Official Attitudes Toward Post-mao Chinese Film

Evaluating Official Attitudes Toward Post-Mao Chinese Film Through a Quantitative Lens

Ian Lamont August 14, 2006 [email protected] (For Graduate Credit)

Prof. Charles Hayford Visiting Scholar, History, Northwestern University Harvard Summer School Film And History Postwar Japan and Post-Mao China (History S-1855)

What does China’s central government think of Chinese film? Attempting to answer this question in the post-Mao era usually involves qualitative evaluations of two types of sources: Statements by Chinese officials and agencies, or observations of the actions of Chinese officials and agencies. Paul G. Pickowicz took this approach when evaluating Chinese film of the post-Mao era,1 as did Jianying Zha in her comparisons of “Fifth Generation” Chinese filmmakers Zhang Yimou and Chen Kaige.2 However, I believe quantitative methodologies can also be used to explain the evolution of official attitudes toward Chinese film and Chinese filmmakers. In this paper, I will present the results of a quantitative content analysis based on news published by the New China News Agency, the state-run wire service of the People’s Republic of China. Using the NCNA news items as a gauge of official policy, my analysis will explore the state’s attitude toward Chinese film in general, as well as toward specific directors, in the context of social and political trends taking place during the five years before and after the June 1989 Tiananmen incident. I will also compare the NCNA data to some of Pickowicz’ and Zha’s qualitative research.

Background In the 1980s and early 1990s, the media gatekeepers of Chinese film — that is, the individuals and organizations responsible for filtering film content before audience consumption — were mostly associated with the government. According to Pickowicz, they included regional studios, the Film Bureau within the Ministry of Culture (and, after 1

Paul Pickowicz, “Velvet Prisons and the Political Economy of Chinese Filmmaking,” in Deborah Davis et al, ed., Urban Spaces in Contemporary China (Cambridge: Cambridge University Press, 1995) 2 Jianying Zha, China Pop: How Soap Operas, Tabloids, and Bestsellers Are Transforming a Culture (New York: The New York Press, 1995)

1

the late 1980s, the Ministry of Radio, Film, and Television), the China Film Distribution Corporation, the state-controlled Chinese Filmworkers Association, state-funded film publications, and even the government organs charged with running film awards programs.3 Yingchi Chu identifies additional bodies that issued documents concerning film policy: The Chinese Communist Party Central Committee, the State Council, the Ministry of Propaganda, the General State Administration, specially constituted “leading groups,” and party leaders.4 These organizations and individuals could make or break a film project at any stage of production, from vetting script ideas to regulating distribution in domestic movie theaters. Pickowicz proposes that Chinese filmmakers in fact worked in a “velvet prison” of state control in the post-Mao era. The velvet prison theory was originally described by Hungarian author Miklos Haraszti in the early 1980s to explain post-Stalinist control of artists in Communist countries. Haraszti said it was more effective to give artists special perks and a little power, as long as they were actively or passively loyal to the state and were not outright anti-Marxist.5 However, Haraszti described an additional aspect of this system: the option for the state to briefly revert to Stalinist methods of control should the need arise. And this, suggests Pickowicz, is one model for evaluating official treatment of Chinese film following the Tiananmen square demonstrations in 1989. In the 1980s, “entertainment” films featuring romance and violence were typical fare, because the Film Distribution Corporation believed these genres were popular among urban viewers. The 3

Pickowicz, p. 207 Yingchi Chu, “The Consumption of Cinema in Contemporary China,” in Stephanie Hemelryk Donald et al, ed., Media in China: Consumption, Content, and Crisis (London: RoutledgeCurzon, 2002), p. 43 5 Pickowicz, p. 195, citing Miklos Haraszti, The Velvet Prison: Artists Under State Socialism (New York: Noonday Press, 1989) 4

2

corporation did not really care if the studios managed to get an “art” film past the censors; the costs for producing such films were borne by the studios and the corporation could make money off of international distribution rights (which it controlled) while limiting domestic distribution to a handful of prints.6 However, following the 1989 unrest, the state reverted to a “crude Stalinist mode of control.” Existing films with controversial content were banned, while the Film Bureau “ordered the studios to produce a small flood of films on contemporary and historical topics that praised the army, the party, and the police, while condemning the polluted ways of foreign cultures.”7 There were variations in the way certain directors operated in this environment, say Pickowicz and others. Zhang Yimou’s Ju Dou (1990) and Raise The Red Lantern (1991) were initially banned in China. Pickowicz disputes that dissident politics were the reason. Rather, he says, the bans related to bureaucratic issues as well as the films’ “reverse Orientalism,” which promoted the exotic — and often ugly — aspects of China to appeal to foreign tastes.8 However, Zha suggests criticism of Fifth Generation films came mainly from intellectual circles.9 Whatever the causes, the domestic bans on these films were lifted in 1992, and Zhang even won some Chinese film awards the following year. Pickowicz says that Xie Jin, a much older Chinese director whose career experienced a comeback after the Cultural Revolution, had a far different experience than Zhang. Calling Xie a “thoroughly establishment figure,” Pickowicz says Xie is a director who is “able to work comfortably in both Stalinist and velvet prison environments and

6

Pickowicz, p. 210 Pickowicz, p. 211 8 Pickowicz, p. 213 9 Zha, p. 94 7

3

whose amazing popular films must be regarded as ‘official culture.’” Furthermore, while Xie’s films express popular discontent with the Communist system, “Much of what he does legitimizes and perpetuates the dominance of state socialism.”10

Methodology Pickowicz and Zha base most of their conclusions on their own appraisals of Chinese films, the actions of official agencies responsible for creating film policy and/or film distribution, books, journal articles, and other primary sources. Zha was also able to conduct interviews with Chinese film workers, including Chen Kaige. Can a quantitative analysis give insights into the issues Pickowicz and Zha explore, or test their qualitative findings? I believe the answer is yes, for certain issues and certain findings. Ideally, my quantitative methodology would be based on documents created by official agencies, or articles produced by state-controlled film publications. These documents are written by knowledgeable people within the state system and reflect official film policy. When subjected to computer-assisted analysis in aggregate or according to a sampling technique, patterns would allow us to evaluate official policy and existing qualitative research. Unfortunately, this ideal approach is impractical for several reasons. First, I am unable to read Chinese. Second, even if I could, it would be very difficult to locate appropriate documents from the Ministry of Culture and other state-run organizations. Third, most documents from the pre-Internet period are not in digital format, which is required for the software I am using (Microsoft Excel and Yoshikoder). I could transcribe

10

Pickowicz, p. 217

4

the documents or use a quantitative methodology based on human coding, but these two options would take too much time. I have therefore used content from the English-language news wire of the New China News Agency upon which to base my research. Besides being written in my native language, there is a complete digital archive of NCNA news items11 going back to January 1, 1977, accessible from the LexisNexis online database. The NCNA is the official news agency of the central government, “authorized to issue communiqués, statements and important news on foreign affairs and to provide domestic and international news for newspapers and broadcasting stations across the country.” The aim of the English service has an additional mission: to let foreigners and foreign governments better understand China’s people, policies, and cultural life. While the NCNA service does not necessarily reflect the views of official groups within the state film industry, during the Deng Xiaoping era it reported to the Chinese Communist Party Department of Propaganda, which in turn reported to the Chinese Communist Party Central Committee.12 These bodies and the film organizations ultimately serve the same masters, and would presumably be on the same page when it came to outlining official views about domestically produced films and filmmakers. My content analysis has several stages. In the first stage, I performed frequency counts on terms associated with content variables using the results of LexisNexis searches entered into an Excel spreadsheet. For instance, a news item associated with the “Film” variable (or “F”) contained the terms “film” or “movie” in the full text of the article (LexisNexis automatically counts plural forms as well). In 1984, I counted 348 F 11

Articles, official statements, news briefs, and captions Won Ho Chang, Mass Media in China: The History and the Future (Ames: Iowa State University Press, 1989), p. 195. 12

5

items in English NCNA news items. The following year, there were 404 F items, but as a percentage of all NCNA items, there was actually a decline from 1.27% in 1984 to 1.13% in 1985.13 In certain cases, I performed searches of multiple variables — e.g., “Zhang Yimou” (Z) items and “International” (I), in order to determine what percent of NCNA items about Zhang Yimou were associated with international film festivals or awards in a given year. In the Results section of this paper, there is a description of the issues I analyzed with frequency counts. The Appendix contains a full description of this methodology, including the terms that I chose to represent each content variable. In the second stage of my content analysis, I used a software tool called Yoshikoder to count the frequency of positive and negative terms in NCNA news items about Zhang Yimou and Xie Jin over time. The aim was to analyze official attitudes toward Xie and Zhang, as well as evaluate test Pickowicz’ conclusions about the two directors. The results are described in Issue 4 of the Results section, and more information about the methodology is described in the Appendix. It must be noted that the quantitative methodologies I use are a far from perfect gauge of official government attitudes toward Chinese film and filmmakers. General problems with computer-assisted text analysis include researcher bias at the design and interpretation stages; poorly designed studies; inconsistencies in the source documentation, which range from misspellings to changes in writing styles and terminologies (e.g., Mao Tse-tung was used in NCNA stories in the late 1970s, but the pinyin spelling — Mao Zedong — was used in the 1980s and later decades); ambiguous terms, or words with multiple meanings; and a general inability to discern subtleties in

13

i.e., 27,461 NCNA items divided by 348 F items vs. 35,789 NCNA items divided by 404 F items

6

language — the use of metaphors, attempts by writers to have readers “read between the lines,” etc. Specific problems I encountered were numerous. They included a small number of NCNA news items for certain content variables. For the Yoshikoder analysis, I was limited by a small number of news articles to sample from, as well as a dictionary that was not tailored to film studies or NCNA content — I relied on the basic positive and negative categories within the General Inquiry dictionary, which was originally developed for the study of political texts. Lastly, there was a quirk in NCNA content that ruined most of the data relating to certain content variables in 1990 and 1991. This was caused by a daily “What’s on in Beijing” arts and culture listing, which was apparently aimed at local expatriates and tourists visiting the capital. The column ran almost every day in 1990 and for the first four months of 1991, ending on May 5, 1991. It caused a spike in references to some, but not all, variables under study during this period. They included the F (film) and O (opera) variables.

Results In the following section I will detail the issues that were subjected to quantitative analyses. Issue 1: NCNA coverage of Chinese film The first issue I attempted to gauge was how the government — through the barometer of state news agency coverage — viewed Chinese film. My methodology did not allow for me to discern official policy per se, but it did let me measure the amount of NCNA coverage devoted to Chinese film as a percent of all film coverage. I also

7

searched NCNA news items for groups of opera- and Chinese opera-related terms, to provide a basis of comparison with the film and Chinese film variables. The results were intriguing. Neither film (F) nor opera (O) received much attention from the NCNA from 1984 to 1993 relative to the entire census of NCNA news items. Film references were three to five times more common than opera references, but the gap narrowed in the early 1990s, as F references declined sharply and opera references declined and then leveled out (see chart, “NCNA Film vs. Opera Items, 1984-1993”).14

The gap between Chinese film (CF) news items and Chinese opera (CO) news items was narrower still. The total number of Chinese film items dominated from 1984-1989, but these levels dropped by about half in 1991, 1992, and 1993 (In 1990 there was a spike of Chinese film items, but this is related to the daily “What’s on in Beijing” arts and 14

An article that contains references to both film and opera will count as a single reference for each category. This convention applies to all comparative frequency counts described in this paper.

8

culture listing, which sometimes contained references to Chinese film.) The number of Chinese opera items rose post-1989, beating out Chinese film for four consecutive years. This was partially the result of the “What’s on in Beijing” arts and culture listing, but even in 1992 and 1993 the number of Chinese opera items was notably higher than in 1989 (see chart, “Chinese Film vs. Chinese Opera Items, 1984-1993”).

When the Chinese film and Chinese opera items are charted as a percent of all film and opera items from 1984 to 1993, Chinese opera comes out firmly on top. Chinese opera was consistently above 50% of all opera-related items.15 Chinese film shows a gradual decline, starting out at 24% of all film items in 1984, and dropping to 13% by 1993 (see chart, “NCNA Chinese Film and Chinese Opera Coverage, 1984-1993”)

15

1988 is the exception, with 47%. There was also a spike to 80% or more in 1990 and 1991, but this was caused by the “What’s on in Beijing” arts listing, noted previously.

9

I believe the prominence of Chinese opera in NCNA coverage had much to do with the increased domestic emphasis on established Chinese art forms (i.e., Chinese opera) at the expense of foreign-influenced art (i.e., Chinese film). Furthermore, in the postTiananmen geopolitical landscape, in which the Chinese government was scorned for its actions in 1989, there was also an increased emphasis in NCNA on innocent “cultural exchanges,” which often involved tours of Beijing opera troupes to foreign countries. It must be noted that the relative lack of non-Chinese opera items when compared with foreign film in NCNA coverage can partially be attributed to entertainment preferences among NCNA journalists. Many urban Chinese during this period had seen foreign movies on television or even in theaters.16 But live Italian, Russian, or German

16

Hollywood productions weren’t regularly shown in domestic cinemas until 1995, but movies from eastern Europe, India, Hong Kong and elsewhere were sometimes screened.

10

opera was a rarity, probably owing to logistical issues, high production costs, and expensive tickets. Issue 2: Chinese film and “reverse Orientalism” Much has been written about the tendency of Fifth Generation Chinese filmmakers to receive accolades abroad while being ignored at home. This is a simplistic assessment, but Both Pickowicz and Zha point to the early 1990s as being particularly rough for directors. During this time, certain Fifth Generation films were celebrated at foreign film festivals while being banned or heavily edited at home. Pickowicz laid out the reasons behind the domestic freeze as follows: “[This was] more to do with what might be called [Fifth Generation directors’] ‘reverse’ Orientalism. In a word, these trendy new works, funded by foreign sources and made primarily for foreign audiences, reveal the exotic and erotic Chinese world that foreigners like to see rather than a Chinese world that is recognizable to the Chinese themselves.”17 If this is the case, the NCNA, as the official press mouthpiece of the government, should logically de-emphasize Chinese film during this period. As I have demonstrated earlier, this is exactly what happened, at least in 1992 and 1993 (see chart above, “Chinese Film vs. Chinese Opera Items, 1984-1993”). However, there are several problems with this logic. First, not all Chinese films are made by Fifth Generation directors. Chu notes a rush of “mainstream melody films” (i.e., films that support the government’s social agendas) around this time, such as Founding Ceremony of the People’s Republic (1989) and Zhou Enlai (1991). As these types of films toe the government line, they should not be subjected to the same silent treatment in 17

Pickowicz, p. 213

11

the NCNA news wire as Fifth Generation films and directors. A second problem: Not all NCNA news items that discuss Chinese film mention specific directors or films from either camp, but rather discuss the general state of the domestic film industry. Nevertheless, I think it is possible to reconcile these issues, even though I did not run searches on all Fifth Generation filmmakers and their “mainstream melody” counterparts, or separate the general film industry articles from reviews or news about specific filmmakers and films. First, if the level of so-called mainstream melody film items remained the same or even rose slightly, and Fifth Generation news items as well as general industry items declined, overall references to Chinese film should still decline. This is reflected in the NCNA data. Second, I ran a series of NCNA searches on a single Fifth Generation filmmaker, Zhang Yimou. His NCNA items peaked in 1988 at 7, dropped to 1 in 1989, rose to 3 in 1990, dropped to 0 in 1991, before rising back to 7 in 1992 and 5 in 1993. The data for the last four years is particularly intriguing, as it roughly corresponds with the release of three of his films: Ju Dou (1990), Raise the Red Lantern (1991), and The Story of Qiu Ju (1992). The first two films were temporarily banned in China, and in those two years there was very little news about Zhang reported by the NCNA. But in 1992, the number of Zhang references shot back up to pre-Tiananmen levels. Pickowicz and Zha point out that Chinese officials liked The Story of Qiu Ju, released that year, because, among other reasons, it made the Public Security Bureau look good. I conclude that the frequency of the NCNA’s Zhang Yimou news items over these three years supports Pickowicz’ and Zha’s findings. As for the 1993 data, even though none of Zhang’s films were released that year, Pickowicz says that Zhang was “allowed to win”

12

some domestic film awards in 1993, which correlates with the 5 NCNA news items mentioning Zhang. Issue 3: Chinese film and international awards After reading Pickowicz and Zha, and listening to class discussions, I was interested in learning more about Chinese film in the context of recognition abroad. How much do international film festivals and awards matter to the Chinese government, especially if some of these filmmakers show China’s ugly underbelly to foreign eyes? According to Zha, the praise from foreign film critics put the government in an uncomfortable position: “Amid all these hosannas in the West, Chinese officialdom has been gritting its teeth, more perturbed than pleased by the situation. Like many apparently inward cultures, China has an almost paradoxical craving for the laurels of external validation.”18 Aiming to add a quantitative perspective to existing qualitative research, I designed a simple frequency analysis that counts the number of Chinese film news items that were mentioned in the context of international film festivals, awards, and prizes (CF+I). Domestic awards and festivals such as the “Golden Rooster Award” and the “Dragon Film Festival” were excluded. The purpose of these searches was to gauge how the Chinese government valued Chinese film. Were Chinese film and filmmakers worthy of being mentioned as an art form and entertainment media in their own right? Or were they primarily mentioned when receiving “face” abroad, in the form of screenings and award ceremonies at international film festivals? In the pre-Tiananmen period (the five years from 1984 to 1988) between 40% and 50% of all NCNA news items that mentioned Chinese film also mentioned international awards, festivals, and prizes. Beginning in 1989 the percentage of Chinese film news 18

Zha, p. 79

13

items with these international associations dropped. The 1990 and 1991 data has to be discounted owing to the “What’s on in Beijing” column, but in 1992 the rate was still below 40%. In 1993 it shot back up to pre-Tiananmen levels (see chart, NCNA Coverage of Int'l Chinese Film vs. Int'l Chinese Opera, 1984-1993).

I also ran a test of Chinese opera items in an international context (CO+I), to provide a basis of comparison. This was problematic. I needed to identify appropriate Chinese opera items that not only involve international recognition, but are also a valid point of comparison with the international recognition awarded to Chinese films. In the entire 1984-1993 period there was only one NCNA news item that mentioned Chinese opera in the context of an international opera festival. In that same year (1993) there were 25 Chinese film items that mentioned international awards, prizes, or festivals.

14

Owing to this disparity, I decided to use “cultural exchange” as the international term to apply to the Chinese opera searches. This was a poor substitute, as such mentions in NCNA news items relate more to Chinese diplomacy than international recognition. But at least it provides some basis of comparison, and leads me to conclude that Chinese film is not viewed as a local art form as much as it is a source of prestige abroad. Chinese opera, on the other hand, stands on its own as a popular art form, entertainment medium, and cultural treasure, even though foreigners may not appreciate it. Issue 4: Zhang Yimou vs. Xie Jin The last segment of my study tests the idea that Fifth Generation filmmakers were viewed and treated differently than non-Fifth Generation filmmakers by China’s film bureaucracy. Zha’s chapter on Chen Kaige and Zhang Yimou did not mention any nonFifth Generation directors, but Pickowicz does compare directors from the two groups. He qualitatively analyzes the films of Zhang Yimou, and describes Xie Jin, the much older director whose career was re-established in the 1980s, as an “old-school melodramatist” and “thoroughly establishment figure.”19 If Xie is indeed more of an establishment figure than Zhang, then official news coverage of the two directors should be quite different, with the NCNA generally favoring Xie Jin over Zhang Yimou. With this assumption in mind, I designed two quantitative methods to evaluate NCNA coverage of the two directors. The first method was based on frequency counts in the LexisNexis database, and hypothesized that the more politically acceptable director would receive more coverage. The second method uses Yoshikoder dictionary reports to evaluate the relative amount of positive and negative words contained in news items about Zhang vs. news items about Xie. 19

Pickowicz, p. 217-218

15

Data derived from frequency counts supports Pickowicz’ qualitative assessment. When news items containing references to Zhang Yimou are compared to those mentioning Xie Jin as a percentage of all articles about Chinese film, Xie comes out firmly on top in the pre-Tiananmen period, and marginally on top in the post-Tiananmen period (see chart, “Zhang Yimou vs. Xie Jin in NCNA References, 1984-1993”). Only in 1992 did Zhang have slightly more NCNA references than Xie, and that was the year the officially sanctioned The Story of Qiu Ju was released.

However, data derived from the Yoshikoder results ran contrary to my initial expectations. This stage of the content analysis involved measuring the number of positive and negative terms in seven news items about Zhang Yimou, and seven news items about Xie Jin. As there were not enough news items to chart the results from year to year, they were organized in two groups: The five years preceding the Tiananmen

16

demonstrations (1984-1988), and the five years including and after the Tiananmen demonstrations (1989-1993). The four groups of news items were then entered into Yoshikoder, and parsed by a dictionary of positive and negative terms based on the General Inquirer content analysis program, which is regularly used for evaluating political texts. My expectations, partially based on Pickowicz’ findings and the frequency counts described in the chart above, “Zhang Yimou vs. Xie Jin in NCNA References, 19841993,” were that Xie would be associated with positive terms throughout both periods, whereas Zhang would be mostly negative. This is not what the data showed. Even though Xie was covered more often by the NCNA in the pre-Tiananmen period than Zhang, the sampled news items about Xie had a much higher level of negative terms, and a slightly lower level of positive terms (see chart, “NCNA And Xie Jin, Zhang Yimou: General Inquirer Positive and Negative Terms”). At first, this was quite puzzling. While it is conceivable that Zhang’s growing international stature would have resulted in more positive coverage than Xie during this period, there is no obvious reason why Xie-related items would have received a higher negative score. Then I looked at the news items that up the sample. One of the Xie items describes problems in the film industry.20 Another item recounts the tragic plot of Xie’s Hibiscus Town.21 Both items include numerous terms that would trigger the negative General Inquirer dictionary. Moreover, the two pre-Tiananmen items about Zhang22

20

“Leading Film Director Sets Up In Business,” Xinhua General News Service, December 6, 1988. “‘A Town Called Hibiscus,’ A Big Hit,” Xinhua General News Service, April 2, 1987. 22 “Xi’an Director Putting China’s Movies On World Screens,” Xinhua General News Service, July 29, 1988, and “Orders Pour In For Prize-Winning Red Sorghum Film,” Xinhua General News Service, March 3, 1988. 21

17

discussed his successes as a filmmaker, without delving deeply into the depressing ending of Red Sorghum and other films shot during the 1980s.

As for the post-Tiananmen news items, those about Zhang had, as expected, a higher negative score than items about Xie. However, the Zhang-related items also had a higher positive score. There are several explanations for this. First, high positive and high negative scores are not mutually exclusive using this type of methodology. Second, Zhang’s official profile was rehabilitated in 1992 and 1993, thanks to The Story of Qiu Ju. Indeed, four of the five Zhang items from the post-Tiananmen sample discuss this film, and do so in a positive manner. Third, because the NCNA reduced coverage of Zhang in the three years following Tiananmen, the sample size was small, and were weighted heavily toward items about the officially endorsed The Story of Qiu Ju. While this may be seen as a significant error on my part, I was left with no choice: There simply

18

were not many NCNA articles about Zhang to input into Yoshikoder. In 1991 he was totally shut out from NCNA news coverage — there wasn’t a single mention of Zhang in the 71,080 news items NCNA published that year.

Conclusion I hope the research described in this paper will serve as basic examples of how Chinese media content can be used to answer historical questions about Chinese film. This type of methodology is currently not a popular way to study history. Quantitative data is sometimes used to support qualitative research, as in Yingchi Chu’s skillful use of annual film production figures and box office receipts. Additionally, many researchers use Chinese media content selectively to support their hypotheses. But very few historians use computer-assisted methodologies or even basic statistical analyses of Chinese media content to understand Chinese mass media or Chinese media policy. I hope this will change, as more Chinese- and English-language media sources are included in digital databases such as LexisNexis and Factiva, and researchers are exposed to the quantitative tools and methodologies that can help them answer their research questions and support their hypotheses. And while there are certainly some drawbacks to using such methods, as noted above, I believe there is a bright future for quantitative research in the fields of Chinese film and mass media history.

19

Appendix: Full Methodology My computer-assisted content analysis was based on data gathered from the LexisNexis Academic “Guided Search” Form, accessible via the password-protected Harvard Libraries Website. A series of drop-down menus and text fields were used to select World News, Asia Pacific News, the search terms (described below), and the Source, “Xinhua” (i.e., the New China News Agency). Searches were performed on NCNA news items, which are any text content in the database that has a unique NCNA item number. Most news items are articles, reports, and official statements, but some are captions, news briefs, and entertainment briefs. The NCNA English wire service released tens of thousands of news items each year from 1984-1993. The news item totals are included in the master Excel spreadsheet that accompanies this report. All searches were full text searches, looking for terms located in any part of the news item, as opposed to just the headline, or just the headline and lede. The information entered in the Search fields were groups of terms organized around the following seven content variables, and associated search terms: Variable: F (Film) Search terms: “film” or “movie” Variable: CF (Chinese film) Search terms: “chinese film” or “china's film” or “china's film” or “chinese movie” or “Beijing film” or “chinese director” or “changchun film” or “jiangxi film” or “shanghai film” or “jiujiang yangtze film” or “domestic movie” or “zhang yimou” or “xie jin” or “chen kaige” Variable: O (Opera) Search terms: opera and not “soap opera” Variable: CO (Chinese Opera) Search terms: “china's opera” or “peking opera” or “beijing opera” or “chinese opera” or “local opera” or “traditional opera” or “guangdong opera” or “sichuan opera” or “tibetan opera” or “cantonese opera” or “taiwanese opera” or “shanghai opera” or “fujian opera” Variable: Z (Zhang Yimou) Search terms: “Zhang Yimou” Variable: X (“Xie Jin”) Search terms: “Xie Jin” Variable: I (International Festival/Award)

20

Search terms: “festival” or “award” or “prize” and not “dragon film festival” or “zhuhai film festival” or “golden rooster award” or “one hundred flower award” or “changchun film festival” or “shanghai international film” or “china international sports film festival” Exception: I associated with O (International+Opera) Search terms: “cultural exchange” (replaces “international opera festival” or “international opera prize” or “international opera award”, which only yielded one result) Searches could be combined, as the LexisNexis interface has three search fields connected by “and/or/not” drop-down menus. A combined search might be CF+I, etc. LexisNexis results of the searches included total number of results, headlines, date, and first sentence of the news item. I only used the total number of results for each search in a given year, and entered this figure in an Excel spreadsheet. I customized the spreadsheet to calculate relative percentages, and also used Excel’s “Chart Wizard” to create charts that appear in this report. For Issue 4, “Zhang Yimou vs. Xie Jin”, I identified seven articles about Xie Jin and Zhang Yimou in the NCNA search results from LexisNexis. These articles were about the directors, or their films, as opposed to simply mentioning them in passing. I copied and pasted the articles for Xie and Zhang into four text files: Xie 1983-1988, Xie 1989-1993, Zhang 1983-1988, and Zhang 1989-1993, and imported them into the Yoshikoder content analysis software program. I then applied the “General Inquirer” positive and negative dictionary to each of the text files, to measure the relative proportions of negative and positive terms, as defined in the GI dictionaries. The numerical results were then pasted into an Excel spreadsheet, and turned into a chart, which I used to compare NCNA coverage of the two directors.

21

Bibliography “‘A Town Called Hibiscus,’ A Big Hit,” Xinhua General News Service, April 2, 1987. Chang, Won Ho. Mass Media in China: The History and the Future. Ames: Iowa State University Press, 1989. Chu, Yingchi. “The Consumption of Cinema in Contemporary China,” in Hemelryk Donald, Stephanie et al, ed., Media in China: Consumption, Content, and Crisis. London: RoutledgeCurzon, 2002. Haraszti, Miklos. The Velvet Prison: Artists Under State Socialism. New York: Noonday Press, 1989. “Leading Film Director Sets Up In Business,” Xinhua General News Service, December 6, 1988. LexisNexis Academic. Reed Elsevier Inc., New York. Microsoft Excel X for Mac Service Release 1. Microsoft Corp., Redmond, Washington. “Orders Pour In For Prize-Winning Red Sorghum Film,” Xinhua General News Service, March 3, 1988. Pickowicz, Paul. “Velvet Prisons and the Political Economy of Chinese Filmmaking,” in Davis, Deborah et al, ed., Urban Spaces in Contemporary China. Cambridge: Cambridge University Press, 1995. “Xi’an Director Putting China’s Movies On World Screens,” Xinhua General News Service, July 29, 1988. Yoshikoder. Weatherhead Center for International Affairs, the Center for Advanced Study in the Behavioral Sciences, Harvard University, Cambridge, Mass. Zha, Jianying. China Pop: How Soap Operas, Tabloids, and Bestsellers Are Transforming a Culture. New York: The New York Press, 1995.

22