“In Other News”: China’s International Media Strategy on Xinjiang—CGTN and New China TV on YouTube

In the Western world China stands accused of severe human rights violations regarding its treatment of the Uyghurs and other predominantly Muslim minorities in its northwestern Xinjiang Uyghur Autonomous Region. This is the first article to systematically analyze the response of China’s international state media to these allegations. By studying the YouTube channels of two leading Chinese state media, China Global Television Network (CGTN) and New China TV (operated by Xinhua News Agency), it presents an in-depth understanding of how China’s foreign-facing propaganda works in a crucial case. The quantitative content analysis highlights how China reacted to increasing international (mostly United States) pressure regarding its Xinjiang policies by producing higher volumes of videos and putting out new counternarratives. The qualitative analysis that follows provides in-depth treatment of the most important discourses that Chinese media engage in to salvage the nation’s international image, namely those on development, culture, nature, and terrorism. It finds several ways of countering criticism, ranging from presenting a positive image of China, in line with traditional propaganda guidelines and President Xi Jinping’s assignment to state media to “tell the China story well,” to more innovative approaches. Thus the development narrative becomes more personalized, the discourse on culture supports the “heritagization process” to incorporate minority cultures into a harmonized “Chinese civilization,” representations of nature firmly tie Xinjiang into the discourse of “beautiful China,” the “terror narrative” strategically employs shocking footage in an attempt to gain international “discourse power,” etc. The article provides an up-to-date picture of China’s state media strategy on a highly contentious international issue.

to "tell the China story well," to more innovative approaches.Thus the development narrative becomes more personalized, the discourse on culture supports the "heritagization process" to incorporate minority cultures into a harmonized "Chinese civilization," representations of nature firmly tie Xinjiang into the discourse of "beautiful China," the "terror narrative" strategically employs shocking footage in an attempt to gain international "discourse power," etc.The article provides an up-to-date picture of China's state media strategy on a highly contentious international issue.

Keywords
Xinjiang, YouTube, discourse analysis, Chinese state media Since 2018 criticism of China's human rights record has intensified and China's government has been challenged in the international court of public opinion with respect to its mass detention of Uyghurs and other mostly Muslim ethnic minorities in Xinjiang.After initial denial of the very existence of "reeducation camps," China's leadership changed tack and admitted that it had built "vocational education and training centers" (hereafter, VET centers) for them.The purpose of these centers was said to be the "de-radicalization" of those suspected of being influenced by "extremist thinking" and the training of them in Mandarin Chinese, legal knowledge (including patriotism), and occupational skills.This was said to improve the livelihoods of the local minority populations while eradicating the socio-cultural basis for (ethno-national) separatism, (religious) extremism, and (violent) terrorism.Yet among Western observers these official claims were widely dismissed as a thin veneer for forced political indoctrination and even "cultural genocide" (Roberts, 2020).Instead of relenting, accusations targeted new areas such as torture, coercive sterilizations, and forced labor in Xinjiang (Alpermann, 2022).Compounded by other pertinent issues, such as Sino-US frictions, protests in Hong Kong, etc., it is fair to say that the Chinese government faces a veritable international image crisis (Jacob, 2020;Silver, Devlin, and Huang, 2020;Uyghur Human Rights Project, 2020: 12).
In response to this crisis, the government has taken a range of countermeasures to mitigate the fallout, manage its image, and, if possible, create a counternarrative that casts China in a more positive light.One of the forums in which this contestation over meaning making has played out most directly and prominently is the video-sharing platform YouTube.Though blocking access to it for domestic users, China's government is, ironically, making highly active use of this platform to advance its messages globally.This article focuses on the role played by Chinese state media-exemplified by China Global Television Network (CGTN) and New China TV-in this contest.Our hypothesis is that China has adapted its strategy in three main ways: First, it has intensified the propagation of its messages in quantitative terms to counter the international image crisis created by Xinjiang-related criticism.Second, the content of its messaging has become more varied as it has intentionally tried to link Xinjiang policies with international terrorism but at the same time also use it for a more general critique of "Western double standards" and promotion of its own perspectives on human rights.Third, and most intriguingly, it has used innovative and creative ways of reporting to generate higher impact and more trustworthiness.CGTN and its parent company China Central Television (CCTV) have long copied the style of CNN, the BBC, Al-Jazeera, and other international news networks to stake out their own ground in terms of content.But more recently China's media products have become more variegated with the greater use of social media for message dissemination.
Our primary data consist of all CGTN and New China TV videos relating to Xinjiang posted on YouTube from January 2018 to December 2020.In total, the sample contains more than eight hundred videos of varying length.Inspired by Matteo Vergani and Dennis Zuev (2011), we use a mixed-methods approach for analysis.First, we present a quantitative overview based on content analysis to outline the general trends of Chinese state media strategies on Xinjiang as found on YouTube.Second, we employ discourse analysis (Keller, 2011) for a more in-depth examination outlining the main features of the discursive strategy adopted by Chinese state media in this case.Thus our study contributes to an understanding of China's crisis managementwhich can be compared to other crises such as the COVID-19 pandemic.It also sheds light on China's participation in media globalization more generally as well as on the specific representations and constructions of China's Xinjiang policies and issues.It shows how the channels adopt Western (primarily US) media formats and discourses to frame their policy agenda in terms familiar to the target audience, while "bandwagoning" on the current fragmentation and politicization of news media in an era in which the terms "fake news" and "alternative facts" have become everyday accusations in order to provide "other news" to the world.But first, we discuss the internationalization of China's state media influence more broadly.

The Globalization of China's State Media
Like all communist-party-led polities, China has a long history of outwardfacing propaganda, using print but also audiovisual materials.But two factors have combined to fundamentally transform the official face that China shows to the world.First, the media landscape has evolved considerably along with technological advances such as satellite and cable television, the internet, social media, etc.These present China with new opportunities to send its preferred messages to target audiences worldwide; but, of course, they also involve new challenges (Thussu, de Burgh, and Shi, 2018).Second, along with its economic rise, China has become more confident and developed its own soft-power strategy to "tell the China story well," as the Chinese Communist Party (CCP) general secretary and state president, Xi Jinping, has repeatedly admonished the state media to do (Brady, 2015: 56;China Daily, 2016).
To counter a perceived lack of "discourse power" (huayu quan 话语权), the Chinese government has allocated large sums of money to globalize staterun media (Yang, 2018).Its goal is to break through the perceived monopoly of international (i.e., Western) media.A significant step in this direction was the creation, in 2016, of CGTN-emerging from CCTV-9, the first twentyfour-hour English-language news channel in China, broadcast by CCTV since 2000 (Hu, Ji, and Gong, 2018: 69).While CCTV continues to operate, for instance, in the US cable television market (Cook, 2020: 23), its brand has taken a step back, 1 and CGTN is being promoted as the primary television channel for international audiences, followed by New China TV, the channel operated by Xinhua (or New China) News Agency.According to its website's mission statement, CGTN "aims to provide global audiences with accurate and timely news coverage as well as rich audiovisual services, promoting communication and understanding between China and the world, and enhancing cultural exchanges and mutual trust between China and other countries." 2Arguably, this is the internationally more palatable version of Xi Jinping's original statement during a January 2014 Politburo meeting in which he tasked the state media to "tell the China story well": "China should be portrayed as a civilized country featuring a rich history, ethnic unity, and cultural diversity, and as an Eastern power with good government, a developed economy, cultural prosperity, national unity, and beautiful scenery" (cited in Brady, 2015: 55).
Given the acrimonious relations between China and the United States, not everyone is convinced of such benign descriptions.A report published by Freedom House in January 2020, for instance, delineated Chinese state media goals much more bluntly: "To promote a positive view of China and of the CCP's authoritarian regime; to encourage foreign investment in China and openness to Chinese investment and strategic engagement abroad; and to marginalize, demonize, or entirely suppress anti-CCP voices, incisive political commentary, and exposés that present the Chinese government and its leaders in a negative light" (Cook, 2020: 3).
Reports such as this one have become more frequent in recent years and generally warn against Chinese influence operations, although their effectiveness cannot be easily established (Ohlberg, 2018;Uyghur Human Rights Project, 2020).A report published by the Hoover Institution, for instance, noted that between June 2019 and June 2020 CGTN's YouTube channel had "1.7 million subscribers and 1.1 billion views" (Diresta et al., 2020: 42).But compared to RT (previously known as Russia Today, the main foreign-facing state media of Russia), with 3.9 million subscribers and three billion views between June 2019 and June 2020, there appeared to be room for improvement. 3Moreover, CGTN's rate of active viewer involvement over the same period was about eleven times lower than RT's (Diresta et al., 2020: 42).Nevertheless, the authors obliquely warn that "China's potential to refine and increase its capabilities . . .remains of significant concern to liberal democracies" (Diresta et al., 2020: 47).In February 2021, Ofcom, the British regulatory agency responsible for television broadcasting, even stripped CGTN of its UK broadcasting license for being "party controlled" (Hern, 2021).
Interestingly, China's own media analysts tend to use the same benchmarks but often seem less sanguine about their state media's success.In fact, there is considerable debate on the most promising strategy.For example, Zhu Zhibin and Jiang Yunai (2018) explicitly study RT's YouTube strategy as a role model for China's media.Finding "hard topics" such as politics, war, and economics to be predominant on RT's YouTube channel over the years 2007-2017, they argue that China's state media should likewise take a clearer stand in international affairs and try less hard to please audiences.In contrast, Wei Xiao and Fan Pan (2018) contend that CGTN is following a strategy of "depoliticization" 去政治化 on its Facebook main page, with some success.But they criticize a mismatch between content supplied by CGTN and users' preferences.Thus they suggest focusing on more engaging nature broadcasts rather than on social topics that do not enlist as much user interaction.
They are not alone in this regard.Zang Xinheng and Lu Xinran (2019) also conduct a comparison between RT and CGTN videos posted on YouTube and find that the Chinese channel's impact on viewers significantly lags behind that of RT.They suggest that the much larger share of "entertainment" items on CGTN, compared to RT's emphasis on hard news, is weakening the former's authoritativeness.They too are of the opinion that CGTN needs to position itself clearly and speak up forcefully on China's behalf.On the other hand, they argue for showing more personal stories that touch the audience emotionally and, explicitly, for "more pandas." Perhaps the most authoritative is Jiang Heping's (2019) analysis, since he served in a number of leading roles within CCTV and is currently responsible for CGTN's North America office.He lists several major problems afflicting the station.These include being too closely wedded to domestic propaganda; lacking good storytelling skills, internationally influential programs, and its own style of reporting; as well as unbalanced regional coverage and limited online interactivity.Jiang explicitly recommends creating several flagship programs such as The Point with Liu Xin or Dialogue with Yang Rui (more on these below).Anchors like Liu Xin 刘欣 and Yang Rui 杨锐, Jiang argues, should become crucial commentators on social media and "key opinion leaders" in their own right.With a more personal style of communication, their social media profiles will be more convincing than the official CGTN channel.In terms of regional targeting, earlier studies found that China's state media had concentrated on developing countries.Now, however, Jiang and others recommend targeting the G7 and G20 countries more explicitly (Jiang, 2019: 61;Zhang, 2019: 24).
Of importance for our discussion is that despite their notable disagreements, there is consensus among these authors that short videos-anything from below one minute to several minutes in length-are much more effective on social media than longer shows.They tend to be shared more often and therefore have a higher impact (Zhu and Jiang, 2018;Zang and Lu, 2019).
Given these domestic and international debates on Chinese state media's global presence and strategies, it is important to study specific examples of implementation.COVID-19 is one crucial case (Jacob, 2020), but so is Xinjiang, on which we will focus below.While some studies have addressed the clandestine use of social media to produce and disseminate personalized propaganda (Kao et al., 2021;Ryan, Impiombato, and Pai, 2022;Steenberg and Seher, 2022), this is the first study to focus on the YouTube channels of China's official state media.

Methods and Data
Our goal in this article is not to assess Chinese state media's effectiveness in influencing viewers or public opinion at large.Rather, we want to analyze the discourse that state media carefully construct to counter the Xinjiang-related revelations of recent years.We examine how different aspects of this official discourse are being foregrounded to repel specific accusations.Moreover, we are interested in understanding the evolution of new formats of foreign-facing propaganda.Thus we focus the analysis on one particularly important social media platform, namely, YouTube.As discussed above, this is among the primary platforms identified by Chinese media analysts as battlefields in ideological contestation with the outside world.It is also a comparatively low-cost channel by which to influence international public opinion and craft a positive national image for China.With around two billion monthly active users in 2022, YouTube is almost on par with Facebook as a social media platform in terms of users and ranks as the most popular video-sharing platform worldwide (Mohsin, 2020).
To allow comparison not just across time but also between media outlets, we identified and downloaded all videos relating to Xinjiang posted either by CGTN or New China TV from January 2018 to December 2020.To do this, the full list of videos by the two channels was manually screened for content on a regular basis by the second author.The sample includes all videos covering or set in Xinjiang, regardless of whether the name Xinjiang actually appears in the title.Related videos (e.g., regarding human rights or terrorism discussions in general) were also included.The total corpus comprised 834 videos with a combined running time of almost 797 hours.Some duplicate videos (e.g., fragments of longer documentaries already categorized) were later removed from the sample.In September 2020, CGTN started livestreaming fixed-camera footage of Xinjiang scenery from various locations with added music on an almost daily basis.More than forty such livestreamsbest described as screensavers and lasting up to twelve hours-were hosted until the end of that month alone, and another thirty-five in October 2020.We exclude these, as well as any "travelogue" livestreams of longer than one hour, from the final sample used in subsequent sections because they would have skewed quantitative analysis.The total running time of the 711 videos which made it into the final sample used for the quantitative analysis is slightly over 69 hours.
To get an overview of the material, both authors watched approximately 50 percent of the material and independently conducted inductive, open coding (Richards, 2009: chap. 5).Next, the authors compared and discussed their codes to standardize them.For the remainder of the material all videos were again independently coded by each author, then compared and divergences between the coders discussed to resolve them.Though laborious, this process ensures the highest intercoder reliability and intersubjectivity.This coding was then used for the content analysis outlined in the following sections.At the same time, the first author wrote short memos that form the basis for the qualitative discourse analysis to follow.In this, we generally follow Reiner Keller (2011) and his sociology-of-knowledge approach to discourse while keeping in mind the particularities of Chinese political discourses (Alpermann and Fröhlich, 2020).While the authors independently worked on different aspects of the qualitative analysis, all interpretations were discussed and cross-checked.Thus we combine quantitative and qualitative methods to present a more comprehensive picture.
As a baseline for comparison, we analyzed the state-propagated discourse on Xinjiang over the past two decades.Here, we selected the relevant white papers of the Chinese government, published by the State Council Information Office (SCIO) in 2002-2020.Eleven of these white papers address Xinjiang directly, while the rest of the sample concerns human rights and ethnic policies more generally (see Appendix Table A1).These materials serve as background to better understand the narratives presented via state media.For the time period under scrutiny our analysis reached the point of "theoretical saturation" (Corbin and Strauss, 2008: 263).New trends emerging from 2021 onwards will require separate analysis building on the baseline established here.

Timeline
To gain perspective, we first establish the timeline of this study and highlight important episodes of contention.When we started this project in early 2020, it had already become clear that China's state media were upping the ante on YouTube with respect to Xinjiang.Yet little did we know that this was to become such a hot topic over the next few months.Figure 1 presents an overview of the number of Xinjiang-related videos uploaded to YouTube by either CGTN or New China TV as well as the total running time by month.We start our analysis in 2018, when the international media's and the scholarly community's attention to mass incarceration in Xinjiang was picking up steam (Zenz, 2018;Roberts, 2020: 213-14), but find that YouTube did not at first play a significant role in China's strategy to counter mounting human rights criticism.For most of 2018, China's official spokespersons were fully denying the very existence of reeducation facilities for Xinjiang's Muslim population (Cumming-Bruce, 2018a;Roberts, 2020: 208-18).Only in October of that year did the People's Republic of China (PRC) government conduct its about-face, admitting to the existence of the camps and framing them as VET centers, as noted above (Cumming-Bruce, 2018b).In November 2018, the SCIO published a new white paper on Xinjiang (White paper L; see Appendix Table A1 for full details of this white paper and subsequent white papers cited by letter), but its main topic was the protection of cultural heritage.While it included some ominous statements-such as "adapting to local society is essential for the survival and development of any religion" and "Xinjiang upholds the tradition of religious localization and provides guidance to religions on adaptation to China's socialist system"-its major thrust was to present a positive image of the Chinese government's efforts to protect local cultures, in line with the "heritagization process" discussed below.
This strategy changed, however, in early 2019 as international pressure mounted.We can see an uptick in frequency and volume of Xinjiang-related videos in our sample over the first three months of that year.This was clearly in connection with a March 2019 white paper, the first to address the allegations head-on during this crisis (White paper N).Two more white papers were to follow in July and August 2019 (White papers O and P), alongside diplomatic tussles (Yellinek and Chen, 2019).Though production of Xinjiangrelated videos was not sustained at the same level throughout the year, there was another minor peak in July, as alleged coercive labor practices were made public (Zenz, 2019).These accusations were rejected in several videos and dismissed curtly-with VET centers described as conducting "practical training . . . in classes rather than through employment in factories or enterprises, or obligatory labor"-in the August 2019 white paper.
Xinjiang-related videos surged toward the fall and eventually reached a peak in December 2019 with sixty-four videos (excluding duplicates) and a total running time of well over six hours in that month alone.This overlapped with deliberations in the US Congress over legislative action to protect the Uyghurs and other ethnic minority groups in Xinjiang against alleged human rights violations (Lipes, 2019), which are frequently directly addressed in the videos.After the US Senate passed its Uyghur Human Rights Policy Act of 2019 in September, the House of Representatives passed its own, stronger version in early December.A series of leaks bringing to light classified and other official Chinese documents added further fuel to the fire: the "Xinjiang Papers" (Ramzy and Buckley, 2019) and the "China Cables" (International Consortium of Investigative Journalists, 2019) were both published in midto late November 2019.These revelations in Western media heightened international attention on the Xinjiang situation and explain the spike in Chinese counterpropaganda.
After a short hiatus in February 2020 (i.e., during the COVID-19 lockdown), March and April again saw higher levels of Xinjiang-related video output in response to renewed allegations of the use of Uyghurs as forced labor (Xu, Kim, and Reijnders, 2019).This also coincided with yet another leak, of the "Karakax List" (Zenz, 2020a), in mid-February.While May was surprisingly quiet-maybe due to COVID-19 and other topics dominating news coverage-the Xinjiang topic heated up again in June and July.It was at this time that the US Congress passed a harmonized version of the Uyghur Human Rights Policy Act of 2020 (on May 14 in the Senate and on May 27 in the House of Representatives); President Donald Trump signed it into law on June 17.Following this, the US government announced sanctions against top Xinjiang officials, including Chen Quanguo 陈全国, the region's party secretary, in early July (Verma and Wong, 2020).Adding a new line of attack, a report published by the US-based Jamestown Foundation in June 2020 accused China of forcefully depressing birthrates among Uyghurs and other non-Han populations in Xinjiang (Zenz, 2020b).This corresponded with another spike in our sample data for July, with fifty-four videos in that month.
Finally, as the US government also discussed broad restrictions on cotton and apparel produced in Xinjiang using allegedly coercive labor in early September (the bans eventually promulgated in mid-September were not quite as wide ranging, see Swanson, 2020aSwanson, , 2020b)), Chinese state media on YouTube went into overdrive.CGTN and New China TV posted a total of 197 videos that month with a jaw-dropping playing time of 399 hours and a half, followed by another 180 clips with almost 360 hours the next month.For sanity's sake, we exclude the livestreams of over one-hour length from our analysis in the following sections.But this still leaves us with 148 videos with some twenty-two hours of footage in September and 134 videos with a total running time of fifteen hours in October.September 2020 also saw the publication of another white paper on Xinjiang (the fourth such paper in the preceding two years)-unsurprisingly on "employment and labor rights" in the region (White paper R).Thus we can see that traditional forms of foreignfacing propaganda and new methods mostly worked in concert to address urgent challenges to China's international image.Most importantly, we can identify several episodes of international contention and an increasing use of YouTube as a platform to fight against accusations from abroad.

Content by Category
Next, we delineate the content of these videos.As previously discussed, the coding proceeded in a bottom-up and open-ended manner based on the final sample of 711 videos.Therefore, not all categories are at the same level of specificity, and the larger ones contain several subcategories.Also, looking at the total number of videos or at the total running time of those videos presents us with differing pictures.Starting with the primary coding, which broadly structures the data, Table 1 shows that by far the largest category by total number of videos is the "development narrative."Many of these are among the shortest videos in the sample.Thus this category accounts for 39 percent of the total number of videos, but only around 16 percent of total running time.However, as will be argued below, running time is not a good measure of the importance attached to a narrative by the producers or, in fact, of audience impact.Prominent subcategories in this category are "infrastructure" (15% of within-category running time), "poverty alleviation" (12%), "tourism" (8%), and "agriculture" (6%).However, by far the largest subcategory is "personal development stories" that portray an individual (103 videos, 45% of within-category running time).These will be presented in more detail below.Suffice it to say for now that state media churned out these portraits at a "Great Leap Forward pace" during July-October 2020.
The opposite relationship between frequency and length is observed for the second-most important category, namely, "culture."This accounts for 17 percent of the number of videos but some 34 percent of running time.This can be explained by the frequent use of livestreaming in this category.Livestreams are usually low on information density.Though livestreams are produced by journalists, they arguably approach user-generated content (UGC), which, according to Chinese media analysts, has several advantages: UGC is low cost since it needs minimal editing; its rough edges make it appear more authentic and, hence, more trustworthy to viewers; and it enhances the audience's feeling of participation (Zhu and Jiang, 2018: 121).We submit that a similar logic may explain the growing use of livestreams over the summer and fall 2020.They are meant to create the feeling that China is fully transparent about what is going on in Xinjiang. 4 The subcategory "tourism and travel" makes up about one-third of the running time within the "culture" category, whereas "ethnic culture" constitutes the largest subcategory with 41 percent of running time.That ethnic culture is a focus is not surprising given how keen China's government is to show how well this is being preserved (cf.White paper L).A final subcategory to note is "urban culture" (11% of total running time and 10% of the total number of videos), which often shows the modern face of Xinjiang.The category "nature" accounts for some 13 percent of the total number of videos in our final sample and 4 percent of running time.Recalling Xi Jinping's admonition to present China's "beautiful scenery," scenery and wildlife videos make up virtually the entire subcategory, with ecological issues and natural catastrophes addressed in a tiny minority of videos.
The category "countering criticism" accounts for 13 percent of both videos and running time.This category takes various forms.There are frequent very short videos-often from official press conferences-in which Chinese government officials reject Western accusations as "unfounded," "fabricated," etc.These "rebuttals" account for some 10 percent of within-category running time.The opposite-positive statements on China's Xinjiang policy coming from authoritative non-Chinese figures-are labeled "endorsement" and make up 14 percent of within-category running time.The political leverage gained by having a foreign politician, journalist, or academic issue such an endorsement can be quite considerable, especially when that person comes from a Muslim-majority country (Videos 22, 23, and 24; see Appendix Table A2 for full details of these and subsequent videos cited by number).This is what Zhang Shilei (2019: 23) refers to as "'borrowing mouths' to speak" "借 嘴"说话.But by far the largest subcategory here is "myth busting," with 51 percent of within-category running time.These include shorter explainers of several minutes intended to show how Western media and politicians have gotten it wrong about Xinjiang.But they also include longer formats such as Headline Buster, a special edition of The Point with Liu Xin solely devoted to debunking specific news articles or foreign researchers such as Adrian Zenz.Finally, a quarter of the running time in this category goes to wider-ranging talk shows that touch on so many topics that we simply coded this subcategory as "talking politics." Our next-largest category is the "terror narrative" and accounts for 4 percent of videos but 9 percent of running time.As will be explained below, its political importance is larger than these shares suggest.This category includes some short videos such as a report about an anti-terror drill showing the preparedness of security forces or a story about a person "saved" from "extremism" through a spell at a reeducation facility.But we also find several long documentaries that will be discussed in more detail below.
Finally, there is the category of "Xinjiang policy amplifiers," accounting for both 5 percent of videos and running time.In terms of content, Xinjiang policy amplifiers are, by definition, redundant to other narratives, so we will not analyze them separately below.Regarding their function, they are similar to endorsements but come from Chinese sources.Often these are short videos or footage taken from Chinese news that repeat what an official has said or what a recently released white paper contained.While these videos are dull, they are arguably less so than the source material thanks to the inclusion of charts, animated graphs, and (usually dancing) ethnic minority individuals to stimulate the eye.Thus their purpose is to drive home the message and it is probably hoped that they will be shared more widely and reach a broader (or different) audience than policy announcements usually do.

Strategic Shifts
Before going into a deeper analysis of the content, this section explores how Chinese state media strategies have evolved over time-beyond the quantitative increase in Xinjiang-related videos already mentioned.For clarity, Table 2 outlines the total number of videos posted in the most important content categories for the period January 2019-December 2020. 5Whereas two recent books (Roberts, 2020;Tobin, 2020) argue that the "Global War on Terror" allowed China to legitimize its crackdown on Uyghurs through "securitization," we find that the terror narrative in purely quantitative terms is far from the most prominent aspect of Chinese state media legitimation attempts via YouTube.There are two spikes in this category to notice, in December 2019 and June 2020.As noted above, these coincide with relevant legislative action in the United States, and it is compelling to argue that Chinese state media tried to use the terror narrative to sway the opinion of policy makers and the international public to accept the PRC's Xinjiang policies as necessary to prevent further terrorist attacks.But even during these episodes of contention, the terror narrative was neither the only nor the predominant counterargument advanced by Chinese state media.Admittedly, however, this finding is only quantitative, and the qualitative analysis below reinforces the importance of the terror narrative.
What we see instead is a deluge of development-, culture-, and naturerelated videos.While these are the most prominent in every month except December 2019 and January 2020, their lead has grown by leaps and bounds toward the end of the sample period.Thus our tentative conclusion is that having failed at the attempt to get the terror narrative message across, Chinese state media changed tack and reverted to positive propaganda.This can be read as an attempt to present "other news" on Xinjiang to distract public scrutiny of China's human rights record in the region.Distraction and diversion are strategies regularly employed in domestic propaganda in the PRC as well (M.Roberts, 2018).Whether these will be successful internationally depends not just on quantity of output but even more so on the messages and how they are being conveyed.Thus our next section analyzes the major discourses of Xinjiang-related content on YouTube.

The Development Narrative
The vast majority of videos throughout all categories included in our sample contain an either implicit or explicit "development narrative" that portrays China as being on its way to a more prosperous, technologically advanced, scientifically rational, and ecologically progressive country.While this notion of linear progress and modernization is the baseline of most videos, the ones specifically included in this category tell stories of state-led infrastructure projects (such as rail-and highways, power plants, or industrial agriculture), poverty alleviation, employment and social welfare programs, and the advancement of the tourism industry.The narratives of these videos closely follow the discourses of government white papers (e.g., White paper K).Mostly, statistical figures are presented that show a tremendous increase (invariably read as an improvement) in a given metric, though often over long time periods.Typically, either the 1950s or late 1970s are taken as the point of departure for these long-term comparisons.
As a major innovation in presentation, personal development stories are told through portraits of seemingly ordinary citizens.These everyday stories differ substantially from depictions of larger-than-life propaganda heroes à la the Mao Zedong-era figure Lei Feng 雷锋, but also from earlier prominent Uyghur media figures that typically featured "distinct cultural characteristics" that distanced them from the Han Chinese (Chen, 2016: 116).This is exemplified by New China TV's Xinjiang-My Home series from summer 2020 (with 64 videos in the sample, usually slightly less than two minutes in length) and CGTN's A Rare Look into Xinjiang (with eight episodes in September 2020, each around four minutes).It is easy to see the family resemblance between these series and I Am from Xinjiang 我从新疆来, a sixteen-part sequel written by Uyghur author Kurbanjan Samat and produced by CCTV for a domestic audience in 2016.Because of its success, this was followed up by another series by Samat, I Am Going to Xinjiang 我到新疆 去, in 2018. Li Fuli (2019: 41) of Tarim University lauds these works for breaking the mold of the usual stereotypical media depictions of Uyghurs as either engaging in terrorism or singing and dancing. 6He lucidly criticizes these stereotypes as "othering" and "alienating" the Uyghurs, thereby preventing an integration with mainstream society.He concludes that successful depictions of Xinjiang would have to "dilute the taste of propaganda and wrap the message into a story" (Li, 2019: 42).These stories should use well-known tropes, such as fighting for one's dreams, overcoming obstacles, etc., to emotionally appeal to the audience.The emphasis should be on commonalities with mainstream Chinese society and everyday lives, not on the "mysteries" of Xinjiang, and Uyghur culture should be introduced in small doses only (Li, 2019: 42).While it would be far-fetched to claim that Li's article provided the blueprint for the New China TV and CGTN series mentioned above, he provides a perfect description of how they are designed.In deliberately "ouring"instead of "othering"-Uyghurs, the personal development stories promoted in 2020 differed markedly from earlier propaganda depictions of Uyghur individuals (Chen, 2016).
The development narrative itself is not a new phenomenon in Chinese official discourse (Tobin, 2020: 72-73), nor is it restricted to Xinjiang or other ethnic minority regions.It has been embedded in a Marxist-Leninist materialist view of history since the beginning of the PRC era but is based on an even older concept that Marius Meinhof (2017: 58) calls "colonial temporality."According to this concept, the experience of being (semi)-colonized by Western powers in the nineteenth century is seen as a source of national humiliation and yet has also given way to a developmental "reference point" that China has been aiming to catch up to (Meinhof, 2017: 65).The notion of China still being "in the middle of history," ever striving toward a better future, allows for both the "imagining of a different path yet to be taken and pride with respect to the steps already achieved" (Meinhof, 2017: 62).China has increasingly framed development as an "inalienable human right" and has argued that "there is no end either to development or to realizing the right to development" (White paper J), thus de facto institutionalizing the notion of being in the "middle of history" in state policy.
Of course, the path trodden in striving for development varied over different historical periods-at times embracing revolution and at other times following more technocratic methods (Meinhof, 2017: 60).An early example of the latter narrative, arguably a direct precursor to the contemporary YouTube development stories, can be found in the English-language China Today newsreels of the late 1950s and early 1960s.For instance, a short video from 1963 (in the comparatively calm interim period between the Great Leap Forward and the Cultural Revolution) introduces the transformation of Xinjiang's capital, Urumqi: a visual contrast to its impoverished past is established by juxtaposing scenes of the then clean and orderly city streets with undated stock footage of unpaved mud roads.The narrator describes Urumqi as now being marked by "prosperity and vitality" thanks to the development of industry, agriculture, and livestock breeding.Urumqi is host to a "booming market" stocked with goods attractive to local ethnic groups (then still translated as "nationalities") and the city is seeing a steady rise of consumer goods in department stores. 7 Today's CGTN and New China TV narratives follow a similar logic.The video "Roads in Xinjiang" (Video 32) introduces the region's major highway, bridges, and railway infrastructure, interspersed with archival pictures of earlier construction efforts and the promise that these projects will further improve people's livelihoods in the future.Many videos focus on poverty alleviation, which is linked to various forms of cash-crop agriculture (e.g., almonds [Video 62] or watermelons [Video 51]), the development of tourism (e.g., a lavish night market in Khotan [Video 21]), and state-led relocation and urbanization strategies, which focus on people's improved living conditions (Video 26).
What these narratives have in common is to portray the state as the sole actor capable of instigating these processes and positively transforming society.While individual people are frequently given a voice, they often narrate their stories through the lens of the state framework-generally the implementation of a policy is structured as the turning point toward the better.For instance, two brothers surnamed Ismael from Akto county used to barely make ends meet by working as herdsmen in the mountains, in an environment described by New China TV as "uninhabitable."Their situation only began to change in 2017 thanks to a local government poverty alleviation program, which trained the younger brother in planting melons.Seeing his brother's success, the older sibling decided to also give up herding and move to the melon plantation as well.He says that no matter how hard he worked as a herdsman, he could never earn more than 11,500 yuan (around US$1,720) a year-but his situation has now vastly improved through the training provided by the government (Video 51).Thus, while the brothers were undoubtedly hardworking in their old occupation, they lacked the abilities and foresight to engage in a more "modern" and lucrative trade.Only the benevolent help from the government was able to provide them with the skills necessary to channel their efforts into a more efficient and appropriate project-a logic also reflected in Xinjiang's internment camps being framed as "vocational schools." This and other similar narratives are in line with a general "discourse of gratitude" toward the state, which Emily T. Yeh (2013: 13) has identified in the Tibetan context and which can also be observed in Xinjiang.Development comes in the form of state programs, but it is also supposed to happen within the individuals themselves.New China TV's Xinjiang-My Home series provides the viewer with a variety of personal development stories in which the state's role is usually less visible than in the example above, but which show subjects who have already internalized the way to lead a "proper" life.All videos of this series focus on an individual who may have struggled in one way or another in the past, but who has had a transformative phase that led her or him to become a successful and satisfied citizen of Xinjiang.For instance, young urbanite Osmanjan describes himself as formerly having been used to "indulging himself in fun" by going to bars, hanging out with friends, and sharing his life online.However, he then noticed-what triggered this realization is not made apparent-that he was not making progress in his life.His social circle began to shrink, and boredom set in.He then became inspired by his parents' hard work in their self-owned restaurant and decided to change his life.He started helping out in the restaurant by barbecuing lamb skewers (a Xinjiang specialty) and waiting tables.Soon his life became more meaningful.He expresses that he hopes to make more effort to help his family and compares his progress with the "vast and vibrant development" of the city he lives in (as he says this, a bird's-eye perspective of urban high-rises is shown) (Video 67).Osmanjan never experienced the poverty the Ismael brothers had to face as herdsmen; he is depicted as living off his parents' income, being idle, and indulging in slightly improper activities like going to bars (which still look very tame to a Western viewer). 8 While Osmanjan's story does not include any direct hint at government policy, it equates the relationship between personal development and family with a city's development and the nation (cf.Hoffman, 2003).He comes to realize what he owes to his parents and adopts a more filial attitude and more diligent work ethic.The same kind of attitude, it could be argued, is expected from Xinjiang as a whole within the Chinese nation-after all, development is a "gift," and gifts demand reciprocity (Yeh, 2013: 233).
An important trope that is tied up in this narrative is women's liberation.Since its founding the CCP has claimed to work toward this goal, although this has often been mere rhetoric and gender equality has declined in the recent past (Zuo, 2016).But with respect to Muslim cultural and religious practice, even the most conservative politicians become defenders of women's rights.In China, as elsewhere, a primary goal is to "liberate" Muslim women from their headscarves (Leibold and Grose, 2016).The wearing of the burqa is presented as the epitome of "extremism" (Video 42).But the Chinese authorities do not stop there.Since 2011, Xinjiang's government has run Project Beauty 靓丽工程, which is complemented since 2017 by the Three News Campaign 三新活动 (referring to "new lifestyles," a "new atmosphere," and a "new order").In its wake, thousands of new beauty parlors have been established and even higher numbers of Uyghur women have been trained in cosmetology (Grose, 2019).The aforementioned March 2019 white paper (White paper N) on fighting terrorism explicitly mentions "cosmetology and hairdressing" as two of just eight skills to be taught at VET centers.Thus it should not come as a surprise that we find this trope also in the personal development stories presented by Chinese state media.One episode of Xinjiang-My Home portrays a divorced woman and mother who becomes the family breadwinner by opening her own beauty parlor (Video 52).Another is still working toward that goal, expressing her aspiration that "all of us will be pretty and stylish every day" (Video 71).One female makeup artist neatly sums up the philosophy behind this campaign: "As a woman, I have a right to be pretty!"(Video 25).
However, it would be unfair to suggest that beautification is all there is to the trope of women's liberation.For one, Uyghur males are also "liberated" of their customary beards-though this is not made explicit, one hardly sees a Uyghur man sporting a beard or moustache in any of the videos, even among the elderly.In addition, in many videos the more orthodox understanding that women's liberation could only be achieved by bringing them into productive work outside the home is clearly in the foreground.One formerly "self-abased" factory worker avers that "confidence is the basis of beauty" (Video 60), suggesting that only through working for her own money can she feel like a "proper" woman.Not being allowed or not wanting to work outside the home are presented as signs of "extremism" in several salvation stories-stories of Uyghurs brought back from the brink of "extremism" through "de-radicalization" in a reeducation facility (Videos 43,44,49,61,and 74).But these salvation stories are rare compared to the many videos that portray Uyghur women working in various occupations.We note an uptick in stories of women (and men) happily participating in labor migration schemes (Videos 50, 75, and 76) and village-based "anti-poverty workshops" (Videos 31, 35, and 68) after the accusations of coercive labor in Xinjiang became virulent.
But after China was criticized for forcing birth control, stories of female Uyghur homemakers were also published.One is Zulhumar, mother of a "sensible" son, whom she is seen picking up from school and taking to the supermarket (Video 64).Another example is Zulpiya, the twenty-four-yearold wife of a hardware store owner and mother of a son who describes herself as a "stay-at-home mom of the new era" 新时代的全时太太 9 as well asslightly unconvincingly-an "independent woman."She is seen doing housework, taking her son to the bookstore (speaking Mandarin only), and to various children's leisure activities.While he is in school, she goes out to drink tea, shop with friends, and dance (Video 69).Intriguingly, while the first images in her home suggest a humble background, her leisure life appears a bit too glamorous for the wife of a hardware store owner-moreover, this narrative stands in contrast to the "women's liberation" portraits above.Thus the personal development stories flexibly weave together a diverse set of characters to convey values that mainstream Chinese society can relate to, such as filial piety, hard work, education for self-improvement, and devotion to one's family (cf.Tobin, 2020).
Even more significant, from the perspective of how foreign-facing propaganda is changing, is the fact that this style of presenting the development narrative in a more convincing "micro-frame," instead of only as macro-level data, has begun to enter the official government white papers.The white paper "Employment and Labor Rights in Xinjiang" (White paper R), published by the SCIO in September 2020, by our count contains no fewer than seventeen personal development stories in miniature, on top of the usual statistics.It seems that different parts of the propaganda apparatus are eagerly learning from each other to better solve messaging problems.

Culture and History
Many of the personal development stories described above are set in an urban environment and already provide a glimpse at a modernized city culture not dissimilar to that of China's inland cities.In addition, videos that we categorized under "urban culture" show a Xinjiang that has already largely assimilated to secular and Han-dominated cultural forms (Kobi, 2018: 210-11).For instance, Uyghur comedian Memettursun Memeteli attempts to counter stereotypes of Xinjiang people riding horses to school and dancing in traditional style by making videos about youth cultures, anti-drug campaigns, or traffic police (Video 3).Another artist, "the new face of Xinjiang hip-hop," Ai Re 艾 热, 10 is locating his music as part of both international and (Han) Chinese culture.Singing in Mandarin and competing in televised national rap competitions, Ai Re claims to be frequently asked what elements in his music are particular to Xinjiang-to which he simply answers that him being a Xinjiang native is in itself such an element (Video 66).Another video portrays Gulzar, a young Xinjiang 11 woman who dresses in business clothing, likes going to bookstores, frequents cafes, and plays badminton.She claims that she had planned to move to larger cities in China's heartland 内地 in the past, but is now happy to stay in Xinjiang, as the region is now at a comparable level of development to the rest of the country (Video 59).In other words, these types of videos are trying to de-emphasize ethnicity and depict Xinjiang as part of both sinicized and globalized urban culture.Thus these videos' significance lies in showing the party-state's envisioned end goals of cultural assimilation within the logic of the Zhonghua minzu 中华民族 discourse (Tobin, 2020).The individuals depicted in them approach what Joanne Smith Finley (2013: 13) calls "would-be-urban 'cosmopolitans' . . .new world citizens," who are presented as an ideal to strive for.
Yet a different and much more prominent category of videos simultaneously portrays Xinjiang as a minority region and host to a cornucopia of ethnic cultures.Suppressed during the Cultural Revolution and long having been perceived as backwards and uncivilized, minority cultures became objects of an emerging fascination within China's search for new forms of cultural expression.Ethnic cultures became marketable and attractive to Han tourists, who perceived them in what Louisa Schein (1997) has described as "internal Orientalist" terms (for a critique of the qualifier "internal," see Anand, 2019).In fact, in the case of Xinjiang what characterizes the Chinese state media's depiction of ethnic culture is not just the colonial gaze on the ethnic "other" and an infatuation with the "mysteries" of the Silk Road or peculiarities of non-Han cultures, as critically noted by Li Fuli (2019).
While these are in evidence as well, 12 the more politically important point is that the display of ethnic diversity and multiculturalism constitutes a discourse legitimizing Xinjiang's incorporation into China.As the debate surrounding mummies discovered in the Tarim Basin in 1980 shows, archeological finds and anything relating to Xinjiang's history can easily be politicized (Millward, 2007: 15-17).The official Chinese line-that Xinjiang has been multiethnic, multicultural, and multireligious since ancient timesis a repudiation of Uyghur claims of indigeneity to the region (Bovingdon, 2010: 30-31).The 2009 "White Paper on Development and Progress in Xinjiang" (White paper B) declared that the Han were "one of the earlier peoples who inhabited Xinjiang," that is, before the Uyghurs.This assertion of long-term Han settlement of the region is also to be found in a two-part documentary on the Buddhist Kucha (Gaochang) kingdom (Videos 18 and 19).It is repeated in various government white papers; in comparison to the one from 2009, the 2018 white paper "Cultural Protection and Development in Xinjiang" (White paper L) is even more explicit and forthright, while "Historical Matters Concerning Xinjiang" (White paper O) extends the cultural contacts between China proper and Xinjiang backwards into prehistoric times.In one fell swoop these white papers declare all expressions of the ethnic cultures located in Xinjiang to be part of the larger "Chinese civilization" 中华文化-a concept that itself is Han-centric and perpetually situates other ethnic groups lower down in the hierarchy (Tobin, 2020).Marina Svensson and Christina Maags (2018: 19) show that the cultural "heritagization process" in China clearly serves political ends, and these are even more pronounced in the case of Xinjiang.While not all videos in our category "culture" contain the full gamut of this "authorized heritage discourse" (Svensson and Maags, 2018: 15), they constitute discourse fragments that, taken together, reproduce it in an audiovisual format.
This can be seen, for instance, in the many videos on ethnic music and dance, which appropriate these cultural artefacts as part of the wider "Chinese civilization."One prominent example is the set of musical suites known as the Twelve Muqam; the muqam, as Nathan Light (2008: 1-2) puts it, are at once "the most valued exhibitions of Uyghur cultural attainment" and a "highly politicized art form."In our sample there are both lengthy livestreams on the Twelve Muqam (Video 84) as well as shorter pieces (Videos 17,53,and 82) that introduce viewers to this gem of Uyghur music.Invariably the presentation foregrounds how this Uyghur musical form was performed at the Tang court, suggesting that it stems from a part of the imperial realm, while any religious connotations are ignored-a frequent feature of heritagization processes in China (Svensson and Maags, 2018: 19).The same can also be said with respect to the Uyghur mäshräp tradition that alongside the muqam has been turned into part of China's official intangible cultural heritage as recognized and protected by UNESCO (Harris, 2020).In short, Uyghur folklore, with its unique blend of various cultural influences and indigenous traits (Light, 2006), is being presented from a specific Chinese national perspective and thus is being appropriated by the state.
Apart from song and dance, food is one of the most common markers of ethnicity and place-based identity appearing within the range of the "ethnic culture" category.Some videos introduce colorful food markets (Videos 21 and 40), while travelogues cover local specialties or night market snacks (Video 83).There are videos showing lamb skewers (Video 78) or more unusual items such as chickpeas (Video 28) and grape leaves (Video 29).Most prominently, an almost tiresome number of videos introduce the traditional Uyghur naan (flatbread), presenting it as the flagship Xinjiang culinary specialty (e.g., Videos 5,9,and 79).Consuming local food is a popular pastime for (but not only for) Chinese tourists, and popular food documentaries like A Bite of China are altering people's perceptions of localities, inspiring tourist trips, and helping to shape place-based identities (Xu, Kim, and Reijnders, 2019).Food is a relatable and common way to depict a culture, and CGTN's and New China TV's narratives subtly place Xinjiang among the national discourse of regional cuisines and its overarching logic of "unity in diversity."At the same time, these relatable food specials are drawing attention away from other aspects of ethnic culture, which are less marketable and perceived as more problematic.
One such "problematic" element that is conspicuously missing from the depictions of ethnic culture is Islam, the predominant religion of the Uyghurs and nine other officially recognized minorities in China.While the danger of radical Islam is discussed within the terror narrative (see below), Islam has been almost fully eliminated from any "ethnic culture" depictions during the time span of our sample.However, CGTN videos from 2017 show Uyghur Muslims praying at the beginning of Ramadan (Video 1) and Hui Muslims (in Xinjiang) celebrating the Eid al-Fitr festival to mark the end of Ramadan (Video 2).One year later, in line with Xi Jinping's strategy to sinicize religions, such depictions of Islamic practices (in and beyond Xinjiang) were already no longer possible (see also Malzer, 2020).

Nature and Scenery
While the politicization of nature-related videos may not be quite as obvious as in the case of videos on culture, we argue that this genre has to be regarded as "alter-political."It provides a deliberately depoliticized vision of Xinjiang and is strictly in keeping with Xi Jinping's call to display a "beautiful China" to the world.Thus it feeds into the state-promoted discourse of "ecological civilization" (Marinelli, 2018).Accordingly, the most common type of videos are aerial views of gorgeous landscapes, showing lavender fields, apricot or almond trees in full bloom, colorful autumn leaves, etc., often set against mild Western piano music (e.g., Videos 11,12,and 16).This creates images of humans living in full harmony with their natural environment, enjoying the benefits of nature.In other cases, additional information is given on ecological development.Here, there is an overlap with the above-described development narrative (Videos 27 and 46).A noteworthy thematic emphasis is placed on showing an abundance of wetlands and water resources (Video 80), which can be read as an implicit refutation of allegations that intensive agricultural use of water resources had inflicted great harm on Xinjiang's fragile ecosystems.The Xinjiang Production and Construction Corps (XPCC)-targeted by US sanctions since 2020-in particular, is often identified as a culprit in this regard (Uyghur Human Rights Project, 2018: 30-33).In this context, the 2014 white paper "The History and Development of the Xinjiang Production and Construction Corps" (White paper D) explicitly emphasizes the XPCC's contributions to water conservancy.While the role of the XPCC is not discussed in any of the videos in our sample 13 -in itself a notable finding, given its strong position in the regional economy-water conservancy projects are on display in the sample (Video 63).
Occasionally, however, a different side of nature is shown that is more reminiscent of the Mao-era "humanity overcoming nature" trope (Shapiro, 2001).In these videos, viewers are presented with railway workers fighting against blistering cold and blizzards (Video 8) or local officials bringing food to herders whose livestock are on the verge of death because of an untimely early snowfall (Video 58).Moreover, antidesertification efforts are a special feature of videos in this vein (e.g., Videos 14 and 15).
A handful of videos show natural disasters occurring in Xinjiang.However, the short items on earthquakes (Videos 6 and 47) or an avalanche (Video 10) do not look particularly dangerous: images show next to no damage to or in buildings (a fallen beer bottle seems to be the only casualty in one video [Video 47]), and the narration reports hardly any cases of injuries.While natural disasters are largely beyond human control, the disasters in these videos appear no more than practice runs for Xinjiang's rescue teams, without constituting any real danger.
Lastly, some videos show rare animals spotted in the wild, such as snow leopards (Videos 4 and 56), Tibetan wild asses (Video 55), or salamanders (Video 7), emphasizing successful ecological protection and improved natural resources in Xinjiang.Some animal-related videos can even be interpreted through a subtle political lens: one item shows a wounded wild deer being rescued by Xinjiang police (Video 54)-the police are thus depicted as benevolent helpers of the weak.Another shows a breeding station for rare wild horses, which are said to be too genetically close to interbreed and therefore need partners from abroad (Video 13)-which, one might argue, could be understood as a subtle call for interethnic marriages.For instance, in the opening banter during an episode of Headline Buster, host Wang Guan 王冠 claims to be genetically "16 percent Mongolian" and concludes that "we're all mixed" (Video 70), while reporter Liu Yina 刘依娜, visiting a Xinjiang village, postulates that "probably my origins actually started from here a very long time ago" (Video 81).With everyone allegedly belonging to a Zhonghua community of common ancestry (cf. the aforementioned 2009 white paper on ethnic policies [White paper B]), origins, genes, and territory become conflated into a unified entity, which by definition appears to be inseparable.This interpretation chimes with recent state promotion of interethnic marriages in Xinjiang-especially between Han men and Uyghur women (Byler, 2019;Dayimu, 2020).

The Terror Narrative and Countering Criticism
Our analysis above found that the terror narrative is not the predominant legitimation strategy of Chinese state media on YouTube in terms of quantity and that it was foregrounded only during two episodes of contention.That said, alleged terrorism and religious extremism as its root cause are at the heart of the Xinjiang issue from the perspective of the Chinese government, as its white papers repeatedly make clear.It has been pointed out before that China, against the backdrop of the US-led "Global War on Terror" after 9/11, has reframed its Xinjiang troubles for the purpose of international consumption from a focus on "separatism" to one on religious "extremism" and "terrorism" (Clarke, 2018; S. Roberts, 2018Roberts, , 2020)).This much is also evident from the government's white papers: Of course, China condemns each of the so-called "three evil forces" 三股势力, namely terrorism, separatism, and extremism, but the emphasis has clearly shifted.In the 2002 white paper "'East Turkestan' Terrorist Forces Cannot Get Away with Impunity" (White paper A) the term "terrorism" or its derivatives are used 130 times, while "separatism" and "extremism" appear six and four times, respectively.In the 2019 white paper "The Fights against Terrorism and Extremism and Human Rights Protection in Xinjiang" (White paper N), these terms are mentioned 174, 22, and 85 times, respectively.Taken at face value, this corroborates Sean R. Roberts's (2018Roberts's ( , 2020) ) contention that separatism has taken a back seat while extremism is foregrounded in official explanations of terrorism.
However, the relationship between the three concepts is not so straightforward as this crude measure would suggest.In fact, there seems to be some confusion even among the drafters of these white papers as to their exact relationship.The March 2019 white paper "The Fights against Terrorism and Extremism and Human Rights Protection in Xinjiang" states that "separatism is the hotbed in which terrorism and extremism take root in Xinjiang," suggesting that separatism is the basis for the other two "evil forces."In contrast, the September 2019 white paper "Seeking Happiness for the People: 70 Years of Progress on Human Rights in China" (White paper Q) seems to reverse this causal relationship by saying that "religious extremism, which constitutes the ideological base of ethnic separatism and terrorism, is by nature antihuman, anti-society, anti-civilization, and anti-religion" (emphasis added).While the relationships between these three concepts certainly are complex, very likely interactive, and, at least in our view, can exist independently (e.g., there are cases of separatism without extremism), it is still remarkable that Chinese official positions on this seem not to be entirely fixed.In our sample videos, we also find experts invited as studio guests expressing slightly differing opinions as to their relationships (Video 38).
The core of the terror narrative during the two episodes of contention identified above are three CGTN documentaries that show previously unreleased surveillance camera footage of attacks and other details as evidence of terrorism motivated by religious extremism.The first, the fifty-minute documentary video "Fighting Terrorism in Xinjiang" (Video 30) was released with great fanfare at the height of US congressional debates on the Uyghur Human Rights Policy Act of 2019.This was followed a week later by the half-hour video "The Black Hand-ETIM and Terrorism in Xinjiang" (Video 36).The third documentary video of this kind, "Tianshan: Still Standing-Memories of Fighting Terrorism in Xinjiang" (Video 57), just short of an hour in length, was released right before President Trump's signing of the final version of this act.Each release was accompanied by several amplifiers-such as shorter videos taken from the longer version as "teasers," talk shows debating the significance of the documentaries, etc.
Despite this sophisticated publicity, these documentaries initially did not live up to the expectations of their producers.From the subsequent reactions, it is obvious that the makers of the documentaries had banked on their being a game changer in the debate on Xinjiang.Apparently, their understanding of the issue was that policymakers and publics in the West lacked a real appreciation of the security challenges faced by China.Therefore, the reasoning continued, if China revealed how serious the issue was, there would be more sympathy with its "counterterrorism" and "de-extremification" measures.Showing the-often shockingly brutal-surveillance footage of terrorist attacks and violent riots happening in China was certainly a risk.Even the Chinese domestic audience had not been shown these scenes, which was justified as an attempt to prevent ethnic discrimination and hatred against Uyghurs.Yet the risk did not pay off.Instead, the first two documentaries were widely ignored in the West, much to the chagrin of Chinese newsmakers.The Global Times carried a blistering commentary on Western media's "selective deafness" (Shi, 2019), blaming them for ignoring an "inconvenient truth."The spokesperson for the Chinese Ministry of Foreign Affairs, Hua Chunying 华春莹, put international journalists on the spot during her regular press briefing by asking who among them had watched the documentary.As not a single reporter raised their hands, she scolded them like schoolchildren who had neglected their homework.Footage from this scene was then used repeatedly to criticize the lack of uptake by "biased" Western media (Videos 33,34,and 37).The Point with Liu Xin produced a whole show discussing the issue in which anchor Liu Xin was visibly rankled by her Western colleagues' intransigence (Video 38).When some international media eventually commented on the documentaries, they were quick to dismiss them as distraction tactics (Buckley and Ramzy, 2019).
This failure to change the international discourse on Xinjiang became a case study for media analysts of the challenges faced by Chinese media in their attempt to break up the perceived "Western media monopoly" (Li, 2020).One of the problems identified was that YouTube did not even notify those users subscribed to CGTN's channel of the new video, which led to speculation that the platform was deliberately restricting its circulation (Shi, 2019;Li, 2020).The longer 2019 CGTN terror documentary was even removed at one point because of alleged copyright infringements and had to be reuploaded (Video 39).This is compounded by the fact that YouTube has deemed the footage so disturbing that it requires users to register with proof of age before they can watch it.This obviously complicates wider dissemination and defeats CGTN's purpose. 14The third terror documentary of June 2020, therefore, changed tack and emphasized the personal suffering induced by terrorism through individual perspectives amplified by a series of narrative and cinematographic innovations (Zhang and Zheng, 2020).This is not the first time that China's state media have failed to garner international support through a documentary on Xinjiang-related terrorism.The first such attempt, building on the aforementioned January 2002 white paper (White paper A), was undertaken in July that year with simultaneous airing of "On the Spot Report: The Crimes of Eastern Turkestan Terrorist Forces" via CCTV-4 (in Chinese) and CCTV-9 (in English).While we do not have access to the original footage in this case, the seasoned China scholar and Xinjiang expert Yitzhak Shichor (2006) provides a detailed account of its content.His interpretation accompanying the summary is, however, deeply skeptical regarding whether the violent episodes recounted here could actually be construed as terrorism and whether they all stem from separatism or extremism.This reaction to China's terror narrative is fairly typical among international observers (Steele and Kuo, 2007;Clarke, 2018;Roberts, 2020).It is important to understand how deep-seated the interpretation is that all violence ultimately is triggered by repressive policies coming from Beijing.In the case of Tibetan protests, in particular self-immolations, Robert Barnett (2012) has characterized this as a "policy-response" approach, contrasting it with the "outside instigation" explanation offered by China officially, and, in conclusion, found both wanting-though for different reasons and to varying degrees.
It is against this longstanding "policy-response" interpretation that the Chinese government in its white papers and via foreign-facing state media has attempted to recast the violent attacks by Uyghurs not as being directed against the Chinese state in particular but as terrorist acts targeting humanity as a whole.Thus "extremism" is always presented as a "global threat" and linkages between the East Turkestan Islamic Movement (ETIM) and al-Qaeda or the Islamic State of Iraq and the Levant (ISIS) are stressed. 15This ties in neatly with China's attempts to present itself as working toward the greater good of humankind on all kinds of issues in global governance (Zhang, 2018;Yang, 2020).Two sources of such "extremism" are particularly highlighted.First, online videos-whether sought out by the "radicalized" individuals themselves or deliberately fed to them by middlemen-are at numerous points referred to as having contributed to converting susceptible youths to terrorism (Videos 30 and 36; cf.Harris and Isa, 2019).This strengthens the "outside instigation" thesis, in that those committing terrorist attacks are shown (often in interview excerpts with convicted prisoners) as having been manipulated to the point of becoming remote-controlled instruments in the hands of outside forces.This shifts the blame for violent resistance away from China's Xinjiang policies, pointing instead toward "global jihadists" as a common enemy of the civilized world.In addition, this line of argument taps into debates on "homegrown terrorists" in countries around the world who "self-radicalize" via the internet, and thereby gains persuasive power.
Second, and more importantly, a lack of education or even an active refusal to be educated or send one's children to school is repeatedly stressed (e.g., Videos 45 and 48).Like the August 2019 white paper "Vocational Education and Training in Xinjiang" (White paper P) and the September 2020 white paper "Employment and Labor Rights in Xinjiang" (White paper R), many studio guests argue that the "right to education" is being infringed on in ethnic minority communities in Xinjiang.Thus they reposition the "reeducation" campaign as a benevolent and necessary government intervention that protects human rights rather than violating them.Invariably, "education" is narrowly defined as learning Mandarin Chinese (the "national language"), understanding the law, and acquiring job skillsexactly what the VET centers are supposed to provide.Conversely, a lack in any of these fields is now being interpreted as a direct consequence of being impacted by "extremist thinking" (White paper N; Video 41).A frequently used argument in this context is what we call the "bombs versus camps" comparison: while China is said to react to extremism by providing (re)education, the United States is accused of dropping bombs on any country that threatens it because of extremist ideology.Of course, this argument is also an example of "whataboutism," a generally popular strategy on Chinese state media.Yet, as comedian John Oliver points out, while some of these remarks may be "fair hits," "it's also completely possible for two things to be wrong at the same time." 16But for Chinese state media it allows the use of the "double standard" accusation, which is also commonly marshalled in the white papers.
Most significantly, we argue that a new twist in China's official human rights discourse is underway here.Previously its crucial point was to stress the "right to development" as most fundamental, such that other rightsincluding the political and civil rights championed by the West-had to take a back seat.Terrorism was deemed an attack on human rights exactly because it prevented development (White paper C).Now, however, another "most fundamental" human right is propagated, namely the right to safety.As long as this is threatened, not even the right to development can be realized, let alone any of the other human rights (White paper N).Thus, instead of having one precondition before political rights take priority (development first), there are now two preconditions that need to be met (human safety first, development second).It is easy to see how this argument can be extended from a terrorist threat to a health emergency like the COVID-19 pandemic.It is an important strategy of Chinese state media that this reasoning is often left to be outlined by experts such as US "current affairs commentator" Einar Tangen.Following the "borrowing mouths to speak" logic, this gives the argument added weight.At any rate, the sequential reading of human rights in official Chinese discourse is at odds with the debate in liberal democracies that centers on the parallel realization of different human rights.Instead of asking which comes first, in the latter case the debate is over how to balance the different human rights appropriately.
China is working hard diplomatically and, as we have shown, on the media front to make inroads with its own human rights understanding (Yang, 2020: 11-12).

Conclusion
This is the first study to comprehensively and systematically analyze the Chinese state media's handling of the current Xinjiang human rights crisis.A previous report by the Uyghur Human Rights Project (2020), though insightful, in our view focuses too narrowly on what we call "rebuttals" and "myth busting" (the report uses the term "hostage propaganda").Thus it fails to consider the variegated counternarratives constructed by Chinese propagandists.Admittedly, some of these trends have only become more pronounced after the Uyghur Human Rights Project report was published.But as our quantitative analysis demonstrates, Chinese state media use YouTube to methodically disseminate mostly positive depictions of Xinjiang, while directly countering criticism has been the primary mode only during heightened episodes of contention.Since July 2020 in particular, they scaled back the terror narrative and the showing of atrocities committed in China and tremendously expanded the development narrative as well as culture and nature broadcasts.Thus Chinese state media are more productive in spinning their own yarns than refuting accusations directly.
Our qualitative analysis delves more deeply into these narratives.It finds that the media messages and forms of delivery have become more variegated and sophisticated over time.Within the development narrative we highlight the recently introduced genre of personal development stories as a major tactical innovation to increase persuasive power.This has even spilled over into the usually staid and statistics-laden white papers themselves.We also stress how different tropes, such as women's liberation, expected to hold sway among international audiences, are woven into the fabric of the development narrative.The culture-and history-related narratives clearly serve political ends too, with their emphasis on long-term territorial integrity, multiethnicity, multireligiosity, and multiculturalism in Xinjiang.In addition, they present a glimpse of what the Chinese state envisions as goal of its current Xinjiang policies: a reformed, secular, modern lifestyle and a Uyghur identity centered on the Chinese nation-state in which ethnic difference is only expressed in nonthreatening (or nonprofane) ways, such as different food preferences.This is indeed "social re-engineering of Uyghur identity" (Smith Finley, 2019: 10), as critics allege, however benign may be the outcome presented in official propaganda videos.
Nature-related videos take up a considerable share of our sample, even if we exclude the extremely long livestreams.While it is tempting to dismiss them, as previous studies have done (Uyghur Human Rights Project, 2020), we argue that these broadcasts have an alter-political motivation: they present a deliberately depoliticized vision of Xinjiang, not just to boost tourism and development but also to change the subject of the conversation.Though subtly hidden, there are even more directly political allusions embedded in some of these videos.
The most overtly political narrative is the one on terrorism, which is central to China's understanding of the Xinjiang issue and its foreign-facing propaganda to legitimize its response.Though the temporal pattern in this regard is one of waxing and waning, we ultimately interpret this as tactical adjustments in the face of limited success.In qualitative terms, the terror narrative remains central to China's legitimation efforts, since these documentaries provide a graphically violent contrast to the harmony usually depicted in state media.This follows the logic of "letting facts speak" 用事实说话 to reduce barriers against accepting China's message (Liu 2021: 84).China appeals to global solidarity in fighting "terrorism, separatism, and extremism" by presenting itself as a victim like many other countries.This deflects blame from the Chinese government and its policies and attributes it to outside forces.China's "counterterrorism" and "de-radicalization" drive is presented as part of a global effort that benefits humanity.This pattern of argumentation is analogous to that in many other issue areas, such as poverty, public health, ecology, economic development, or demographics.Moreover, we find that an evolution of China's official human rights discourse is underway.Instead of prioritizing the "right to development" over all others, another even more fundamental right, the "right to safety" is now being propagated as a precondition to development.In our pandemic times, we expect this argument to gain traction not just within China but also abroad.
Our study contributes to a growing body of research on China's international media strategy and attempts to gain "discourse power."With an issuespecific in-depth examination, we provide a case study that can be fruitfully compared with others, for example, China's COVID-19 response or environmental policies.It demonstrates how Chinese state media use indirect as well as more direct methods to counter criticism, create a counternarrative, and manage its international image.Future research can build on this to further elucidate the effects of these strategies.12.An illustrative example is a short piece in which a young CGTN reporter sits next to an elderly player of the Kazakh dombra string instrument and explains (or "Han-splains") in flowery words how the dombra reflects Kazakh traditional lifestyle.Without the man ever speaking himself, she concludes: "For Rahemjan, this is what brings him back to the grassland, back to his childhood, and back to the root of the Kazakh culture" (Video 65).13.The closest we could find is a piece on a veteran soldier who took part in the construction of the Duku Highway, losing many of his comrades in the process (Video 77).In addition, there is a livestream on the history of oil production in Karamay since the 1950s (Video 72) and a short clip on a Karamay restaurant tapping into nostalgia for the 1950s and 1960s (Video 73).The restaurant owner proclaims that by displaying items from this past he wants to remind younger generations of the sacrifices of their forebears.This fits perfectly with the "XPCC spirit" analyzed by Neddermann (2010).However, none of these videos explicitly mention the XPCC.Note that this would change with Xi Jinping's July 2022 visit to Xinjiang, which prominently featured in his trip to Shihezi 石河子 the XPCC's headquarters and the site of the Army Reclamation Museum of the XPCC.14.Though Chinese state media have not commented on this, it would be fair to point out that equally harrowing scenes of Tibetan self-immolators can be watched on YouTube without age restriction.15.Intriguingly, though, none of the three major terror documentaries makes mention of alleged plots to carry out attacks outside China, although these are taken seriously by Western observers (e.g., Pantucci, 2018: 160-63;Hastings, 2019) and could have provided additional legitimation for the claim that terrorism is a common enemy.16. "China & Uighurs: Last Week Tonight with John Oliver," YouTube video, 18:30-18:36, July 27, 2020, https://www.youtube.com/watch?v=17oCQakzIl8.
Note: Excludes duplicate videos and livestreams with a running time of longer than one hour; n = 711.Source: Authors' data.

Table 2 .
Number of Videos by Category over Time in the Sample, 2019-2020.
Note:Table focuses on the most relevant categories and on the period 2019-2020; n = 597.Source: Authors' data.

Table A1 .
Chinese Government White Papers.

Table A2 .
Videos Cited in the Main Text.Fighting Terrorism in Xinjiang