A small ad hoc comparable corpus was built for the Chinese and English articles on Sketch Engine, an online corpus software. The Chinese and English news articles were firstly converted into plain text format then formatted and noise eliminated. The Chinese texts were segmented by the CoreNLP tokenizer in Sketch Engine. The Chinese and English news texts were then compressed into an archive file and each uploaded to Sketch Engine.

The corpus consists of 35 pairs of translated Chinese articles from Culture Weekly (CW) and their original versions from various English-language media outlets. The Chinese articles were published in the eight issues of CW published in Reference News (RN) between February and March 2022. Several globally significant issues occurred during this period. Of particular relevance here are the Beijing Winter Olympics and Paralympics because they naturally attracted high global media attention and were a stage for displaying Chinese culture. The Russia-Ukraine conflict also broke out in this period.

The articles selected for this study were chosen because their sources of information and publication dates were identifiable, making it easy to identify their STs. The sources are presented in Figure 1. Five articles were selected from the New York Times, making it the most frequent source. Two articles each were chosen from the Spectator, New Scientist, Independent and Atlantic journals' websites and the Vox and CNN websites. The other 18 articles come from a range of global media. Most of these articles are in English, with the rest in French, Spanish, or German. The STs were retrieved by searching online for the media outlet, date of publication and news headlines provided at the beginning of translated articles in CW. The headlines of STs and TTs are listed in the Supplementary material.

The English subcorpus consists of 80,515 tokens or 68,685 words, whereas the Chinese subcorpus contains 37,410 tokens or 31,576 words. Thus, the size of the Chinese subcorpus is less than half (46%) that of the English subcorpus, which may be attributable to the omission of parts of the texts, given that “omission is a key strategy in the translation of news items, where material is tailored to the needs of a specific local audience” (Bielsa and Bassnett, 2009, p. 8). Table 1 provides the details.

Researchers in agenda-setting mainly employ traditional methods such as content analysis and manual coding to identify media and public agendas. As Dearing and Rogers (1996, p. 18) note on the measurement of agendas, “[t]he media agenda is usually indexed by a content analysis of the news media to determine the number of news stories about an issue or issues of study”. With the rapid growth of online media, researchers adopt innovative methods such as “big data” to explore agenda-setting in media spheres (Neuman et al., 2014). The present big data-based study utilizes MonkeyLearn, a text analytic tool powered by artificial intelligence and natural language processing to examine the comparable corpus, identifying keywords and conducting collocation analysis.

The analysis is conducted at two levels. First-level agenda-setting examines the cultural issues in the English and Chinese news respectively. The study employs the corpus tool of keywords analysis to identify the most salient keywords from the STs and TTs and classifies these keywords into various cultural issues. The term “keyword” is used in corpus linguistics to mean “a word that is more frequent in a text or corpus under study than it is in some (larger) reference corpus” (McEnery and Hardie, 2012, p. 245). The word list of the English subcorpus is compared to that of the reference corpus English Web 2020 (enTenTen20). Likewise, the word list of the Chinese subcorpus is compared to that of the Chinese Web 2017 (zhTenTen17). These keywords were classified by the topic classifier in MonkeyLearn which automatically detect topics in text data. This was then followed by a second-level agenda-setting analysis of the salient cultural attributes in the English and Chinese news agendas. The word sketch corpus tool compared the collocation of key cultural attributes in English and their equivalents in the Chinese news agenda, as described in Kilgarriff and Tugwell (2001). The collocates of these attributes were examined in their contexts to identify patterns of translation strategy.

Table 2 illustrates the top 100 keywords in the English news subcorpus. These keywords mainly fall into the eight topic categories of politics, health, humanities, food, entertainment, environment, science, and sports. Over half of the keywords relate to politics, making it the most salient issue in the English news agenda. These keywords relate mainly to two political matters: the activities of the Saudi crown prince and the Russia-Ukraine conflict. A total of 32 keywords come from “Absolute Power”, an article about Saudi Arabian Crown Prince Mohammed bin Salman published in the April 2022 issue of the Atlantic. The Russia-Ukraine conflict, which started in February 2022, generated 26 related keywords. The second most salient issue in the English news is health. Six keywords on COVID-19 originate from an article published in Time in January 2022 (Park, 2022). An equal number of keywords were derived from a Vox story in January 2022 (Pandell, 2022). The third most prominent topic in the English news agenda is humanities. Six keywords are from a story which appears in the March 2022 issue of National Geographic magazine. A few topics related to humanities, such as the history of the handkerchief, science fiction, the tragic deaths of famous persons, and the evolution of Chinese characters. The food topic features five keywords from three recipes. The entertainment topic is marked in a Vox article about how Rayne Fisher-Quann, a famous writer and cultural critic on TikTok, spends her day online. Environment issues such as climate change and waste recycling are highlighted in two articles, including one about Thomas Graedel, a professor of industrial ecology at Yale University, and the other about the Caldor Fire in California and Thomas Gorden, a member of Fire Twitter. The science issue is featured in an article in the Atlantic about Karen Richon, an aerospace engineer at NASA Goddard Space Flight Center. The sports topic occurs in just three keywords relating US skier Mikaela Shiffrin, Finnish ice dancer Matthias Versluis, and US ice dancer Kaitlin Hawayek being interviewed about the food they ate in the Beijing Winter Olympic village.

Table 3 shows the top 100 keywords in the Chinese subcorpus. These keywords mainly fall into the six topic categories of politics, humanity, environment, health, food and sport. In the Chinese news, political issues are still the most prominent, with a total of 33 keywords featuring the Russia-Ukraine conflict and four keywords featuring the Saudi crown prince. The environmental issue of wildfire is highlighted with nine keywords from an article in Wired. The health issues of trauma and COVID-19 were discussed in two articles in Vox and Time with five and three keywords respectively. Food topics, such as the recipes already mentioned, reoccur. Besides these, the closing of McDonald's restaurants in Russia as a result of the war in Ukraine and rice as a popular food with foreign athletes in the Olympic village were presented. Like the English news, the sports topic has the least prominence on the Chinese news agenda, with two keywords featuring ice dancing athletes from Spain and the Czech Republic, and the US ice dancer Kaitlin Hawayek.

Political issues rank among the first issues in both the English and Chinese news agendas. The Russia–Ukraine conflict was a political issue that is both salient in the English and Chinese news agenda. The activities of the Saudi crown prince were another political issue that is salient in the English news agenda with 32 keywords. However, it is less salient in the Chinese news agenda, with five keywords. The issue of the Saudi crown prince's activities received considerably less attention in the Chinese news agenda than in the English one. The royal attributes are largely maintained, with a set of keywords representing monarchical power, whereas the cultural attributes are greatly undermined, with only a few keywords on the historic and political attributes. The environmental issue of wildfire is given more prominence, whereas waste recycling receives less. The health issues of COVID-19 and trauma had less emphasis in the Chinese news agenda than its English counterpart, particular in the case of the issue of trauma. The topic of food was given slightly more attention in the Chinese agenda.

This subsection compares the salient attributes of the cultural issues in the English and Chinese news agenda and examines whether the salience of cultural attributes in the English news is transferred to the attribute agenda of the Chinese news. It compares the collocations of “culture” and “cultural” and their Chinese equivalent “文化” (culture) in the English and Chinese subcorpora. These examples were selected to discuss instead of others because they are prominent collocates and examples are presented to show how these collocates are translated in specific contexts.

Figure 2 shows the word sketch of “culture” in the English subcorpus. Table 4 illustrates the scores and frequencies of these collocates in the English subcorpus. “Culture” appears 33 times in the English subcorpus and mainly collocates with words in five categories. The modifiers of “culture” include “many”, “Asian”, “pop” “American”, and so on. Regional or national cultures like African cultures or Saudi and religious cultures such as Christian, Islamic and Jewish are represented in the English news. “Culture” also encompasses popular and “throwaway” culture, as well as subcultures of a specific historical period, like Pythagorean culture and plantation culture. This word often collocated with others like “religion” and “language”. Culture is often placed on an equal footing with religion and language, as the roles of all of these are important in society.

The adjective “cultural” occurs 17 times in the English subcorpus. The word sketch of “cultural” in the English subcorpus falls into two categories, as shown in Figure 3. The scores and frequencies of the collocates of “cultural” in the English subcorpus are shown in Table 5. The nouns modified by “cultural” are “evolution”, “bankruptcy”, “touchstone”, “reference”, “decade”, “artifact”, “value”, “difference”, “shift”, “Fisher-Quann”, “term”, “trauma” and “language”. “Evolution” is the most frequent noun modified by “cultural” in the English subcorpus. The adjectives used together with “cultural” are “intellectual”, “seismic”, “technical”, “religious”, “current”, “historical”, “social” and “political”.

The Chinese equivalent “文化” (culture) occurs 32 times in the Chinese subcorpus, almost the same number as in its English equivalent. Figure 4 shows the word sketch of “文化” (culture) in the Chinese subcorpus falling into six categories. The frequency and score of its collocates in the Chinese subcorpus are shown in Table 6. The nouns which are used to adjectivally modify “文化” (culture) are “亚洲” (Asia), “美国” (America), “当今” (present), “美国人” (American), “民族” (nation), and “英国” (Britain). “文化” (culture) is used together with phrases including “语言” (language), “纽带” (tie), “医生” (psychiatrist), “知识” (knowledge), and “历史” (history). “文化” (culture) modifies “巨变” (shift), “大臣” (secretary), “价值观” (value), “试金石” (touchstone), “象征” (symbol), “文物” (artifact), “术语” (term) and “创伤” (trauma). The other nouns in possessive constructions with “文化” (culture) are “隔阂” (barrier), “琳达·郑武” (Linda Trinh Vo), “传统” (tradition) and “变化” (change).

The word sketches of “culture” and “cultural” and their Chinese equivalent “文化” (culture) represent the cultural attributes in the English and Chinese news agenda. Some cultural attributes of the English news agenda were maintained in the Chinese agenda. As Example 1 shows, one New York Times article says that there are many kinds of cultures, the most prominent of which are Asian, American and Asian American cultures. As most of RN's readers are Chinese, its translated article maintained this cultural attribute by translating it faithfully to “亚裔美国文化” (Asian American culture), thus maintaining the cultural attribute that resonates with the readers.

As Figure 3 shows, one of the most frequent nouns modified by “cultural” is “evolution”. Examples 2 and 3 show two of the four instances of “cultural evolution” which occur in a New Scientist 26 March cover story about how languages evolve. These two examples relate the evolution of languages to cultural evolution which is translated into “文化进化” (cultural evolution) or “文化的进化” (cultural evolution) in the two instances in the corresponding 28 March RN article.

Some cultural attributes of the original news agenda regarding popular cultures were omitted in the translated news agenda, as is shown in Example 4 from a Deutsche Welle article about the meaning of numbers, which was originally written in German and translated into English. However, the mention of pop culture in the middle of the sentence was omitted as well as the name of the director Sean Cunningham at the end of the sentence. This might be because the movie is the only example of pop culture in this sentence and the film director may not be familiar to the Chinese readers.

The religious attributes of the cultural agenda in the Chinese news are greatly downplayed in the Chinese versions. When “culture” occurs with “religion”, it is not translated into Chinese, as is shown in Examples 5 and 6. Example 5 is from a Newsweek cover story about the Ukraine refugee crisis which quotes the deputy director of an NGO describing Ukrainians as sharing cultural and religious connections with Europeans. Example 6 quotes author Udo Becker comparing the meanings of numbers in different cultures and religions. These two sentences are not translated into the Chinese versions.

The linguistic attribute in the Chinese news agenda is strengthened. When “culture” occurs with “language”, it is translated into Chinese in the way shown in Example 7 from the Wall Street Journal review of Kingdom of Characters on the history of Chinese characters. “文化” (culture) and “语言” (language) are used together in a similar footing, as is shown in Example 8 from an Independent piece about the TV show “Servant of the People” that features now-Ukrainian President Volodymyr Zelensky. “Cultural” and “language” are used together as a modifier for “barriers”, which is translated into “语言和文化的隔阂” (language and cultural barriers) in Chinese.

Example 9 from the New York Times describes close cultural and linguistic bonds between Russia and Ukraine. The adverbs “culturally” and “linguistically” were translated as “在文化和语言上” (culturally and linguistically) in Chinese.

This paper has argued that news translation, as a form of recontextualization, performs the agenda-setting function in the case of CW. Translation sets the agenda of the original news texts for the readers of the TTs. When translated into Chinese, the salience of cultural topics in the English news is mostly transferred into the issue agenda of CW. Some of the cultural topics were given more prominence, such as the linguistic attribute, whereas the religious attribute is greatly undermined in translation. A possible explanation for this may lie in the news outlet's editorial policy, which discourages the dissemination of religious beliefs among its domestic readers. However, the findings of the current study do not support those of Cheng et al. (2016) that limited salience of issues and attributes occurs between the news produced by Xinhua News Agency (XNA) and the US news outlets. This result may be explained by the fact that the data examined in Cheng et al. (2016) are not necessarily translations.

As XNA runs RN and heavily regulates the views expressed in it. Many regard XNA as an agent of the Chinese central government and the Communist Party of China (CPC). The current study has found that the translated news in CW clearly aligns with the central government's view on various cultural issues. This finding supports previous studies of the ideological positioning of RN's translation practices (e.g., Pan, 2014; Xia, 2019; Pan and Liao, 2021). The newspaper strictly adheres to the XNA style guide which bans the cultural use of sensitive words in relation to religion. This may be because, as Shirk (2011, p. 24) suggests, that XNA “has undergone a remarkable makeover” in recent years in order to compete with emerging commercial media outlets. These results also accord with He's (2003, p. 198) findings which showed that the Chinese media bodies, including XNA, employ a discursive strategy of “ideological re-pitching” which commingle the narrative of “order” with communist ideology and maximize patriotic sentiment.

Consistent with previous ones (e.g., Wu, 2018; Xia, 2019), this study has found that translated news in RN undergoes a recontextualization process. RN translators are often required to recontextualize their textual productions to make them suitable for their diverse readerships and the target culture (Davies, 2006). Rogers and Dearing (1988, p. 571) note that “[m]edia gatekeepers have a general idea of the news interests of their audience, and this perceived priority of news interests is directly reflected in the news values with which media personnel decide the media agenda”. The media outlet converts news articles designed for another, largely foreign, readership and rewords them for readers in different environments who may have different ideological and linguistic backgrounds. Thus, many ideological and linguistic shifts are likely to occur during this process of news translation. The original news is intended for readers of English who are mostly outside the Chinese mainland. The readership of translated narratives in RN is mainly Chinese mainland residents. RN also runs an official website and social media platforms accessible around the world, but much of the material online differs from that which is in the newspaper (Zeng and Li, 2021). RN readers are mostly members of the CPC or government cadres who have received higher education. Hence, one could reasonably argue that the newspaper is a party mouthpiece creating articles which cater for its clientele's reading habits.

Although news stories claim to present an objective account of an important event (Tuchman, 1972) and the translator purportedly aims to provide a faithful rendition of the original narrative, many factors contribute to this complex agenda-setting process, including translation conventions, institutional procedures, and socio-cultural contexts. Taking all these factors into consideration, the original news tends to undergo linguistic changes to cater for the audience's needs. The news translator needs to deal with various power dynamics relating to the political stance of the original journalist, that of the editor and owner of the media outlet, and those of the target readership. Translated news is often the product of mediation between different power relations in the media outlet and a manifestation of the relationship between the media outlet and its target readers. From the perspective of the function that translation plays, translated news set the original agendas both linguistically and ideologically through various forms and mechanisms. This creates a range of different text trajectories and recontextualizations that seek to cater for a diverse target readership.