Abstract
This study conducts a corpus-assisted discourse study of framing responsibilities for climate change in China Daily (CD) and The New York Times (NYT). Based on the distinction between causal and treatment responsibilities, it focuses on the framing of human and non-human causal responsibilities as well as developed and developing countries’ causal and treatment responsibilities for climate change in the two newspapers. The findings suggest that CD tends to show consensus on the human causes of climate change while NYT is inclined to problematize human causes for climate change. While both newspapers favor treatment over causal responsibilities, CD prefers to underline developed countries’ historical causal responsibilities for climate change and urges developed countries to take more treatment responsibilities for climate change, whereas NYT prefers to underscore developing countries’ current causal responsibilities for climate change and their shared treatment responsibilities for climate change.
Keywords
Introduction
Responsibility attributions have drawn growing attention in academia due to their important role in shaping public understandings and opinions towards a certain issue (Iyengar, 1996). Previous studies have made a distinction between two categories of responsibilities, namely “causal” and “treatment” responsibilities (Iyengar, 1991). Causal responsibility refers to the causes of a problem, whereas treatment responsibility highlights the alleviation/solution of a problem. The former underlines the past, while the latter points to future solutions (Iyengar, 1990). These concepts are especially important in controversial issues because they are concerned with the parties responsible for the causes and solutions of the issues (Kim et al., 2010) and public understanding of the issues (Schlesinger and Lau, 2000). Numerous studies have been conducted on the discursive constructions of responsibilities in media discourse (e.g., Holton et al., 2012).
Previous studies argue that media are central sites of blame games (Anderson et al., 2018). Social actors are likely to blame others so that they can protect themselves against criticism. This can be witnessed in media’s preference for attributing social issues to individual deficiencies rather than governmental policies (e.g., Koteyko et al., 2008; Peng and Tang, 2010), because it can help to shift the blame on the government (see Kim et al., 2017). Particular ways of framing responsibilities in news media are subject to political orientations and professional routines (Kim et al., 2010). This has been evidenced in cross-national studies related to conflicts and international concerns (e.g., Chen and Wang, 2020; Kim and Lee, 2008; Li, 2010; Liu and Li, 2017).
Despite the controversial nature of climate change (Liu and Huang, 2022; Villar and Krosnick, 2011), media representations of climate change responsibilities in different countries have been extensively examined, such as the US (Kuha, 2009; Liang et al., 2014), the UK (Nerlich et al., 2012) and India (Billett, 2010). Nonetheless, few studies have given a critical examination of the particular ways of framing climate change responsibilities by news media from the US and China, two main contributors to climate change (Boykoff, 2012). Therefore, this study combines the methods and theories of corpus linguistics (CL) and critical discourse analysis (CDA) to give a corpus-assisted discourse study (CADS) of the framing of causal and treatment responsibilities for climate change in two international newspapers in China and the US, namely China Daily (CD) and The New York Times (NYT). It has two primary objectives: (1) to reveal their preferential ways of framing responsibilities for climate change in the two newspapers; and (2) to examine whether the two newspapers align with their national interests in framing climate change responsibilities.
Framing responsibilities for climate change
News media can frame an issue in a particular way (Gitlin, 1983). The framing process involves selecting certain aspects of reality and making them more salient (Entman, 1993). Therefore, it has the potential to shape people’s interpretations of a certain issue by defining what the problem is, what/who causes it and how to deal with it (Entman, 1993). News framing of responsibilities for climate change can potentially lead the audience to identify the causes of and solutions to climate problems (Billett, 2010; Olausson, 2009). For example, acknowledgment of human impacts on climate change will contribute to people’s active involvement in addressing climate change (Boykoff and Boykoff, 2004). Recent studies have investigated media framing of responsibilities and found that media coverage may pay less attention to causal responsibilities than treatment responsibilities (e.g., Freeman, 2017). The discussion of causes is mainly concerned with debates on human interference in climate change. For instance, the US newspapers are largely found to be skeptical about human contributions to climate change and are likely to attribute causes to natural fluctuations (Kuha, 2009). This can potentially deflect the responsibility of the government and lead to ineffective actions. The doubtful voice may be attributed to the need for balanced reporting (Boykoff and Boykoff, 2004) or Republicans’ denial of human-caused climate change (Carvalho, 2007). By contrast, media in other countries like the UK are found to preferably present consensus on human impacts on climate change (see Freeman, 2017; Grundmann and Krishnamurthy, 2010).
Some other studies also focus on the debates about the causal and treatment responsibilities of both developed and developing countries. In most cases, news reports follow the in-group-protecting mechanism and tend to put less blame on in-group members (Post et al., 2019). For example, Indian media seldom attribute causal responsibilities to other developing countries (Billett, 2010). However, some studies also discover that media in developed countries prefer to blame their own countries or other developed countries for contributing to climate change (Liang et al., 2014; Post et al., 2019). This can be attributed to the global consensus on developed countries’ historical responsibilities for climate change (Post et al., 2019). As regards treatment responsibilities, studies have found that if the countries are framed as the main contributors to climate change, they are inclined to have obligations to solve the problem (Billett, 2010; Liang et al., 2014; Pan et al., 2021). Billett (2010) shows that a negative image of developed countries is overwhelmingly presented in Indian mass media through an emphasis on their lack of actions in global efforts. In some cases, media are also found to present a positive national image by attributing treatment responsibilities to their own countries. For instance, Liang et al. (2014) argue that the US TV news stories tend to frame the US as having more capabilities to solve climate change. Post et al. (2019) discover that media in developing countries are likely to depict their own countries as actively involved in the global issue partly because they tend to avoid attributing too much power to developed countries (see also Pan et al., 2021). These studies have demonstrated that the framing of climate change responsibilities tends to be shaped by national interests. As Lee et al. (2002) suggest, media representations of international news events tend to be filtered by the prism of national interests. However, there is little information about whether news media from China and the US frame climate change responsibilities in a way that is consistent with their national interests.
Methodology
The present study is based on two large corpora: the CD corpus and the NYT corpus. The CD corpus collects news reports concerning climate change from the newspaper CD. The NYT corpus collects news reports concerning climate change from the newspaper The NYT. CD is selected because it is the most important official English-language newspaper of China and serves the important function of communicating the voice of the Chinese government. As a “newspaper of record” in the US, NYT is known for its liberal stance and influences on shaping international news agendas (Lee et al., 2002; Liu et al., 2022). The two corpora were built by collecting news reports with climate change in their headlines. All news texts were extracted from the electronic database LexisNexis and the time span set for data collection is from 2001–2020. Previous studies have shown that climate change started to receive much media attention after 2000 and has received growing attention in the media since the mid-2000 (Schmidt et al., 2013). To make sure that all news texts collected are topic-related, these news reports were further manually checked. The CD corpus consists of 438 news reports with 281,958 tokens, and the NYT corpus consists of 1071 news reports with 1,106,660 tokens. The number of news reports suggests that climate change receives less attention in CD than in NYT. This can be attributed to the lack of attention to climate change by the Chinese government before 2007 (Heggelund and Nadin, 2017).
A corpus-assisted discourse study approach (CADS) is adopted in this study by combining the theories and methods of corpus linguistics (CL) and the discourse-historical approach (DHA) in CDA (Reisigl and Wodak, 2016). CDA views discourse as a social practice and underlines the examination of discourse in its socio-political contexts (Fairclough, 1995). One of its primary concerns is to expose the dynamic relations between language use and the wider socio-political contexts (Catalano and Linda, 2020; Chen and Waugh, 2020; Wodak and Meyer, 2016). Viewing climate change as not only a social phenomenon but also a discursive phenomenon, this study considers news discourse as an important site for discursive struggle (Fairclough, 1995) and examines the particular ways of framing responsibilities for climate change in different news media. It is “critical” in that it aims to “make contradictions apparent” and demystify the dynamic relations between news discourse and society (Wodak and Ludwig, 1999: 12).
The DHA is known for its emphasis on the importance of socio-historical context in the explication of the findings of textual analysis as well as the three-dimensional analysis of texts at the macro (e.g., topics/themes/frames), meso (e.g., discursive strategies) and micro (i.e., linguistic means and linguistic realizations) levels (Kijratanakoson, 2022; Reisigl and Wodak, 2016). With this framework, the study starts with the macro analysis of the particular ways of framing responsibilities and then moves to analyze the specific discursive strategies and linguistic realizations in the constructions of causal and treatment responsibilities of different parties. Discursive strategies refer to “a more or less intentional plan of practice” to “achieve a particular social, political, psychological or linguistics goal” (Reisigl and Wodak, 2016: 33) and this study pays particular attention to the use of four prominent discursive strategies: (1)
CL features the use of computer-assisted corpus analytic tools for the automatic identification and analysis of language patterns (Cheng, 2013). It can benefit CDA by providing an automatic analysis of large data efficiently, identifying the language patterns which cannot be acquired through mere manual analysis and providing entry points for the close analysis of some language patterns in their specific contexts of use (Baker, 2006; Liu and Zhong, 2020). CDA can benefit CL by providing proper interpretations and explanations of the findings generated by computer-assisted corpus analytic tools. Therefore, CADS underlines the “balanced” combination or “synergy” of the methods and theories of (C)DA and CL (Baker et al., 2008; Partington et al., 2004). It makes no distinction between “corpus-based” and “corpus-driven” approaches (Tognini-Bonelli, 2001). An analyst may approach a corpus with or without prior assumptions, but the analyst has to move constantly between the findings generated by computer-assisted corpus analytic tools and the examination of some linguistic features in their specific contexts of use to identify the most proper and ingenious route in analyzing the data and answering the concerned research questions. In this sense, it is “data-driven” rather than “corpus-based” or “corpus-driven” (Partington, 2010).
Following DHA, this study starts from an overall analysis of responsibility framing by identifying whether those words related to causal and treatment responsibilities stand out in the two corpora in view of the fact that framing can be “manifested by the presence and absence of certain keywords” (Entman, 1993: 52). It starts from the automatic semantic processing of two corpora with the online corpus-analytic tool Wmatrix 4.0. The UCREL Semantic Analysis System (USAS) Wmatrix 4.0 incorporates can categorize the vocabulary of English into 21 semantic fields, which can be further classified into 232 semantic categories (SMCs) (Rayson, 2008). It can also help to generate a key SMC list by comparing the SMCs of one corpus with the SMCs of a reference corpus. Key SMCs refer to those SMCs which show statistically significant differences in a subject corpus when compared with their use in a comparable specialized corpus or a general reference corpus. Key SMC lists can rank the key SMCs by their log-likelihood (LL) values. The higher their LL values, the more key these SMCs. To generate their respective key SMC lists, the two corpora are compared in turn with a general reference corpus incorporated in Wmatrix 4.0, i.e., the American English 2006 Corpus (AmE06) (Liu, 2017; Liu and Ma, 2021). The top 30 key SMCs of each corpus are examined to see whether they have key SMCs related to responsibility framing in the two corpora. It is followed by a close examination of the key SMC “Cause and Effect and Connection” (A2.2) in both corpora to examine how they contribute to the constructions of causal responsibilities of climate change in two newspapers. The most frequently used tokens in this SMC (i.e., cause*) are closely examined to reveal the discursive strategies of constructing causal responsibilities in the two newspapers.
It is complemented by a close examination of the discursive construction of the causal and treatment responsibilities of two key parties: developed and developing countries (Reisigl and Wodak, 2016). This is achieved through a close examination of the token countries and their concordance lines to identify the
Therefore, this study is mutually informed by the methods and theories of CDA and CL. CDA, DHA in particular, provides not only the theoretical framework for this study but also the proper interpretations and explanations of the findings generated by corpus-analytic tools, whereas CL can contribute to the efficient processing of the data, the identification of the most prominent language patterns for further detailed linguistic analysis (Baker et al., 2008).
Findings
Analysis of key SMCs
Top 30 key SMCs in the two corpora.
A close examination of the tokens in each shared key SMC finds that they foreground three topics: (1) environmental problems; (2) ways of tackling the problems; and (3) their causes and consequences. The first topic can be revealed in such key SMCs as
The top 20 tokens in A2.2 in CD and NYT.
This study focuses on the different forms of cause* in CD and NYT to see whether they are used to express the causal responsibilities for climate change. They are caused (88) and cause (54) in CD, and cause (301), caused (240), causes (139), and causing (126) in NYT. An examination of their concordance lines finds that cause* can be used to express either the causes or the impacts of climate change in both corpora (see Examples 1 and 2). (1) (CD, 2019/11/22) (2) (CD, 2009/04/01)
Overall, 29% (41) of cause* (142) in CD and 36% of cause* (805) in NYT are used to express the causes of climate change. It suggests that cause* is less likely to be used for the causes than for the impacts of climate change in both corpora. Nevertheless, cause* is more likely to be used in NYT than in CD to express the causes of climate change. Besides, a distinction can be made between human and non-human causes of climate change, as in the following: (3) (CD, 2014/09/20) (4) (CD, 2017/11/17)
The frequencies of human and non-human causes of climate change.
This can also be revealed in the (5) The Heartland Institute is a conservative think tank that disputes the established science of human-
(NYT, 2018/05/08)
The occurrence of human-caused immediately before climate change makes human-caused climate change a nominal group that can be discussed and debated (Fairclough, 1995). Although CD also underlines climate change as caused by human beings, it tends to express it in a clause, as in the following: (6) Nowadays, the view that human activities have (CD, 2010/01/28)
Therefore, CD tends to align with the statement that climate change is caused by humans. It represents a unified voice over this issue. By contrast, NYT tends to foreground the dispute over human-caused climate change in the US, as in the following: (7) It is widely believed that most Republicans are skeptical about human-
(NYT, 2018/07/28)
To sum up, cause* is used less frequently to emphasize causes over impacts but more frequently to emphasize human- over non-human causes in both newspapers. Nevertheless, while CD shows consensus on human causes for climate change, NYT suggests the dispute over human-caused climate change. This is consistent with the increasing polarization of political parties over human-caused climate change in the US (Boykoff, 2007). Especially after Donald Trump became the US president, he and the Republican party showed growing distrust of human-caused climate change. The following sections give a close analysis of the token countries to examine the framing of developed and developing countries’ responsibilities for climate change.
Analysis of the collocates of the token countries
Collocates of countries at L1 position.
Causal and treatment responsibilities of developed and developing countries in two corpora.
Causal Responsibility
CD highlights the causal responsibilities of developed countries and downplays the causal responsibilities of developing countries. An examination of the concordance lines (56) of causal responsibilities for developed countries finds that CD tends to resort to different strategies in the construction of the causal responsibilities of developed and developing countries. First, CD prefers to refer to the (8) Developed (CD, 2011/11/23)
Besides, CD also uses (9) The almost unrestricted emission of greenhouse gases by developed (CD, 2007/11/16)
By contrast, CD also downplays developing countries’ contribution to climate change by resorting to the (10) The United Nations Framework Convention on Climate Change (UNFCCC) 1992 pointed out clearly that developed countries have contributed the biggest proportion of greenhouse gas emission historically and at present, the per capita emission of developing (CD, 2007/10/31)
In the meantime, they also refer to the (11) This is because the globalization process has shifted most of the manufacturing from developed (CD, 2009/12/07)
Causal responsibilities of developing countries are rarely mentioned in CD (6), and (12) Increased demand has led to high oil prices and developing (CD, 2007/10/31)
By contrast, NYT not only (13) Developing (NYT, 2014/09/21) (14) Developing
(NYT, 2015/12/07)
Overall, CD gives a negative representation of developed countries by constructing them as the main contributors to climate change and underlining their historical contributions to the current climate crisis. Meanwhile, it gives a positive representation of developing countries by constructing them as victims of the problems caused by developed countries. By contrast, NYT not only downplays causal responsibilities by giving an equal emphasis on developed and developing countries but also underlines the current contributions of developing countries to climate change.
Treatment Responsibility
A close examination of the concordance lines of developed (388) countries in CD finds that two themes emerge: (1) developed countries must take the lead in cutting emissions and help developing countries (200, 76.6%); and (2) developed countries have not fulfilled their due treatment responsibilities (61, 23.4%). The first theme takes the dominant share (200, 76.6%) in CD’s representations of treatment responsibilities, as in the following: (15) The developing countries are insisting that the industrialized (CD, 2009/06/15)
Among them, deontic modal verbs should (82) and must (4) are often used to underline developed countries’ treatment responsibilities. Besides, developed countries often occur as the “Target” of some verbs used to express the verbal process, as can be seen from the expressions urged (14) or called on (14) (Halliday, 1994: 141). They contribute to the construction of treatment responsibilities of developed countries. However, CD also gives a negative representation of developed countries by highlighting their failure or reluctance to fulfil their due treatment responsibilities (61, 23.4%). Developed countries are negatively (16) It is a pity that developed (CD, 2007/10/31) (17) However, the developed (CD, 2011/03/03) (18) A recent review of the INDCs by civil society groups shows that the ambition of all large developed (CD, 2015/11/27) (19) However, in these 10 years, developed (CD, 2009/08/12)
(20) On top of the fact that rich countries are responsible for today’s global warming, these countries are even reluctant to give the funds and technical support that developing (CD, 2007/05/24) (21) Developed (CD, 2014/12/05)
An examination of 228 concordance lines of developing in CD finds that four themes emerge: (1) developed countries’ responsibilities to help developing countries (100, 44%); (2) developing countries’ treatment responsibilities for climate change (69, 30%); (3) developing countries’ differentiated responsibilities for climate change (41, 18%); and (4) developed countries’ too high expectations of developing countries’ treatment responsibilities for climate change (18, 8%). In other words, they underline the important role of developing countries to fulfil proper treatment responsibilities on the one hand, and highlight developed countries’ contributions to developing countries on the other hand. (22) (CD, 2008/04/24) (23) (CD, 2007/10/31) (24) (CD, 2010/10/08) (25) (CD, 2011/11/23)
An examination of the 44 concordance lines that address the treatment responsibilities of developed countries in NYT finds that although some of them acknowledge developing countries’ request for developed countries to fulfil their commitments (11) and developed countries’ failure to fulfil their responsibilities (5), the majority of them (28) still give a positive representation of developed countries by underlining their agreement to help developing countries to cut climate change, as in the following: (26) Developed (NYT, 2016/12/14)
Besides, it also justifies the US government’s failure to fulfil its responsibilities by attributing it to the unfair commitments made by developing countries or other developed countries, as in the following: (27) Until now, negotiations had followed a divide put in place by the 1997 Kyoto Protocol, which required that developed (NYT, 2014/12/14)
On the contrary, developing countries are generally represented as passive in solving the climate issue. Their inactive attitude is highlighted by using (28) But in the years after the protocol was announced, developing (NYT, 2005/12/04)
Discussion and conclusion
To sum up, a corpus-assisted discourse study of the framing of responsibilities in CD and NYT finds that they show explicit differences in their representations of causal and treatment responsibilities for climate change. An overall analysis of the two corpora finds that while human causes of climate change have been addressed in both corpora, it is more emphasized in NYT than in CD. Besides, although CD shows consensus on human causes of climate change, NYT tends to highlight the dispute over human causes of climate change. A further analysis of the token countries finds that both corpora favor the distinction between developed and developing countries in framing responsibilities for climate change, but it is more apparent in CD than in NYT. Furthermore, both corpora prefer to highlight treatment responsibilities over causal responsibilities. CD prefers to foreground historical causal responsibilities of developed countries for climate change and urges developed countries to take the lead in fulfilling their treatment responsibilities for cutting emissions and helping developing countries deal with climate change. It is consistent with the emphasis of the Chinese government on “differentiated responsibilities” for climate change (Pan et al., 2021). NYT prefers to highlight current causal responsibilities of developing countries for climate change and argues for the shared treatment responsibilities of developing and developed countries for climate change. This is consistent with the US government’s wish to shirk treatment responsibilities for climate change and shift them to developing countries (e.g., Liang et al., 2014; Post et al., 2019).
Therefore, media representations of international issues tend to be filtered by prisms of national interests (Lee et al., 2002). Be it CD or NYT, the ideological square of “positive self-presentation and negative other-presentation” (Van Dijk, 1998) is still at work in the framing of developed and developing countries’ causal and treatment responsibilities for climate change, even though NYT is known for its liberal stance. Both newspapers align with their national interests in framing the causal and treatment responsibilities for climate change. This can be revealed by not only their choice of topics/themes but also their choice of some discursive strategies and the linguistic means and realizations. Even though the responsibilities for climate change have been widely recognized and acknowledged in scientific and academic circles, their representations in news media are still full of discrepancies and controversies. This further suggests that climate problems are not only a scientific issue but also a discursive issue, and that news media always serve as an important site of discursive struggles (Boykoff, 2007; Liang et al., 2014; Post et al., 2019).
The particular ways of framing responsibilities should merit our further attention and necessitate a critical examination given news media’s significant role in shaping public understanding of certain global issues. The main contributions of this research lie in not only revealing their different ways of framing responsibilities at different levels of discourse but also the CADS approach adopted. A combination of the methods and theories in CDA and CL can present not only a holistic picture of the representations of the issue over a long period but also the detailed discursive and linguistic strategies used for manipulating public perceptions of a contentious issue. It is hoped that it can contribute to the growing literature towards this endeavor.
Footnotes
Acknowledgements
We thank the anonymous reviewers for their insightful comments.
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article. The writing of the paper was funded by the Hong Kong Polytechnic University (UGC) (Project ID: 1-BE68)
