Metaphors and China’s COVID-19 policies: a diachronic analysis of Chinese-English bilingual news editorials

2.1 Metaphors in the COVID-19 context

As previous studies have shown, WAR metaphors have remained the dominant way of representing the pandemic, especially during its early stages (see Charteris-Black 2021; Degani 2023; Hanne 2022; Sabucedo et al. 2020; Wicke and Bolognesi 2020). However, the use of WAR metaphors is quite contentious. On the one hand, they may have counterproductive effects, such as framing COVID patients as “fighters”, which may lead to feelings of guilt if they do not recover (Olza et al. 2021; Semino 2021). On the other hand, WAR metaphors can be beneficial by preparing the public for difficult times, encouraging behavioral changes, and fostering national unity and resilience (Seixas 2021). Charteris-Black (2021: 35) contends that WAR metaphors are not inherently wrong or unethical, but the political motives behind their use need to be examined. For example, Benzi and Novarese (2022) suggest that governments might exploit WAR metaphors to foster obedience and thereby undermine democracy. Musolff (2022) also holds that WAR metaphors may undermine public trust in government communication and potentially fuel conspiracy theories. Besides WAR metaphors, JOURNEY, SPORTS, and NATURAL FORCES/DISASTERS metaphors are also employed, among others (Olza et al. 2021).

Specifically in the Chinese context, WAR metaphors also remain prevalent, but other metaphors, such as RACE and CHALLENGE (Gui 2021), FLOOD and CRIME (Liu and Li 2022), as well as SEASON and HOMEWORK (Zhang et al. 2022), are also used. WAR metaphors in Chinese news media reflect the stringent COVID policies in China, with their frequent use aiming to persuade citizens to comply with measures, such as social distancing and wearing face masks (Liu and Tay 2023). Metaphors with positive connotations were employed to describe China’s role in the pandemic (e.g., portraying China as a “war hero”), whereas metaphors with negative connotations were used to frame the US’s role (e.g., depicting the US as a “sinner”) (Liu and Li 2022). Since these studies rely mainly on data from 2020, their findings might not be readily applicable to later periods, particularly during times when COVID policies underwent significant changes.

2.2 Metaphors in policy making and policy persuasion

Metaphors are often assumed to affect how individuals understand public policies and therefore their acceptance of these policies (Charteris-Black 2011; Chilton and Lakoff 1995/2005). This claim is supported by evidence from studies employing a response-elicitation approach (see Boeynaems et al. 2017 for a systematic review). For instance, based on interviews, Lau and Schlesinger (2005) found that the American public’s understanding of policy metaphors – such as viewing healthcare as a marketable commodity versus a community obligation – could impact their support for American healthcare policies. Ahrens et al. (2022) concluded, based on between-subjects experiments, that novel metaphors, compared to conventional metaphors, could better draw participants’ attention and potentially lead them to evaluate proposed foreign policies more favorably. Hart (2021) reported that using extreme metaphors – such as depicting immigrants as animals or armies – reduced support for anti-immigration sentiments and hostile immigration policies compared to literal framings.

In the COVID-19 context, few studies have examined the association between metaphors and the acceptance of COVID policies. One notable exception is Schnepf and Christmann (2022), who investigated participants’ responses to American and German COVID policies using “war” versus “struggle” metaphors.

However, these studies focus on a relatively stable period with minimal changes in policies. Specifically, in the COVID-19 context, shifts in policies could lead to the use of different metaphors, potentially reducing the use of WAR metaphors and introducing new ones. Metaphors might also be used to foreshadow policy changes (Hobbs 2008). Additionally, the way the public interprets metaphor shifts may influence their responses to both old and new COVID policies. Addressing the latter requires a response-elicitation approach, which could be conducted in the future. This article will primarily focus on the former – whether and how shifts in policies could lead to the use of different metaphors.

2.3 Metaphors across languages and over time

Investigating metaphors in multilingual and diachronic contexts benefits our understanding of metaphor universality and variation (see Kövecses 2005). Examples include a compilation of pandemic metaphors in multiple languages (Olza et al. 2021) and a study examining over 100 years of pandemic metaphors in British media and political discourse (Taylor and Kidgell 2021).

In terms of multilingual metaphors, Liu et al. (2024) distinguish between metaphors used by different text producers and those employed by the same text producer(s). While the former typically emphasizes the ideologies of different text producers towards the same event, the latter highlights how the same text producer mediates its ideologies towards the same event when addressing different audiences, possibly through translation practices. This article examines the metaphors used by the same producer, Global Times, in its original Chinese news editorials and their English translations, linking these to how the newspaper frames China’s COVID policy, either similarly or differently, for the two target audiences.

In terms of diachronic analyses of metaphors, some studies have examined changes in metaphors over time in relation to policies, such as metaphor variations concerning Hong Kong’s education and economic policies (Ahrens and Zeng 2022; Zeng et al. 2021). They found that the shift in Hong Kong’s socio-political context following its transition from British to Chinese governance impacted the metaphors used by the Hong Kong government to frame its policies. Similarly, in the COVID-19 context, a pertinent question is whether metaphor usage changes during and after periods of significant policy changes. The COVID policy shift in China provides a valuable opportunity to explore this.

3.1 Data collection

This study comprises COVID-19-related news editorials from Global Times (环球时报, huánqiú shíbào). Global Times (GT) is the only media outlet in China that issues both Chinese and English newspapers. With a daily print circulation exceeding 2.6 million, 8 million daily page views, 2.5 million app users, and 100 million followers across social media platforms,^[1] GT is notable for advocating viewpoints that other state media outlets are reluctant to air openly (Cai 2016) and echoing the Chinese government’s policies (Liu 2022).

To better capture the newspaper’s stance, we chose to collect data exclusively from GT’s editorials, as they often more accurately represent a newspaper’s stance and opinions (Marques et al. 2019). GT’s English editorials are often translations of its Chinese editorials. We focus on both Chinese and English COVID-19-related editorials that contain any of the following keywords in either the title or content: “COVID” (新冠, xīnguān), “coronavirus” (新冠病毒, xīnguān bìngdú), “pandemic” (大流行, dàliúxíng), “epidemic” (流行病, liúxíngbìng), “pneumonia” (肺炎, fèiyán), or “Omicron” (奥密克戎, àomìkèróng).

In the present study, we limited the time span to January 2022 and February 2023 (both inclusive), considering both China’s COVID policies and the date we began collecting the data. China’s COVID policies went through three stages: (1) the strict zero-COVID policy since early 2020; (2) the dynamic-clearing policy since August 2021; and (3) the relaxation of nationwide COVID restrictions since December 2022. However, some Western analysts have suggested that “dynamic clearing” was merely a rebranded version of “zero-COVID”, given that COVID restrictions in China remained stringent.^[2] The selected time range starts in January 2022, which helps us understand whether and how COVID policy shifts occurred gradually and how these shifts might be reflected in the metaphors used by the outlet. The time range ends in February 2023, as we began collecting data in early March 2023. This timeframe should allow for an investigation of metaphor usage following the December 2022 COVID policy shift. We can also easily compare these findings with those from our earlier studies (Liu 2023; Liu and Li 2022), which analyzed data from the zero-COVID policy stage. Sections 4 and 5 will elaborate on this in detail.

The search using the specified keywords and time range yielded 82 Chinese COVID-19-related editorials and their English translations (see Table 1 for data description). Figure 1 below shows the number of collected editorials per month. It demonstrates that COVID-19-related editorials were published each month, increasing from 5 to 10 after the December 2022 policy shift.

Figure 1:

Number of editorials over time.

To answer our research questions, we annotate the metaphors describing COVID-19-related, China-related and the US-related topics in the collected editorials. Specifically, we generated concordance lines (window size: 150 words) using COVID-19-related and China/US-related keywords^[3] and conducted a word-by-word analysis for metaphors. Table 2 below lists the number of concordance lines generated using AntConc.^[4]

3.2 Identification of metaphors

According to both Metaphor Identification Procedure (MIP, Pragglejaz Group 2007) and Metaphor Identification Procedure Vrije Universiteit (MIPVU, Steen et al. 2010), a lexical unit is metaphorical when its contextual meaning contrasts with its basic meaning but can be understood in relation to it. We apply this definition here. Nevertheless, there are some adjustments at the operational level. On the one hand, MIP does not distinguish word classes or include similes in the annotation process, whereas we do. This is because of the fact that whether a word functions as a noun or a verb can affect its basic meaning and, consequently, its metaphoricity; similes, which liken one thing to another, reveal opinions and beliefs in the same way that metaphors do. On the other hand, while MIPVU distinguishes word classes, it also differentiates between direct and implicit metaphors and labels metaphor flags – an approach we do not use. We analyze both metaphors and similes but do not assign separate labels to them. We use the Chinese expression “考卷” and its English counterpart “exam” in example (1)^[5] to illustrate this.

The analytical procedure is as follows:

First, we read the entire text to establish a general understanding of the meaning: China has implemented well the “Class B management” for COVID-19.

Second, we identify the lexical units: “考卷” in the source text (ST) and “exam” in the translated text (TT); both are nouns.

Third, we establish their contextual meanings: both “考卷” and “exam” in the context implies that China’s COVID response was undergoing scrutiny.

Fourth, we determine whether they have a more basic contemporary meaning in other contexts: Yes, the basic meaning of “考卷”/“exam” is “a spoken or written test of knowledge”,^[6] which is often used in an educational context.

Fifth, we decide whether the contextual meaning contrasts with the basic meaning but can be understood in relation to it: COVID-19 pandemic is not education, but the contrast is understandable, as governments’ COVID-19 responses could test their governance.

Sixth, if the answer is yes, we mark them as metaphorical: Yes, Both “考卷” and “exam” are metaphorical.

3.3 Verification of source domains

To verify metaphors’ source domains, we employed the Source Domain Verification Procedure (SDVP, Ahrens and Jiang 2020). SDVP draws upon corpus-based resources including the Suggested Upper Merged Ontology (SUMO, www.ontologyportal.org), WordNet (wordnetweb.princeton.edu/perl/webwn), Handian Dictionary (www.zdic.net), and the Word Sketch function in Sketch Engine (www.sketchengine.eu). Similarly, we illustrate this method using the expression “考卷”/“exam” in example (1):

Step 1: Based on the basic meaning of “考卷”/“exam” and the context, we propose that the potential source domain is EXAM/TEST.

Step 2: By searching the word “exam” in SUMO, we find that the noun “exam” has one sense under the Noun Synset 10719702--TestForm: “a set of questions or exercises evaluating skill or knowledge”.^[7]

By searching the word “test” in SUMO, we find the noun “test” also has one sense under the Noun Synset 107197021--TestForm: “a set of questions or exercises evaluating skill or knowledge”^[8]

Step 3: We conclude that “考卷” and “exam” are conceptually-semantically associated with Exam/Test.^[9]

Step 4: We confirm that the source domain of “考卷” and “exam” is EXAM/TEST.

We identify the target domain as COVID RESPONSE based on the context. Hence, the cross-domain mapping is COVID RESPONSE IS AN EXAM/TEST.

3.4 Categories of translation methods

The study also identifies six types of translation methods for metaphors based on Toury’s (2012) categorization: retaining the same metaphor in the TT (M-M), replacing the original metaphor with a new metaphor (M1-M2), paraphrasing the metaphorical expression into a literal expression (M-P), translating a literal expression into a metaphor (P-M), omitting the original metaphor (M-0), and creating a metaphor from a zero-element (0-M). In example (1), “考卷” was translated into “exam”, with the same metaphor retained (M-M).

The data was annotated by a native Chinese linguist experienced in metaphor analysis for both Chinese and English. The annotator reviewed and recoded the data twice, with intervals of one to three months between reviews, to ensure accuracy. Any discrepancies were resolved through discussion with another metaphor expert. We chose intra-coder reliability over inter-coder reliability because the former ensures consistency and transparency in the coding process, whereas the latter “would not have reduced the ‘subjectivity’ of the coding per se” (Bednarek 2015: 6) but would only guarantee consistency in the use of the coding manual.

4.1 Metaphors over time: shifting narratives of the pandemic, China, and the US

Overall, more metaphors were used to describe the COVID-19-related topics (300 instances in the STs versus 274 instances in the TTs) than China-related topics (125 versus 103 instances) and the US-related topics (66 versus 49 instances). Hence, these editorials centre around framing various COVID-19 topics, such as the virus, the pandemic, and COVID-19 responses, with some references to China and the US as significant players.

Table 3 below lists the top 10 most frequently used metaphors for COVID-19-related topics.

The results show that WAR metaphors remain dominant, accounting for around 45 % of the total. This finding is consistent with our previous studies on GT’s COVID reports in 2020 (Liu 2023; Liu and Li 2022). While WAR, FIRE and SHADOW metaphors clearly convey a negative evaluation of the pandemic (see example 2 below), the evaluative meanings of EXAM/TEST, JOURNEY and WEATHER metaphors require further exploration in context. Example (1) in Section 3 and examples (3)–(4) below provide the contexts of these metaphors.

Example (1) not merely compares a government’s response to COVID-19 to taking an exam but also underscores the positive evaluation of the Chinese government (cf. “China will ace it”). A comparable effect is observed in example (3), where the conventional JOURNEY metaphor highlights China’s significant “strides” towards precision and science in COVID responses. In example (4), GT further emphasizes that the Chinese government has been making prudent decisions by comparing the current COVID condition in China to “clear skies” and “rainbows” after “storms” and “rains”.

When plotting metaphors in STs and TTs over time in Figures 2 and 3 below, we observed that the metaphors accumulated between January and May 2022 and between December 2022 and January 2023 in both STs and TTs.

Figure 2:

Metaphor use for COVID-19 over time in Chinese source news.

Figure 3:

Metaphor use for COVID-19 over time in English translated news.

No metaphors were used to describe COVID-19-related topics in October and November 2022, right before the policy shift, despite the publication of six COVID editorials during this time. The trendlines of WAR and JOURNEY metaphors indicate that GT increased the use of JOURNEY metaphors and decreased the use of WAR metaphors over time. In particular, the use of JOURNEY metaphors peaked in December 2022, coinciding with the time of the policy shift. This may be because GT reduced the need to prepare the audience for difficult times or to persuade them to comply with strict COVID-19 measures using WAR metaphors, which were more necessary during the earlier stages of the pandemic (Liu and Tay 2023; Seixas 2021). Instead, framing the pandemic as a journey could help the audience come to terms with the situation, implying that the return to normalcy is a gradual process (Olza et al. 2021). Following the policy shift, GT increased the use of EXAM/TEST metaphors. Although previous studies had identified EXAM/TEST metaphors in the GT’s 2020 editorials, their use was primarily to emphasize the challenges faced by the Chinese government (Liu and Li 2022). In contrast, in this study, EXAM/TEST metaphors are used to portray the Chinese government as a successful exam-taker.

Despite an overall decrease in the WAR metaphor frequency, WAR metaphors were still employed more frequently than JOURNEY metaphors between December 2022 and January 2023. As a result, the audience might continue to perceive the previous stringent COVID-19 policies as reasonable and necessary based on the ongoing dominance of WAR metaphors. However, the introduction of JOURNEY metaphors could also render the relaxation of these restrictions acceptable, implying that it is time to “move” to the next stage. This warrants further reception-oriented research.

GT paid attention to more varied China-related and US-related topics (see Figure 4 below). Besides the countries themselves, the governments, the party, the people and others were also mentioned.

Figure 4:

Frequency of China-related and US-related topics in metaphor distribution.

Table 4 below presents the metaphors describing different China-related and US-related topics. There are fewer diverse categories of metaphors describing the US, but more diverse categories describing China. Metaphors for China-related topics are evidently positive (e.g., LIGHT, ENGINE and BRIGHT SPOT). However, metaphors for US-related topics are mostly negative (e.g., American politicians as TRICKSTERS and the US as BAD FRUIT).

HUMAN metaphors remain dominant for both China-related and US-related topics, taking up around 70 % of the total. Figures 5 and 6 below plot HUMAN metaphors over time, with a cutoff frequency set at a minimum of 2 instances in either the source news or the translated news. The results corroborate that GT puts more emphasis on China than on the US in the selected time range, as HUMAN metaphors for China occur in 9 out of 14 months, whereas those for the US occur only in 6 out of 14 months.

Figure 5:

Raw frequency of HUMAN metaphors for China over time.

Figure 6:

Raw frequency of HUMAN metaphors for the US over time.

China was often associated with positive human characteristics, frequently described as “a person who shoulders responsibility” and “a person who does not lie flat” before the policy shift but as “a person who has vigor” and “a person who rolls up sleeves and works hard” after the policy shift. This is closely tied to the Chinese government’s shift in focus following December 2022, from a heavy emphasis on the domestic pandemic situation (see example 5) to economic recovery (see example 6).

In contrast, the US was often associated with negative human characteristics, frequently criticized as “a person who attacks others” and “a war loser” before the policy shift. However, after the policy shift, it received minimal mention, with only three instances describing it as an “actor” or “spoiler”. Compared to studies of the newspaper’s 2020 editorials (Liu and Li 2022; Liu 2023), the present study finds that while the newspaper’s positive stance towards China and negative stance towards the US remain unchanged, its focus has shifted over time. With changes in COVID-19 policies, the newspaper now emphasizes addressing domestic issues in China more than criticizing the US. Consequently, the audience may begin to focus on domestic economic recovery rather than China-US relations. This impact requires further research.

4.2 Metaphors across languages: translation practices in Chinese news media

This section addresses the second research question by comparing the metaphors used in GT’s Chinese editorials and their English translations. Figure 7 below plots the percentages of the six translation methods for the top 10 most frequently used metaphors describing COVID-19-related topics.

Figure 7:

Frequency count of translation methods for COVID-19 metaphors.

As can be seen, out of the 10 frequently used metaphors, three metaphors are mainly retained (M-M strategy): MOVEMENT (90 %), EXAM/TEST (83.33 %), and WEATHER (70 %). While MOVEMENT metaphors describe the progression of the pandemic as “sweeping” (“席卷”) and “rebounding” (“反弹”), EXAM/TEST and WEATHER metaphors primarily emphasize the desirable performance of the Chinese government in COVID-19 response and the mild pandemic situation in China, as shown in examples (1) and (4) above. The two target audiences received this information to a similar extent, as these metaphors were largely preserved in the translation process.

However, with 14.88 % of WAR metaphors paraphrased, 6.55 % omitted, and 2.38 % replaced, TT readers encountered fewer militaristic descriptions of the pandemic than ST readers. This applies to both periods before and after the policy shift, as shown in example (7) and (8) below.

In May 2022, when China practised the “dynamic clearing” policy, GT either paraphrased or omitted the WAR metaphor (cf. “抗疫”). In December 2022, when China eased its Covid restrictions, GT continued to use the “war” frame in its Chinese editorials (cf. “大战”) but omitted it in the English version. This observation aligns with the findings of our earlier studies (Liu 2023; Liu and Li 2022). Regardless of the stages of COVID-19 policies, ST readers were exposed to more WAR metaphors than TT readers. However, the impact of these WAR metaphors – whether they foster obedience (Benzi and Novarese 2022) or undermine public trust in government communication and potentially fuel conspiracy theories (Musolff 2022) – requires further investigation through reception-oriented studies.

The translation of FIRE metaphors was handled differently before and after the COVID policy shift. As shown in Figures 2 and 3 in Section 4.1, FIRE metaphors were not used after July 2022 in the English news, but it continued to be used in the Chinese news. Before July 2022, FIRE metaphors in the TTs (cf. “raging”) were primarily used to replace HUMAN metaphors in the STs (cf. “肆虐”), as shown in example (9).

However, in December 2022 and January 2023, FIRE metaphors were exclusively used in the STs, which were paraphrased into non-metaphorical expressions in the TTs (see example 10). Consequently, while before the policy shift, both ST and TT readers were presented with the pandemic as either a “brutal killer” or a “raging fire”, after the policy shift, such negative narratives were maintained more consistently in the Chinese news than in the English news.

HUMAN metaphors describing China and the US were also processed differently, as shown in Figure 8 below.

Figure 8:

Frequency count of translation methods for HUMAN metaphors describing China and the US.

Overall, HUMAN metaphors describing China were more likely to be maintained than those describing the US: 76.92 % versus 60 %. Conversely, HUMAN metaphors for the US are more prone to omission (14 %), replacement (12 %), or paraphrasing (14 %), indicating a tendency 2–5 times higher than that observed for China. Given that HUMAN metaphors for the US predominantly convey negative meanings (as summarized in Section 4.1), the reduced use of HUMAN metaphors for the US in the translated texts could result in GT’s TT readers encountering fewer negative descriptions of the US.

As shown in example (11), the ST metaphor THE US IS A PERSON WHO MAKES ANOTHER PERSON A SCAPEGOAT in the phrase “甩锅” (literally, throw pot) was paraphrased as a non-metaphorical expression “pass the buck”; and the original metaphor THE US IS AN ACTOR in the phrase “上演大戏” (literally, to perform a drama) was omitted in the TT.

HUMAN metaphors describing China were also slightly reduced in the TTs, with 7.69 % of them omitted and 8.97 % paraphrased. One oft-omitted metaphor is “Scapegoat”. Whether this omission was intentional or inadvertent, it could potentially lead ST readers to place greater emphasis on China-US relations and bolster their support for the previously implemented strict COVID-19 policies. However, the actual framing effects require further empirical investigation. In summary, while this study aligns with earlier research (Liu 2023; Liu and Li 2022) by demonstrating that ST readers encounter more negative descriptions of the US than TT readers, it diverges by highlighting a reduction in the use of metaphors framing the US. This change has resulted in a greater emphasis on metaphorically framing China rather than the US, especially following the policy shift in December 2022.