identification

The identification begins with the literature search. The key search term and its variants or synonym phrases were used in searching. In this research, the key search term was < ecological discourse analysis > and it was developed based on research questions and previous academic articles.

As illustrated in Table 1, this study enhanced several key terms by searching relevant publications in the databases of WoS and Scopus. They are both subscription-based platforms offering access to extensive databases and comprehensive citation information across a broad spectrum of disciplines. Numerous researchers chose them because all journals in these two databases undergo a stringent editing process, which ensures the article quality as well as the impact factor rankings of publications (Abas 2023). Since the number of articles in the databases has changed at various times, there is an acknowledgement that the data in this review was extracted from WoS and Scopus on January 15, 2024.

After searching these databases, a comprehensive examination of all 948 articles was conducted to eliminate the duplicated records by using Microsoft Excel and manual checks. After this stage, a final count of 920 articles was left. Subsequently, the remaining 920 papers underwent the next procedure, screening.

2.3.2 Screening

The second stage is screening. According to PRISMA 2020, there are three steps in the screening procedure:

First of all, records are screened. In this step, the authors individually check article titles and the corresponding abstracts of the searched 920 papers. After comparing each person’s selection and discussion where necessary, 830 articles were excluded. There are two reasons to exclude these studies: (1) 746 articles are irrelevant to this research by looking at their titles and abstracts, and they focused on psychology, economics, sociology, education, communication and ecology not ecological discourse in linguistics field; (2) non-research articles, including review articles, book chapters and conference articles, are not included, and the number is 112. Therefore, 62 articles were through for the retrieval step.

The second step is to make reports sought for retrieval. The selected 62 articles in the above step are used for retrieval. The third step is eligibility. After browsing the whole article, 2 articles are not involved with linguistics, so they are excluded. Finally, 60 are included in the review.

2.3.3 Included data and quality appraisal

The selected 60 articles were reported as the data for this SLR. These articles with source information were collected in a repository.^[1]

To ensure the accuracy of the data, two experts who are familiar with both ecolinguistics and SLR were invited. They checked the articles selected from WoS and Scopus twice, including the processes of identification and screening.

3.1.1 The spatial and temporal distribution of selected articles

After the systematic selection process, the qualified data comprise a total of 60 articles, covering EDA-related studies from 25 countries worldwide (as demonstrated in Figure 2). According to the result analyzed by VOSviewer, the country which possessed the largest number of articles was China with 21, and it is followed by the United Kingdom (6 articles) and Canada (5 articles). The number of articles that contribute to EDA was the same in Australia, Italy, and the United States, with 3 ones. In addition, Austria, Germany, India, Indonesia, Malaysia, Netherlands, and Ukraine had 2 articles respectively.

Figure 2:

The spatial distribution and coauthor relationship (a).

Concerning the coauthor relationship, China also had the most cooperative authors with other countries in EDA, and their cooperative researchers included those from Canada, the United Kingdom, Spain, and Malaysia, as shown in Figure 2. UK researchers had the second highest number of collaborations with other countries, which were Austria, China, and the Netherlands, as shown. Other countries where international collaborative research has emerged include Germany and the Netherlands, as well as Germany and Norway. In this case, 9 countries out of 25 demonstrated co-authorship, occupying over a third.

With reference to the temporal distribution (as demonstrated in Figure 3), the 60 articles included were distributed starting from January 2014 to December 2023. Overall, the number of EDA demonstrated an upward trend. In the first five years (between 2014 and 2018), there were fewer studies since the number of articles were below 5 on average. In 2019, the number of published articles increased to 7. Afterwards, the results demonstrated an increased interest in EDA. The number of research increased dramatically during the latter half of the decade, reaching to 14 articles in 2021 and 18 ones in 2023 although there was a drop in 2020.

Figure 3:

The temporal distribution of selected articles.

3.1.2 Source journals

As demonstrated in Figure 4, language-related journals have published more on EDA. Journal of World Languages took the lead, with 5 articles (taking 8 %), while Text & Talk and Theory and Practice in Language Studies published 4 articles (7 %) respectively from 2014 to 2023. They are followed by journals of interdisciplinary research (with 3 articles each), including Environmental Communication – A Journal of Nature and Culture and Journal of Language Teaching and Research. In addition, Journals of Critical Discourse Studies, Frontiers in Psychology, Language Sciences, and Linguistics and the Human Sciences take an equal percentage of 3 % separately.

Figure 4:

The source journals publishing articles on EDA.

There is a category of journals named others, which refers to those that have published less than 2 articles related to EDA. For example, some journals, such as Chinese Journal of Applied Linguistics, Children Literature in Education, Language and Literature, Pragmatics and Society, and Social Semiotics, also include papers on EDA, but only one article appeared in one of them in the past decade, so these journals were classified as others.

3.1.3 The overlay visualization of EDA

VOSviewer was aided in analyzing 2the general trend of research in the selected journal articles. Since the number of articles was not large, the frequency of co-occurrence of author keywords was selected as twice and above. Figure 5 exhibited an overlay visualization of the selected articles in the systematic review, and it is depicted based on average published years. The yellow shades represent the latest research. Concerning the size of the circle, the more related studies are associated with it, the larger the size and vice versa.

Figure 5:

The overlay visualization of keywords in the selected articles.

From 2014 to 2018, scholars primarily concentrated on topics such as environmental sociology, the environment, and the research method of content analysis. Between 2019 and 2022, the research volume increased, and more studies on ecological discourse were around the harmonious discourse analysis and language ecology of ecolinguistics, as well as the critical discourse analysis on news discourse and sustainability. In recent years, EDA on the topic of climate change has gained most widely attentions from analysts (e.g. Hase et al. 2021; Koop-Monteiro et al. 2023; Plastina 2022; Stoddart et al. 2016), particularly regarding its metaphors and framing. In addition, there has been growing interest in the ideology and ecosophy of discourse (e.g. Vaishali and Rukmini 2021; Zhang and He 2021) and children’s literature (e.g. Khan and Zeb 2023; White et al. 2021) although studies on these aspects were still limited.

3.2.1 Research themes

As demonstrated in Figure 6, a total of 14 themes related to EDA have been published in Journals of WoS and Scopus, namely, theory/research approach, news coverage, literature, feedback, language on social media, textbook, political document, report, website advertising, debate, film discourse, leadership commentary, as well as others. Others refer to themes that appear less than twice. In general, the research topics covered a diverse range, and the results of the distribution were relatively average.

Figure 6:

Research themes in the selected articles.

To be specific, both themes of “theory/research approach and news coverage” and “news coverage” took the largest number (with 9 articles). Considering the theory/research approach, it included proposing the research method on EDA (e.g. Huang and Zhao 2021; Stibbe 2014), constructing the analytical framework (Wei 2021), and exploring the linguistic philosophy (ecosophy) (Vaishali and Rukmini 2021). Regarding news coverage, it covered the discourses of climate change (Hase et al. 2021; Stoddart et al. 2016), sustainability (Litofcenko et al. 2023; Yacoumis 2018), ecological events (Mesikämmen et al. 2020; Xiong and Wang 2023), political news report (Zhang et al. 2023), news report of Covid-19 (Xue and Xu 2021), and the selected ecological news (Yuan 2023).

The third frequent studies were on literature (with 6 articles), such as children’s books (Khan and Zeb 2023; White et al. 2021), poetic discourses (Ononye and Chiluwa 2023; Zuo 2019), Portuguese literature (Lopes-Fernandes et al. 2016), and Croatian Renaissance literature (Galić Kakkonen and Mihaljević 2023). The ecological discourse research on feedback also had 6 articles, including multi-faith scholars and leaders who are environmentally active (Boettcher 2021), interviews with forest owners (Takala et al. 2019), and questionnaires of learners’ feedback (Roccia and Iubini-Hampton 2021). The next categories were concerned with the language in social media discourse (e.g. Bhattacherjee and Sinha 2024; Koop-Monteiro et al. 2023) and textbook (e.g. Sharma and Buxton 2015; Song and Perry 2023), with each of them having 4 articles. Ecological discourse studies focusing on political documents, reports, and websites were comparatively less, each comprising 3 articles. Additionally, discourses within advertising, debates, film, and leadership commentary were also considered as research corpora, each represented by 2 articles.

3.2.2 Research approaches

Table 2 presents the research approaches employed in EDA over the past decade. The methods exhibited diversity, predominantly leaning towards qualitative research. Notably, 13 selected articles employed multiple analytical frameworks/methods. However, for clarity in statistical representation, these instances were not categorized as multiple analyses but were counted individually. Consequently, the total sum in the percentage column of Table 2 exceeded 100 %.

Most of the articles adopted the critical discourse analysis although a variety of analytical approaches were used, as illustrated in Table 2. Specifically, critical discourse analysis appeared 19 times (e.g. Stibbe 2014; Stradling and Hobbs 2023; Takala et al. 2019; White et al. 2021; Yacoumis 2018), and this number took 31.67 %. It was followed by theoretical exposition (e.g. Huang and Zhao 2021; Law and Matthiessen 2023; Zou 2021) with 9 articles (16.67 %), and corpus discourse analysis (Frayne 2022; Istianah and Suhandano 2022; Poole 2016; Stradling and Hobbs 2023) with 9 articles (15.00 %).

The frequency of 13.33 % was taken by eco-critical discourse analysis, which was utilized by researchers like Zhang et al. (2023) and Xiong (2014), while content analysis had the percentage of 8.33 %, and it was mentioned by scholars like Stoddart et al. (2016), Takala et al. (2019), and Hase et al. (2021).

Compared to previous research, content analysis, multimodal discourse analysis, Foucauldian discourse analysis, comparative analysis, metaphor analysis and multimodal critical discourse analysis were less frequently used, with each of them taking up 8.33 %, 6.67 %, 5.00 %, 3.33 %, 3.33 %, and 3.33 % respectively.

Research approaches of others refer to those which appeared only once, including: building semantic association network model, cognitive analysis, diachronic discourse analysis, dictionary-based analysis, ecocriticism, ecolinguistic discourse analysis, ecostylistic analysis, frame discourse analyses, harmonious discourse analysis, multilevel analysis, panel analysis, path analysis, positive discourse analysis, sociology-of-knowledge discourse analysis, topic modelling, as well as visual and discourse network analysis. These expressions of research methods were mentioned by the authors in their articles.

4.2.1 Research themes

Based on the results in Section 3.2.1, one of the most frequent themes of EDA is the “theory/analytical approach”. This indicates that linguistic scholars have persisted in exploring ecological discourse studies via multiple dimensions in the past decade. They strive to put forward approaches and diversity in the research of ecolinguistics to deal with ecological discourse, such as proposing ecological discourse analysis (Stibbe 2014) and harmonious discourse analysis (Huang and Zhao 2021). Additionally, Zou (2021) and Zhang et al. (2021) add the dimension of people’s cognition into ecological discourse study, which is achieved by melting the conscious operational opposite-unity cognition procedure (Zou 2021) and landsenses ecology and Maslow’s hierarchy of needs (Zhang et al. 2021) into linguistics approach. They proposed new approaches to accentuate the multidimensional nature of EDA. Since EDA highlights interpreting the relationship between language and its impacts ecologically, it is necessary to consider other disciplines. As Cowley (2018) indicates, by collaborating ecolinguistics with ecological psychology and other disciplines, we can enhance people’s consciousness regarding the interconnectedness between humans and bio-ecology.

Moreover, some analysts probed into the ecosophy or philosophical framework in language analysis. The examination of ecosophy was highlighted as the latest trend in EDA, as depicted in Figure 5, the overlay visualization of keywords in selected articles. Vaishali and Rukmini (2021) explored the linguistic philosophy of Tholkappiyam, which is an ancient linguistic treatise, delving into the connection among culture, ecology, and language through the Tinai theory. Despite the predominantly Western roots of ecolinguistics, the linguistic philosophy of Tholkappiyam has the same potential as ecolinguistics study (Vaishali and Rukmini 2021), so they assess the Tinai framework to highlight its capability in ecolinguistics studies. Zhang and He (2021) proposed a philosophical framework for ecolinguistics by adopting both the shared criteria in Chinese schools of philosophy and the Western human geographical concept of a sense of place. They want to create a solid foundation for analyzing and evaluating linguistic resources, contributing to the essential strand of ecolinguistics research. Just as Huang and Zhao (2021) state, to assess the ecological value of language, we should look beyond Western philosophies and explore the interplay between Eastern and Western thoughts. Researchers are encouraged to identify or establish beneficial ecosophy for evaluating discursive ideologies (Stibbe 2021).

In addition, the wide focus on “news coverage” indicates scholars’ attention to the ecological impact of this discourse genre. Conducting ecological analysis of news discourses can unveil how they construct social issues as well as influence the recognition of individuals or policy makers ecologically. According to Mesikämmen et al. (2020) and Van Dijk (2000), news media shape social reality by wielding discursive power, influencing issue agendas and public discourse. Meanwhile, in today’s interconnected media-society landscape, deciphering how news frames key social issues is crucial for researchers and policymakers tackling these challenges (Yacoumis 2018).

When conducting a more in-depth analysis of the latest trend (in Section 3.1.3) and themes (in Section 3.2.1), we found that a relatively small number of studies were concerned with analysis of ecological discourse, while a majority of them were the ecological analysis of discourse. There were 12 of them delving into discourses with environmental-related topics (ecological discourse). Certain academics delved into the ecological discourse in literature, including the nature poetry of Dickinson (Zuo 2019), the representations of wolves in Portuguese literature (Lopes-Fernandes et al. 2016), the environmental devastation constructed in children’s literature (Khan and Zeb 2023), the Croatian Renaissance text of Petar Zoranic’s Planine (Galić Kakkonen and Mihaljević 2023), the Niger Delta poetry (Ononye and Chiluwa 2023), the ecostylistic analysis on two lectures delivered by John Ruskin about The Storm-Cloud of the Nineteenth Century (Virdis 2022). Apart from EDA in literature, Flusberg and Thibodeau (2023) delved into the metaphors in environmental discourse, and Ponton (2022) examined the websites promoting a nature reserve. Meanwhile, scholars (e.g. Hase et al. 2021; Koop-Monteiro et al. 2023; Plastina 2022; Stoddart et al. 2016) also investigate discourses with climate change topics. The discussions on the ecological discourse can contribute to understanding human relations with the natural world (Galić Kakkonen and Mihaljević 2023; Lopes-Fernandes et al. 2016) and provide ways to environmental protection and ecological restoration (Ponton 2022). Realizing the meaning of ecological discourse, language can be critically used to create a more sustainable society (Khan and Zeb 2023).

Apart from the aforementioned research, most themes are related to the ecological analysis of discourse, emphasizing how discourses reflect certain strategies or political views ecologically, influence readers’ ecological actions, and whether the analyzed discourses convey positive or negative ecological ideologies. This indicates the main focus of EDA lies in interpreting the ecological influence of language. As Penz and Fill (2022) state, EDA is not just restricted to analyzing the environmental discourse. The ecological analysis of discourses would be the main focus of ecolinguistics (Alexander and Stibbe 2014). These discourses may cover a wide range of registers, such as political discourse, news coverage, discourse on social media, and talk shows (Law and Matthiessen 2023).

4.2.2 Research approaches

As demonstrated in Section 3.2.2, critical discourse analysis (CDA) was the most popular approach. When conducting ecological discourse, the critical approach can be used to unveil the ecological ideology within discourses. CDA has been utilized to investigate environmental discourses from the late 1990s onward, aiming to uncover the power relationships and ideologies in society (Penz and Fill 2022). When examining ecological discourse, CDA can be used to reveal the hidden ideology in discourses (Stibbe 2014).

In addition, some researchers (e.g. Huang and Zhao 2021; Xiong and Wang 2023; Zhang et al. 2023) employed eco-critical discourse analysis in their articles. This research approach is proposed by Arran Stibbe, the founder of the International Ecolinguistics Association (Nash 2018), who pioneered the integration of ecosophy into CDA. We can utilize eco-critical discourse analysis to discover the hidden positive or negative ecological ideology within discourses (Stibbe 2014). This approach allows scholars to assess whether the ideology embedded in discourse is ecologically beneficial or not. By analyzing the ideology or values within discourse under a certain ecosophy, we can judge whether the language is ecologically beneficial or destructive (Stibbe 2021).

Additionally, corpus discourse analysis was frequently employed, incorporating both corpus-based and corpus-driven methods upon further examination. Researchers (e.g. Frayne 2022; Gilquin 2022; Xiong and Wang 2023) adopted a corpus-based research method to conduct EDA. However, different from the above studies, Roccia and Iubini-Hampton (2021) built an assessment model based on readers’ feedback to measure the ecological influence induced by an ecolinguistics course online. Although they did not mention the term “corpus approach”, their research was assisted with a corpus-driven approach. In this sense, corpus discourse analysis has been employed in investigating ecological discourse from different perspectives. The approach of corpus discourse analysis can offer an objective demonstration of the linguistic features of the selected discourses. According to Alexander (2018), research based on corpus can reveal that specific linguistic features carry significant weight. In this context, corpus research approaches are deliberately employed to avoid confirmation bias, as emphasized by Frayne (2022). This ensures a more robust and unbiased exploration of linguistic features.

Regarding the multimodal approach of EDA, 4 articles utilized multimodal discourse analysis, and 2 articles referred to multimodal critical discourse analysis. They are mainly concerned with textbooks (Song and Perry 2023; Triyono et al. 2023), websites (Chen and Gunster 2016; Ponton 2022), advertising (Liu and Liu 2020), and animation (Ivchenko 2021). Considering the genres in their research, they are typical multimodal discourses, with predominant verbal and visual modes to construct semantic meaning. When a discourse involves different modes, they should be examined as a whole. This can avoid discrepancies between the meaning conveyed by monomodal discourse and multimodal discourse. The ecological ideology also needs the contribution of multimodal resources in discourse. As Song and Perry (2023) elucidate, the semantic meaning and its reflected ideology by multimodal resources may be different from that of the single text or image.