Between Nationalism and Globalism: Framing Climate Change Risks in Chinese and American Newspapers

1 Introduction

In an increasingly globalizing world, human beings are confronted with growing risks and potential harms, especially with more challenges from technological disasters, political and economic unrest as well as infectious diseases (Beck 1999). Risk has thus become a grave concern in the present-day world (Boholm 2016) and people increasingly use it to understand certain social problems (Battistelli and Galantino 2019). However, risk is related to not only scientific calculation of the likelihood of an unfavorable event (Haimes 2009); but more importantly, to the communication and perception of uncertain features and potential harms (Battistelli and Galantino 2019). It is thus discursively constructed (Ushchyna 2018). Particular ways of communicating risks can influence people’s understanding of political events and social issues. Therefore, news media increasingly report risks to conceptualize certain social issues, such as some controversial health and environmental issues (e.g., Borah et al. 2024; Chan 2016).

Climate change is one of the most pervasive environmental risks to human society in the 21st century (Liu and Huang 2022). To meet the challenge, immediate global attention and coordination are urgently needed (Han, Sun, and Lu 2017). Nevertheless, in recent years, international climate negotiations have been accompanied by controversies and gridlock (Painter et al. 2023). This can be explained by the divergent attitudes in different countries towards the risks of climate change due to their different national interests. As two largest carbon emitters in the world, China and the US are expected to take more responsibilities for addressing climate risks; especially at recent climate conferences, China and the US have become the focus of climate negotiations (Pan, Opgenhaffen, and Van Gorp 2019). However, the US, as a leading developed country, has been shying away from taking the responsibilities for coping with climate change and hesitant to join the international efforts to curb climate change (Merry and Mattingly 2023). By contrast, as the largest developing and a big rising country, China has taken a proactive stance towards climate change in recent years. It views climate change as a good opportunity to expand its global influence in international affairs and improve its international image (Chen and Liu 2024), through actively participating in international climate talks and facilitating global collaboration in carbon emission reductions (Fu and Wang 2022).

Since news media play a paramount role in shaping public understanding of a global issue of which they have no prior knowledge and beyond their immediate concern (Gillings and Dayrell 2023), it is thus of great interest and significance to examine how the news media from the two superpowers, China and the US, compete in the international arena to construct the issue and legitimize their national climate actions. Although there has been growing interest in examining the discursive constructions of climate change in the two countries (e.g., Pan, Opgenhaffen, and Van Gorp 2019; Song et al. 2021), most of these studies focus on their preferential ways of framing the issue as regards the (un)certainty of climate change (i.e., whether climate change is certain or uncertain) (e.g., Bailey, Giangola, and Boykoff 2014) as well as the attribution of responsibilities (i.e., who are responsible for climate change) (e.g., Liang et al. 2014). Still, news framing of climate risks or potential impacts of climate change in the two countries needs more exploration. As emphasized by Liu and Lin (2021), risk framing has become an important discursive strategy in covering social issues because it can be used by news media to evoke public fears and offer the rationale for political actions. This indicates that news framing of risks is not only related to evidence-based scientific knowledge but also associated with politics and national interests. Therefore, this study examines the risk framing of climate change in two leading English-language newspapers in China and the US, namely China Daily (CD) and The New York Times (NYT). It tries to answer three research questions: (1) What are the preferential ways of framing climate risks in the two newspapers? (2) How do they reveal their different ideologies towards climate change? and (3) What are the socio-political factors behind their particular ways of risk framing?

2 Media Framing of Climate Risks

News media are essential in disseminating climate knowledge to the public and raising their awareness to minimize climate risks (Gillings and Dayrell 2023). However, climate knowledge is not simply related to “truths” or “facts”. It is transmitted by the media in specific socio-political contexts (Gillings and Dayrell 2023). Influenced by different ideological viewpoints and political interests, news reports can frame climate risks in a particular way, i.e., to “promote a particular problem definition, causal interpretation, moral evaluation, and/or treatment recommendations” of the risks through selecting certain aspects of reality and making them more salient (Entman 1993, 52). It thus has the potential to shape people’s perception of climate risks and legitimize policies to address climate change (Vu, Liu, and Tran 2019). In recent decades, media representations of the potential impacts of climate change have attracted growing research interest (e.g., Painter 2013; Painter et al. 2023). For instance, Painter (2013) makes a distinction between the implicit/disaster and the explicit risk frames. He proposes that the meaning of risks can involve “a broad sense of a possible adverse impact” and “a more technical sense of assigning probabilities or confidence levels to different outcomes” (Painter 2013, viii). The implicit risk frame refers to the possible adverse impacts of climate change, which is often associated with the disaster frame. In this fear-based disaster frame, climate change is perceived as uncontrollable and unprecedented, which is usually not accompanied by positive messages. It is thus effective to attract media attention but ineffective to improve public engagement. The explicit risk frame refers to the chance that something negative will happen, which is associated with the use of the word “risk” and the related concepts in the same semantic field. This frame appears to be more effective to communicate climate change than strong messages of disaster. However, it is found that the implicit/disaster risk frame is still predominantly used by journalists to highlight the serious outcomes of the changing climate (Merry and Mattingly 2023; Painter 2013).

The choice of different risk frames can communicate different ideologies and suggest different political interests. This can be witnessed in the analysis of the UK and US media. For example, Sonnett (2022) finds that political preferences can possibly influence the way of framing climate risks in the US media. Left-leaning newspapers prefer to link climate change to the word risk, which emphasizes the decision and choice. By contrast, the right-leaning media tend to associate global warming with the word danger, which foregrounds the passive role of humans in managing the risk. In the UK context, left-wing newspapers are found to emphasize the risks of global warming and create a sense of danger, whereas right-wing newspapers tend to question the dangers of global warming (Carvalho 2007). Nevertheless, these studies mainly investigate newspapers in a particular country. Few studies explore the framing of climate risks across countries and illuminate the influence of the national interests (exceptions including Grundmann and Krishnamurthy 2010; Schmidt, Ivanova, and Schäfer 2013; Shehata and Hopmann 2012). Furthermore, previous studies mostly rely on manual content analysis (e.g., Dirikx and Gelders 2010; Painter et al. 2023) rather than corpus-assisted discourse analysis (Liu and Li 2017). While manual content analysis contributes to identifying particular ways of risk framing at the macro-level of texts, it usually fails to address the discursive strategies and linguistic features that contribute to risk framing at the meso- and micro-levels of texts. Besides, the manual content analysis of a large number of texts is also time-consuming and labor-intensive, and the accuracy and replicability of analysis are often questioned. In order to remedy the above problems, this study proposes a corpus-assisted discourse study (CADS) approach to the analysis of the particular ways of framing climate risks in news media from different socio-political contexts.

3 Corpus-Assisted Discourse Studies

This study intends to combine the theories of framing (Entman 1993), critical discourse analysis (CDA) (Wodak and Meyer 2016), and corpus linguistics (CL) (Baker 2006), and gives a CADS of media framing of climate risks in both CD and NYT. CDA regards discourse as a social practice and takes a particular interest in uncovering the hidden power and ideologies behind the discourse (Fairclough 1992; Wodak and Meyer 2016). An ideology refers to “a worldview, a system composed of related mental representations, convictions, opinions, attitudes, values and evaluations” shared by certain social groups (Reisigl and Wodak 2016, 25). Since news media serve as the main site of ideological struggles (Fairclough 1995), CDA has been widely adopted to unmask the ideological standpoints in the newspapers. However, the traditional qualitative analysis in CDA is often criticized for cherry-picking data or linguistic features for analysis (Baker and Levon 2015). Therefore, a growing number of studies have adopted CL methods in CDA (Alexander 2010; Gillings and Dayrell 2023), which are known for the potential to efficiently process a large amount of data, identify linguistic patterns with corpus analytic tools, and provide the basis for further detailed examination (Baker 2006; Islentyeva 2020).

CADS combines both quantitative and qualitative analyses to provide a deeper understanding of the discourse. The quantitative analysis relies on corpus analytic tools to examine a corpus in its entirety (Gillings, Mautner, and Baker 2023). These tools help to identify linguistic patterns and statistical trends which cannot be detected by manual analysis and reduce subjective involvement in the examination. While corpus analytic tools are effective at revealing recurring patterns, qualitative interpretation is necessary to explore the meaning behind these patterns. In CADS, the qualitative analysis can help explain and interpret the patterns and strategies shown by corpus analytic techniques from the critical perspective and expose the underlying ideologies (Baker et al. 2008; Partington, Duguid, and Taylor 2013). In this sense, CADS underlines a synergy of the methods related to CL and theoretical underpinnings related to CDA. According to Entman (1993, 52), framing “essentially involves selection and salience” and can be “manifested by the presence and absence of certain keywords”. There has been growing literature using framing theory to examine media constructions of climate risks. However, previous studies mostly explore the presence of frames in the discourse and rely on the quantitative content analysis (e.g., Painter 2013; Vu, Liu, and Tran 2019). The detailed linguistic means of framing climate risks as well as the more nuanced ideological meanings within the dominant framing, by contrast, remain under-researched (Hart 2023). Given that climate risks are discursively constructed in nature, a CADS approach to framing analysis can combine the quantitative analysis of a large amount of data with the qualitative analysis of some discursive patterns and strategies which are key to news framing and make the analytic procedure more replicable (Alexander 2010; Liu and Li 2017; Wang and Ma 2021).

4 Data and Methodology

Two corpora have been built in this study, namely the CD corpus and the NYT corpus. The CD corpus is composed of news reports related to climate change in China Daily (CD). The NYT corpus consists of news reports concerning climate change in The New York Times (NYT). The two newspapers were selected because they are important in expressing their respective countries’ national perspectives and safeguarding their national interests on the global stage (Lee et al. 2002). NYT is considered as “newspaper of record” in the US (Liu and Huang 2022, 4), and is known for its agenda-setting influences on international news media (Song et al. 2021). It is often tracked by politicians for decision-making because it provides a window into public opinion (Song et al. 2021). As the largest and the most influential English-language newspaper in the Chinese mainland, CD serves as the party organ and shoulders the responsibilities for communicating the voice of the Chinese government to the external world (Liu 2023).

The data were retrieved from the online database LexisNexis and all the news reports containing climate change in the headlines from 2001 to 2020 were collected. The period from 2001 to 2020 captures significant developments in climate discourse, including the US’ withdrawal from the Kyoto Protocol and the Paris Agreement, and the increasing global visibility of climate change impacts. To make sure that all the collected data were relevant to the topic, the downloaded reports were further manually checked. In general, 438 news reports were collected in the CD corpus, with 291,958 tokens. 1071 newspapers were collected in the NYT corpus, with 1,106,660 tokens. It shows that CD pays less attention to presenting climate change, which can be explained by the fact that the Chinese government did not highlight climate issues in the national agenda until 2007 (Broadbent et al. 2016).

This study starts with an examination of the keywords related to danger and risk in CD and NYT by an online corpus analytic tool, Wmatrix 5.0. This software enables an automatic semantic analysis and is distinctive because the UCREL Semantic Analysis System (USAS) it incorporates can categorize the English categories into 21 semantic fields, which can be further grouped into 232 semantic categories (SMCs) (Rayson 2008). Wmatrix 5.0 can generate a key SMC list by comparing the SMCs of a purpose-built corpus with those of a specialized or general reference corpus. The key SMCs are ranked in terms of their log-likelihood (LL) values. The higher their LL values, the more statistically significant those SMCs. Two key SMCs lists are generated by comparing the two corpora respectively with a general reference corpus, the 5-million-sample corpus of the British National Corpus (BNC). This study first focuses on the top 15 key SMCs in each corpus to examine the prominent topics/themes in each corpus. Then this study focuses on the SMC “Danger” (A15-) and investigates the actual use of the top 10 frequent tokens in the two newspapers to examine the prominence of the explicit and implicit/disaster risk frames. It is followed by a close analysis of the most frequent tokens (i.e., risk[s]) in A15-. To further examine their particular ways of risk framing in the two corpora, this study focuses on three aspects of risk framing: (1) the (un)certainty of climate risks; (2) the causes and consequences of climate risks; and (3) the solutions/actions to climate risks (Entman 1993; Liu and Li 2017). Different discursive strategies are further examined to reveal how they can serve ideological purposes.

5 Findings

5.1 Analysis of the Key SMCs

Table 1 demonstrates the top 15 key SMCs in two corpora to identify the most prominent features and major trends in the two corpora. The shared 11 key SMCs indicate that the two newspapers have many similarities in representing climate change. An investigation of the most frequent tokens in each SMC suggests that the news coverage is closely related to four topics, including (1) environment; (2) causes and impacts; (3) science; and (4) politics. The most prominent theme is “environment”, including weather events: W4 (e.g., climate, weather, floods, droughts and storms), climate change: A2.1+ (e.g., change) and O4.6 (e.g., temperatures), as well as environment: W5 (e.g., environmental and environment), W1 (e.g., planet) and W3 (e.g., earth). Besides, causes and consequences of climate change are also highlighted, as can be seen from O1.3 (e.g., gas and carbon dioxide) and F4 (e.g., agricultural, agriculture and crops). Furthermore, both newspapers highlight the important roles of scientific communities (e.g., scientists, science, and scientific in Y1) and government (e.g., government and nations in G1.1) in climate change representations.

Table 1:

Top 15 key SMCs in the two corpora.

Rank	CD			NYT
	Tagset	LL	Semantic Category	Tagset	LL	Semantic Category
1	W4	11,600.08	Weather	W4	15,058.69	Weather
2	A2.1+	6,582.65	Change	A2.1+	5,027.9	Change
3	W5	2,673.52	Green issues	W5	3,268.13	Green issues
4	M7	1,740.8	Places	Y1	3,174.18	Science and technology in general
5	W3	1,582.52	Geographical terms	W3	2,777.2	Geographical terms
6	Y1	1,537.86	Science and technology in general	Z99	2,705.93	Unmatched
7	Z2	1,513.79	Geographical names	O4.6+	2,641.42	Temperature: Hot/on fire
8	O1.3	1,508.16	Substances and materials: Gas	O1.3	2,571.9	Substances and materials: Gas
9	G1.1	1,319.41	Government	Q2.1	2,563.52	Speech: Communicative
10	F4	1,310.69	Farming & Horticulture	G1.1	2,078.43	Government
11	W1	1,207.02	The Universe	W1	1,470.15	The Universe
12	X5.2+	1,129.36	Interested/excited/energetic	O4.6	1,011.38	Temperature
13	Q2.1	1,066.53	Speech: Communicative	A13	865.92	Degree
14	S1.1.3+	966.58	Participating	F4	860.84	Farming & Horticulture
15	O4.6	749.67	Temperature	A15-	756.8	Danger

Nevertheless, different key SMCs suggest their preferential ways of framing the issue. CD prefers to frame climate change as an international issue and focuses on international cooperation in addressing the issue. This can be seen from CD’s preference for M7 (e.g., countries and developing countries), Z2 (China and United States), and S1.1.3+ (e.g., conference, and meeting). This is consistent with the finding of previous studies that Chinese official newspapers prefer to frame climate change from the perspective of global collective efforts (Song et al. 2021; Wu 2009). However, they cannot be found in the top 15 key SMCs of NYT, which can be attributed to NYT’s preference for framing climate change as a domestic issue (Grundmann and Krishnamurthy 2010).

Besides, NYT puts more emphasis on the dangers, as can be seen from the prominence of A15- (e.g., risk, risks, and danger) in NYT. This tendency is also revealed by other key SMCs (LL>6.63) in NYT, such as “dead” (L1-), “worry” (E6-), “fear/shock” (E5-), and “damaging and destroying” (A1.1.2).

To further examine how climate risks are framed in the two newspapers, this study focuses on the most frequent tokens in A15-. The top 10 frequent tokens in A15- in CD and NYT are as follows.

CD: risk (123), risks (68), danger (17), dangerous (15), dangers (10), exposure (10), hazards (8), peril (4), perils (3), jeopardy (2)

NYT: risk (555), risks (414), dangerous (142), dangers (84), danger (75), endangered (66), exposure (32), hazards (31), risky (30), hazard (22)

These tokens can be semantically grouped into tokens expressing the meaning of risks and dangers (Boholm 2012). It suggests that the two newspapers prefer to use both explicit and implicit/disaster risk frames (Painter 2013). Tokens concerning risks are risk and risks in CD (191, 76.4 %) and risk, risks and risky in NYT (999, 70.4 %). Tokens concerning dangers are danger, dangerous, dangers, hazards, peril, perils and jeopardy in CD (59, 23.6 %) and dangerous, dangers, danger, endangered, hazards and hazard in NYT (420, 29.6 %). The percentages of the explicit and implicit/disaster risk frames suggest that while the explicit risk frame is more favored in both corpora than the implicit/disaster frame, NYT shows more preference for the implicit/disaster risk frame than CD. Figures 1 and 2 show the diachronic distribution of the implicit and explicit risk frames. It shows that while the explicit risk frame rises in both corpora, the implicit/disaster frame rises in CD but decreases in NYT. The growing use of the implicit/disaster frame in CD in recent years reflects its emphasis on the consequences, underscoring the urgency of addressing the issue. Conversely, NYT’s increasing reliance on the explicit/risk frame suggests its tendency to highlight uncertainty surrounding climate impacts. However, despite its decline, the implicit/disaster frame remains relatively prominent in NYT. This suggests a continued reliance on fear-based language to frame climate change, which can draw more media attention but also risk creating despair and leading to public apathy (Painter 2013).

Figure 1:

Standardized frequencies of the tokens in each year in CD.

Figure 2:

Standardized frequencies of the tokens in each year in NYT.

6 Discussion and Conclusions

To sum up, this study gives a corpus-assisted discourse study of the framing of climate risks in CD and NYT and reveals both differences and similarities in their framing of climate risks. Both corpora prefer to use the explicit risk frame but NYT shows more tendency to use the implicit/disaster risk frame. It thus indicates its potential to create fear and despair. Although the explicit risk frame is viewed as a more effective way to communicate climate change (Painter 2013), preferential ways of framing risks in the two newspapers indicate their different attitudes towards climate change. CD prefers to frame climate change as a global issue and tends to underline the certainty of climate risks and the impacts on human beings and the world. Therefore, it pays attention to the collective efforts to manage the risks. NYT tends to frame climate change as a national issue and underline the uncertainty of climate risks. It prefers to highlight the impacts of climate change on business and the economy and shift the US government’s responsibilities for addressing the issue.

Their preferential ways of framing climate risks can be explained in terms of the different functions of the two newspapers and the national interests of the two countries concerning climate change (Han, Sun, and Lu 2017; Song et al. 2021; Wu 2009). As an official English-language newspaper, CD serves as the mouthpiece of the Chinese government and shoulders the responsibility for communicating the voices of the Chinese government (Liu and Li 2017). With the release of the first national assessment of global warming report in 2006, the government built the political consensus on climate change and largely accepted the potential consequences of climate change, such as water shortage and food insecurity, on China’s further development (He 2014; Wu 2009). In the last two decades, the Chinese government has played an increasingly active role in addressing climate change (Stensdal 2014). Therefore, CD prefers to “adhere closely to the Chinese government’s guidelines to project one uniform voice in positioning China as the leader in the global battle against climate change” (Song et al. 2021, 18). The preference for a global frame and the emphasis on the certainty of climate risks underline the urgency and necessity of global efforts to address climate change and the positive role China plays in climate change issues. However, the theme of “international responsibility” may also indicate CD’s tendency to place responsibilities in the hands of external agents, especially in developed countries (Liang et al. 2014; Xie 2015), since China tends to underline the differential responsibilities between developed and developing countries (Liu 2023).

Previous studies have demonstrated that countries with higher GDP per capita are inclined to frame climate change from the perspective of domestic politics rather than international relations (Song et al. 2021; Vu, Liu, and Tran 2019). This argument is also supported by NYT which frames climate risks as a contested issue and thus affects the public’s and policymakers’ perceptions of climate change. Besides, the journalism norm of the Anglo-American media system underlines the practice of balanced reporting (Hallin and Mancini 2004), so American newspapers tend to include different viewpoints on climate change (Boykoff and Boykoff 2004). This also contributes to representing the uncertainty of climate risks in NYT, even though some studies have revealed that there is a decrease in skeptical voices in US news representations of climate change (Xie 2015). Furthermore, previous studies have demonstrated that the US news media tend to align with national interests in the coverage of international issues (Lee et al. 2002). This study supports the argument and finds that NYT follows the neoliberalism ideology in framing climate risks that emphasizes economic growth and free market logic and reduces government solutions. The frame of climate change as a domestic issue and the emphasis on the uncertainty of climate risks help to downplay the severity of climate change and shift the global responsibilities of the US for tackling climate change. It can thus help legitimize the US’ failure to adopt an effective climate plan (Fremstad and Paul 2022). For example, although the Clinton and Obama administrations were actively involved in climate negotiations, the US gradually concluded its leadership in international climate politics and abdicated the global responsibilities in climate change actions after President Trump announced the withdrawal from the Paris Agreement in 2017 (Kopra 2019).

This study has several limitations. First, it focuses on two leading newspapers in China and the US. Future research could examine other party-oriented or market-oriented Chinese newspapers and US newspapers with varying partisan orientations. Second, while this study analyzes climate change coverage, future studies could explore news framing of related environmental issues, such as global warming and heat waves. Third, this study investigates news reports from 2001 to 2020. Future studies could include more recent data to uncover new trends in news framing, especially during and after the COVID-19 pandemic. It is expected that future studies can benefit from a corpus-assisted discourse study as it can reveal not only the preferential ways of framing but, more importantly, the hidden attitudes and ideological positions of different newspapers (Song et al. 2021). The use of CL methods can contribute to the identification of frames in a more scientific and replicable way, while the close analysis of some linguistic tokens in their specific contexts can help to reveal the detailed linguistic strategies used by different newspapers for the construction of news frames (Baker 2006). Therefore, a combination of the methods and theories of CL, CDA and framing can expose the news framing of risks at different levels. More importantly, it underlines the socio-political factors behind the particular ways of risk framing and argues for the examination of risk framing in its socio-political contexts (Ushchyna 2018).