News Reporting on Comfort Women

Abstract

This article explored South Korean and Japanese newspaper reports on the “comfort women” who were forced into sexual slavery by the Imperial Japanese Army in the 1930s-1940s, to examine how print media have reproduced the reality of the issue. I conducted a quantitative frame analysis of the contents of news articles (N = 384) on the comfort women in four South Korean and Japanese newspapers. The frames of comfort women articles in all papers can be considered to be very stereotyped, because they have changed little according to the newspaper’s political position (conservative/liberal), attitude (anti-Japan/anti–South Korea), and nationality (South Korean/Japanese). When the relationship of South Korea and Japan has been combative, conflict and morality frames have been abundant. In contrast, when the relationship has been favorable, human interest frames have been ample.

Keywords

comfort women framing frame difference frame changing

On August 10, 2012, Lee Myung-bak, then President of the Republic of Korea (hereafter, South Korea), landed on the islets of Dokdo (called Takeshima in Japanese), over which both South Korea and Japan have declared sovereignty. This action sparked the wrath of the Japanese government of Yoshihiko Noda, and the relationship between South Korea and Japan deteriorated sharply (Nishino, 2014). Lee later disclosed the reason for his action, saying, “I have no intentions to provoke Japan or create a standoff too much, but Japan has been too insincere about the issues” (“Lee Strongly Criticizes,” 2012). The major issue referred to was that of the Korean comfort women (“Korea to Raise,” 2012).

Comfort women were sex slaves held at or near Japanese military bases during Japan’s wars of aggression against other Asian countries and Western colonial powers in the 1930s and 1940s; the largest proportion of the victims were Korean women (Niksch, 2007; Tadaki, Akizuki, Arima, Kawano, & Hata, 2014). Although the term “sex slave” is clear and straightforward, the euphemistic term “comfort women” is widely known and more commonly used even in mass media, not only in South Korea but also in Japan; the English term translates 위안부 (weanbu) in Korean and 慰安婦 (ianfu) in Japanese.

Prime Minister Noda refused to take action on behalf of the comfort women, arguing that all claim for compensation had been settled by the Japan–South Korea Claims Settlement Agreement of 1965, under which the two countries normalized national relations. That attitude ignited the fury of President Lee, who emphasized the fact that all the comfort women still alive at present are very old (over 85 years old) and that Japanese government compensation should be provided quickly while the victims are still able to benefit from it. He landed on the controversial island to express his rejection of the Japanese position.

A bigger clash between South Korea and Japan has emerged since then with the advent of new administrations in both countries. The current South Korean President, Park Geun-hye, has urged Japanese leaders to develop better historical awareness and has implied that a sincere effort on the part of the Japanese government to resolve the comfort women problem is a prerequisite for normalizing the relationship between the two countries, and even for bilateral summit talks (Nishino, 2014). To the contrary, however, the current Japanese administration of Shinzo Abe has repeatedly insisted that there is no evidence of any government or military involvement in the mobilization of comfort women (Tadaki et al., 2014) and has refused to take further action. The conflicting perspectives of the South Korean and Japanese governments on the comfort women issue have become one the of core controversies between the two countries (Niksch, 2007; Nishino, 2014). Even though both South Korea and Japan elected new governments in December 2012, the new leaders did not hold a bilateral meeting until September 2015.

People’s perceptions of comfort women, like those of other issues (McCombs & Shaw, 1972), will be affected by the media they consume. In fact, media reorganize reality by means of framing to help readers process the flood of news reports, and perceive and identify salient messages (Goffman, 1974). Framing is “persistent selection, emphasis, and exclusion” (Gitlin, 1980, p. 7) of content to make a certain aspect or aspects of an issue more or less salient. Frame analysis has been a key method in journalism and mass communication research since the 1980s, and remains so nowadays (Matthes, 2009).

However, despite the importance and volatility of the comfort women issue and the prevalence of framing research in the mass communication field, there has been insufficient scholarly work looking at framing of comfort women. To help fill this gap, the author conducted a quantitative frame analysis of the content of articles on comfort women in four South Korean and Japanese newspapers, applying several previously developed framing schemes: (a) conflict, human interest, economic consequences, morality, and responsibility frames, (b) episodic and thematic frames, (c) positive, mixed/neutral, and negative tone. In addition, the author examined how news frames on comfort women have varied and changed over time.

This study is the first to use concrete quantitative data of news content to analyze comfort women issues as represented in Japanese and South Korean newspapers. It departs from the prevailing approaches to frame research in three ways: (a) the study topic is one that is unique in the field of frame research and is still a source of conflict and controversy at the time of writing; (b) study objects are multi-dimensional, covering media from two conflicting countries and two editorial political positions (conservative/liberal); and (c) the passage of time and changes in framing of the issue over time are considered in the analysis. With this mixture of unusual features, this study tried to advance previous framing theories, especially regarding “frame difference” as discussed by Scheufele (1999) by evidencing new frame concept. Ultimately, this study aims to move toward an understanding of more desirable ways of reporting on international controversies.

Previous Literature

Media Coverage of Comfort Women

The comfort women issue has been heavily covered by news media in both South Korea and Japan. South Korean newspapers, regardless of market (nationwide/regional), political position (conservative/liberal), or content (general/business or economy oriented), have blamed the Japanese government for its attempts to deny legal responsibility and to revise the Kono Statement, which was an official statement issued by then Chief Cabinet Secretary Yohei Kono in 1993 admitting the responsibility of the Japanese military for recruiting women, generally against their will, through various kinds of pressure ranging from coaxing to coercion. For example, South Korean media without exception criticized then Lower House Member Abe’s remark that he would revise the Kono Statement if he became prime minister, made in an interview with the Japanese right-wing paper the Sankei Shimbun on August 28, 2012, approximately 4 months before he actually did take office.

In contrast, the perspectives of Japanese newspapers on comfort women issues have been divided. According to Takekawa (2014), liberal newspapers such as the Asahi Shimbun, Mainichi Shimbun, and Tokyo Shimbun have acknowledged Japan’s misdeed in mobilizing comfort women and urged the Japanese government to face the facts. However, conservative and right-wing newspapers such as the Yomiuri Shimbun and Sankei Shimbun have argued that there is no evidence clearly showing Japanese government or military involvement in the forcible mobilization of comfort women.

Comfort women issues have been a very popular subject for academic study, especially in South Korea. Both Korean and non-Korean researchers have tried to illuminate the Imperial Japanese Army’s sexual, physical, and mental abuse of comfort women and called for Japan to acknowledge its responsibility (Choi, 2013; Kim, 2010). Comfort women have mostly been described as victims of the war ignited by Japan (Boling, 1995; Cho, 2012; Soh, 2009). Recently, in addition to study of the historical events themselves, some scholars have focused on the arguments and statements made about the issue by both the South Korean and the Japanese governments (Doh, 2008; Kang, 2012).

One commonality between all these studies is that they have approached the issues from the perspective of human rights; in contrast, there has been almost no research analyzing the issue from a media studies perspective. To my knowledge there is only one such paper, Lee and Min (2011). They did a content analysis of articles on comfort women in the South Korean newspapers Chosun Ilbo and Hankyoreh from December 1, 2000, to December 31, 2010, and concluded that the most frequently found frames were “Japanese sexual slavery” and “Japanese apology and compensation.” The present research expands the scope of analysis of comfort women framing from that set by Lee and Min (2011) by including another country’s media and adopting various framing schemes and frame comparison conditions, as described further below.

News Framing

Entman (1993) has defined framing as “select[ing] some aspects of a perceived reality and mak[ing] them more salient in a communicating text, in such a way as to promote a particular problem definition, causal interpretation, moral evaluation, and/or treatment recommendation for the item described” (p. 52). One commonly used distinction in classifying news frame analysis is between “inductive” and “deductive” frames. If no previous frame model suggests itself as suitable for adoption to analyze news content, researchers can design new frames inductively, with an open-ended view. This approach takes time and is usually based on “small samples” (Semetko & Valkenburg, 2000, p. 94). In contrast, the deductive approach can be applied when predefined frames exist that are adoptable to the analysis of the news content being considered. This is useful when large samples of material need to be processed. The current study investigates frames of comfort women articles using a deductive method mainly based upon the division in Semetko and Valkenburg (2000), who investigated earlier frame studies and identified the most commonly used deductive frames as the (a) conflict frame, (b) human interest frame, (c) economic consequences frame, (d) morality frame, and (e) responsibility frame. The conflict frame “emphasizes conflict between individuals, groups or institutions as a means of capturing audience interest” (Semetko & Valkenburg, 2000, p. 94). d’Haenens and Lange (2001, p. 850) noted that the conflict frame reflects a “winning and losing” perspective and could also be called the “horse race frame.” It is known to lead the public to mistrust (Cappella & Jamieson, 1997). The human interest frame focuses on the personal and emotional element. As articles casting light on people and their circumstances arouse readers’ sympathy as well as interest, journalists often introduce a case of a person affected by or witnessing some news event in the lead of a complicated article. The economic consequences frame deals with economic impact or economic benefit/detriment. The morality frame “add[s] a religious or moral charge to an event, problem or subject” (d’Haenens & Lange, 2001, p. 850). It can be detected by checking whether an article contains a moral message or makes reference to morality, or also to God or other religious matters. Finally, the responsibility frame “presents an issue or problem in such a way as to attribute responsibility for its cause or solution to either the government or to an individual or group” (Semetko & Valkenburg, 2000, p. 96).

Another way to classify frames is based on “episodicity” and “thematicity.” Iyengar and Kinder (1991) said that “a lack of detail or context in a media article that focuses exclusively on the event and person is called an episodic frame” (p. 2). That is, an episodic frame eschews analytic or explanatory writing and instead prefers to present cases and visuals. When Pfau et al. (2004) compared reports of embedded and non-embedded journalists in the first days of the 2003 U.S. invasion of Iraq, it was found that embedded journalists, who have more chances to cover battlefields visually, more often produced episodic-framed reports. The use of a simple episodic frame presenting selected idiosyncratic and not necessarily representative cases, without context, in a vivid and emotionally evocative way, has the risk of obscuring the reality and leading readers to misunderstand (Niner, Ahmad, & Cuthbert, 2013). In fact, Iyengar and Simon (1994) found that American media coverage of the Gulf War was “heavily episodic or event-oriented” (p. 179) and found that watchers of TV news coverage of the war preferred military solutions over diplomatic methods. In contrast, thematic frames can be found in in-depth analysis dealing with background, cause and effect, context, and similar matters. Whereas an episodic frame would focus on an individual, a single event, and the private realm, a thematic frame concentrates on an issue, trends over time, and the public realm (Iyengar & Kinder, 1991).

Even though the tone of news—positive, mixed/neutral, or negative—is not formally integrated into frame analysis, many studies have adopted tone analysis as part of news frame research (e.g., Bedingfield & Anshari, 2014; Parry, 2011; Pfau et al., 2004). In most of these studies, tone has been detected by a combination of quantitative and qualitative-impressionistic methods, specifically, by qualitative evaluation of quantified language use. For example, articles supporting a certain issue or attitude and including affirmative words or comments were classified as having a positive tone. As negative-toned news generally draws attention and deliver news more vividly and sensationally (Bohle, 1986), media companies often prefer negative tone to positive tone (Rattliff, 2001). However, politically negative political news causes readers to develop cynical attitudes toward politics (Cappella & Jamieson, 1997). Moon and Sung (2006) surveyed 260 South Korean in 2006. They evaluated citizens’ perception on Japan after being exposed to the negative-toned articles and their intention of purchasing Japanese products. They found that the more South Koreans were exposed to the negative articles, the more negative their perception of Japan and in turn, the stronger their negative feelings about the purchase of Japanese products. Moon and Sung (2006) concluded that exposure to mass media content containing negative images of Japan caused negative perceptions toward Japan.

Based upon these previous studies, this article adopts the following research questions to shape its examination of frames of articles on comfort women.

RQ1: What is the relative prominence of the conflict, human interest, economic consequences, morality, and responsibility frames in comfort women articles in four Japanese and South Korean newspapers?

RQ2: What is the relative prominence of the episodic and thematic frames in comfort women articles in four Japanese and South Korean newspapers?

RQ3: What is the relative prominence of positive, mixed/neutral, and negative tones in comfort women articles in four Japanese and South Korean newspapers?

Frame Difference

Scheufele (1999) developed a process model of framing by identifying four key processes: frame building, frame setting, the individual-level process of framing, and a feedback loop from audiences to journalists (pp. 114-118). Frame building refers to how frames are created, constructed, or adopted by media; it is distinct from frame setting, dealing with that how frames are applied by media or are taken up by the public. Scheufele (1999) suggested three potential sources or levels of influence on frame building: (a) the journalist level, covering factors such as ideology, attitudes, and professional norms as manifested in the individual reporter; (b) the organizational level, covering the type or political orientation of the medium; and (c) the external level, such as political actors, authorities, or interest groups. Therefore, the same issue covered by different media outlets can be expressed through different frames according to that outlet’s pursuits and aims (Schudson, 2003). Frame differences exist along various axes: for instance, they have been found not only according to conventional left–right political distinctions but also between elite national newspapers and local newspapers (Carpenter, 2007). In Semetko and Valkenburg’s (2000) study, frame difference was not detected between media (television vs. newspaper), but was detected between sensationalist and serious news outlets.

News frame can be analyzed not only at the organizational but also at the national level. That is, as news coverage and frames are affected by national political systems, culture, and social norms (Hallin & Mancini, 1985; Scheufele, 1999), news frames will be different according to media’s nationality. Oh (2011) analyzed frames in six newspapers in South Korea and Japan covering the 100th anniversary of Japan’s annexation of Korea in 1910, and concluded that there was a frame difference between the two countries’ newspapers: South Korean newspapers more frequently adopted morality frames than Japanese newspapers, which for their part used more future-oriented frames than South Korean newspapers. Thus,

RQ4: How do conflict, human interest, economic consequences, morality, and responsibility frames differ under conditions such as the newspaper’s editorial political position (conservative/liberal), attitude toward the opposite country in relation to the comfort women issue (anti-Japan/anti–South Korea), and nationality (South Korean/Japanese)?

Frame Change

Another important factor to think about when conducting frame analysis is time. Even coverage of the same issues may evince changing frames over time, depending on changes in related events, public attention, and so on (Chyi & McCombs, 2004). In fact, public attention to a given issue is generally very high in the beginning and fades out as time goes by, as Downs (1972) indicated with the idea of the issue-attention cycle, composed of five steps: (a) a pre-problem stage, (b) alarmed discovery and euphoric enthusiasm, (c) realizing the cost of significant progress, (d) gradual decline of intense public interest, and (e) a post-problem stage. When Houston, Pfefferbaum, and Rosenholtz (2012) studied frame change in coverage of major U.S. natural disasters from 2000 to 2010, they found that “the environment frame—which captured the death and destruction resulting from the disaster[s]—constituted the vast majority of all coverage frames in the immediate aftermath of the disaster[s], but quickly declined during the first two months after the disaster[s]” (p. 615). Similarly, coverage of the Iraq War beginning in 2003 on NYTimes.com also changed over time from episodic frames related to conflict and violence to thematic frames containing coverage of reconstruction and the future prospects of Iraq (Dimitrova, 2006). Therefore,

RQ5: How do conflict, human interest, economic consequences, morality, and responsibility frames change across time (favorable/combative relationship)?

Research Methods

Sampling and Data

Two national newspapers, representing conservative and liberal perspectives, were selected for each country: the conservative Chosun Ilbo and the liberal Hankyoreh for South Korea, and the conservative Yomiuri Shimbun and the liberal Asahi Shimbun for Japan. To evaluate frame change over time, the research period was divided into four (sub-)periods based upon the atmosphere of relations between the two countries: basically favorable or basically combative, as follows.

Period 1 (favorable)

From October 8, 1998 to October 8, 1999. Former South Korean President Kim Dae-jung and former Japanese Prime Minister Keizo Obuchi presented a “South Korea–Japan Partnership Statement” after a summit talk on October 8, 1998. Obuchi officially expressed deep regret and apologies for Japan’s colonial conquest of Korea, and Kim promised to overcome the unfortunate history and make efforts to proceed toward a future-oriented relationship. After the meeting, the two countries’ relationship improved dramatically.

Period 2 (combative)

From February 22, 2005 to February 22, 2006. The prefectural assembly of Shimane Prefecture in Japan established “Takeshima Day” on February 22, 2005, igniting South Korean anger because South Korea had de facto control of the island and there was a longstanding dispute around its sovereignty. The two countries’ relationship was drastically worsened as a result.

Period 3 (favorable)

From September 16, 2009 to September 16, 2010. The liberal Democratic Party of Japan (DPJ), an opposition party, took power for the first time in Japanese history in September 2009. Compared with the conservative Liberal Democratic Party (LDP), which had held power for more than 50 years, the DPJ placed a high priority on a friendly relationship between Japan and South Korea. For instance, the first DPJ prime minister, Yukio Hatoyama, banned cabinet members from paying their respects at Yasukuni Shrine in Tokyo, which commemorates soldiers who died during wars but also enshrines 14 A-Class war criminals. South Korea and China have strongly opposed the visits of Japanese prime ministers and cabinet members to pay honors at the shrine.

Period 4 (combative)

From December 26, 2012 to December 26, 2013. The LDP recaptured power in Japan in December 2012 as Shinzo Abe was elected Prime Minister. Abe is very conservative and has pursued a “strong Japan” policy that involves legitimating Japan’s wartime history, and so the Abe administration has caused a lot of friction with South Korea. The comfort women conflict was one of the core diplomatic problems during this period and remains so.

As the influence of a given event on public opinion is very strong in the beginning (Downs, 1972), it is relatively easy to pinpoint the advent or beginning of an issue’s influence. However, the point when the direct influence of an issue can be said to have expired is not clear. The author felt that 1 year would be a sufficient period to capture the emotional aspect of people’s responses to international issues or disputes between Korea and Japan as well as their intellectual positions or opinions. Therefore, the author arbitrarily set a 1-year window for each of these periods, for reasons of comparability.

Articles with keywords comfort women (위안부/weanbu in Korean and 慰安婦/ianfu in Japanese) and sex slave (성노예/sungnoye in Korean and 性奴隷/seidorei in Japanese) were selected; articles with fewer than 400 characters (equivalent to approximately 150 words in English) were excluded, as short articles usually focus on delivering facts and have little room to implement frames, and thus may provide insufficient or skewed data for the evaluation of those frames (Bosman & d’Haenens, 2008). Articles where the main topic was not comfort women (where they were merely mentioned) were also excluded.

The author employed the five most frequently used frames as identified by Semetko and Valkenburg (2000)—conflict, human interest, economic consequences, morality, and responsibility—along with a deductive approach. Twenty questions developed by Semetko and Valkenburg were used to extract frames and each question was marked 1 for the answer yes and 0 for no.

The author conducted a principal component analysis with varimax rotation on the answers to these 20 questions to yield clusters corresponding to the five common frames. The values of the cluster scales ranged from 0 (frame not present) to 1 (frame absolutely present). Despite the conflicting opinions on the legitimacy of using binary data in factor analysis and the risk that correlations between binary variables are low, Semetko and Valkenburg (2000) stressed that they found “a very clear factor structure” using this approach and thus “no compelling reason to perform another kind of (factor) analysis” (p. 99). In addition, their binary data have been reused without any change in many subsequent framing studies (e.g., Bosman & d’Haenens, 2008; d’Haenens & Lange, 2001; Oh, 2011; Semetko & Valkenburg, 2000). Therefore, the author decided to use binary data in factor analysis in this study.

Internal consistency was measured by Cronbach’s alpha, and quantitative data analysis was processed with SPSS (version 18.0). Among the 20 questions, as there was no article using the economic consequences frame, the author removed three questions relevant to this frame, leaving 17 questions for analysis. Principal component analysis with varimax rotation was conducted as a factor analysis; according to the results, the 17 questions clustered under four frames: conflict, human interest, responsibility, and morality. Only items with factor loadings higher than .50 and eigenvalues higher than 1.0 were selected. After obtaining Cronbach’s alphas to measure the internal consistency of these frames as a reliability analysis, the author deleted three more questions that were lowering overall internal consistency. Therefore, 14 questions were ultimately adopted. The alpha value of the responsibility frame was 0.96; that of the human interest frame, 0.96; that of the conflict frame, 0.97; and that of the moral frame, 0.96. The results are shown in Appendix.

The tone of news was divided into three categories: positive (supporting one or the other side in the argument on the comfort women issue with affirmative words or comments), mixed/neutral (delivering only facts or containing both positive and negative stance), or negative (blaming some government or individual(s) for the comfort women issue). In addition, each tone was examined for the direction of the sentiments expressed (regardless of positivity/negativity): toward South Korea or toward Japan.

If an article was somewhat short and focused on events and people, it was sorted as belonging to the episodic frame. In contrast, if an article included in-depth information such as background, cause and effect, or analysis of an event, it was sorted into the thematic frame. Both tone of news and episodic/thematic frame division were evaluated impressionistically (interpretively) by the raters.

Two South Korean graduate students in mass communication who were also fluent in the Japanese language performed the coding. The author first explained the coding standards and let them code some initial articles, then discussed the coding with them and revised it repeatedly until good intercoder reliability was reached. Ultimately, 50 articles were randomly selected and coded. Intercoder reliability evaluated by Krippendorff’s alpha was .86 with 20 framing questions, .94 with episodic/thematic judgments, and .95 with news tone. The average Krippendorff’s alpha was .92, which can be regarded as very high and as adequate to proceed with coding.

Results

The total number of articles used as data in the current research was 384, as presented in Table 1. Chosun Ilbo and Hankyoreh were represented by 124 articles each, the biggest number, and Yomiuri Shimbun by 44 articles, the smallest. Asahi Shimbun was represented by 92 articles. What these numbers reveal is that South Korean newspapers have put higher news value on the comfort women issue than Japanese newspapers. It is particularly striking that the number of articles from the Yomiuri Shimbun was only 44, approximately half that of the other Japanese newspaper and a third that of the South Korean newspapers. This is interesting in light of the fact that the Yomiuri Shimbun has officially denied Japanese government responsibility for the enslavement of the comfort women, saying that there is no evidence showing Japanese government involvement. This is the same as the Japanese government’s own stance. Furthermore, just as the Japanese government does not voluntarily raise the comfort women issue because international society mostly puts the blame for it on Japan, the Yomiuri Shimbun also covers the topic only passively.

Table 1.

News Articles Used for Analysis.

	Chosun Ilbo	Hankyoreh	Yomiuri Shimbun	Asahi Shimbun	Total
Period 1 (favorable)	9	17	6	15	47
Period 2 (combative)	38	56	8	38	140
Period 3 (favorable)	6	12	0	5	23
Period 4 (combative)	71	39	30	34	174
Total	124	124	44	92	384

The number of articles written on comfort women during favorable eras was 47 for Period 1 and 23 for Period 3, whereas the number of articles during combative eras was 140 for Period 2 and 174 for Period 4—at least triple the number during favorable eras. This shows that the comfort women issue seems to be actively raised whenever the relationship between the two countries is struggling. The comfort women problem was the main conflict factor during Period 4, but not during Period 2, when the main conflict factors were territorial issues and the treatment of the colonial and war era(s) in Japanese history books.

RQ1, RQ2, and RQ3: News Framing

Each of the four newspapers used the responsibility frame most frequently, without exception, as Table 2 shows. South Korean newspapers showed higher mean scores for the responsibility frame (Chosun Ilbo = .43, Hankyoreh = .42) compared with Japanese newspapers (Yomiuri Shimbun = .34, Asahi Shimbun = .23). South Korean newspapers frequently delivered expressions of blame toward the Japanese government on the part of the South Korean government, the substance of which was that the Japanese government had not acknowledged its responsibility for the forcible mobilization of comfort women; they often then went on to urge the Japanese government to show greater historical awareness. Those articles were usually coded as belonging to the responsibility frame.

Table 2.

Mean Scores of Four News Frames in Four South Korean and Japanese Newspaper.

Newspaper	Responsibility	Human interest	Conflict	Morality	n
Chosun Ilbo	0.43	0.18	0.17	0.10	124
Hankyoreh	0.42	0.23	0.10	0.06	124
Yomiuri Shimbun	0.34	0.04	0.21	0.18	44
Asahi Shimbun	0.23	0.20	0.20	0.02	92
Whole sample	0.37	0.19	0.16	0.08	384

Note. Scores range from 0 (not present) to 1 (absolutely present).

The second most frequently used frame differed according to newspaper nationality: human interest frames in South Korean newspapers and conflict frames in Japanese newspapers. Examples of the human interest frame included South Korean newspapers’ coverage of comfort women’s bitter experience and Japanese politicians’ stories who denied Japan’s duty to compensate comfort women. In contrast, Japanese newspapers reported many articles dealing with the conflict between the South Korean and Japanese governments on comfort women issues and with the two sides’ positions, which were assigned the conflict frame.

Most newspapers rarely adopted a morality frame; the exception was the Yomiuri Shimbun. However, because the Yomiuri Shimbun sample was small (n = 44), it is possible that the presence of a few morally framed articles like “Comfort Women Were Needed” (May 14, 2013) or “Kick Out the Troublemaker, Lower House Member Nishimura” (May 18, 2013) represented a disproportionately high sample of moral frames.

All newspapers showed a higher proportion of episodic frames (74.0%) than thematic frames (26.0%), as seen in Table 3. In particular, South Korean newspapers showed much higher proportions of episodic frames than Japanese newspapers. South Korean newspapers maintain a simpler focus on events, people, or newsworthy statements, such as the provocative words of Japanese politicians. In contrast, Japanese newspapers run many in-depth articles on this issue, including background, cause and effect, expert analysis, and so on, and running relatively long.

Table 3.

Proportions of Episodic and Thematic Frames in Four South Korean and Japanese Newspaper.

Newspaper	Episodic (%)	Thematic (%)	n
Chosun Ilbo	88.7	11.3	124
Hankyoreh	79.0	21.0	124
Yomiuri Shimbun	56.8	43.2	44
Asahi Shimbun	55.4	44.6	92
Whole sample	74.0	26.0	384

The tone of comfort women articles in the four newspapers was generally mixed/neutral (56.5%) or negative (42.7%) as seen in Table 4. The proportion of a negative tone toward the opposite country was around 50% in the South Korean newspapers. In contrast, it was very low—around 10%—in the Japanese newspapers. One notable finding is that only the Asahi Shimbun blamed the home country (19.6%); all the other newspapers reported with articles negatively characterizing the opposite country or its actions. In fact, editorials in the Asahi Shimbun have often urged the Japanese government to acknowledge its forcible mobilization of comfort women, just as the South Korean newspapers have.

Table 4.

Proportion of News Tones in Four South Korean and Japanese Newspaper.

	Positive (%)			Negative (%)		n
Newspaper	Home	Opposite	Mixed/neutral (%)	Home	Opposite	n
Chosun Ilbo	2.4	0	39.5	1.6	56.5	124
Hankyoreh	0	0	46.0	4.8	49.2	124
Yomiuri Shimbun	0	0	84.1	2.3	13.6	44
Asahi Shimbun	0	0	80.4	19.6	0	92
Whole sample	0.8	0	56.5	7.0	35.7	384

RQ4: Frame Difference

The author conducted an independent-samples t test to determine the factors causing frame difference. Newspapers were divided into two groups on the basis of the criteria of editorial political position (conservative or liberal), attitude on comfort women issue (positive or negative toward Japan or South Korea), and nationality (South Korean or Japanese). Mean scores were compared by group; Table 5 shows the results.

Table 5.

Mean Scores for Each News Frame by Condition.

Newspaper		Responsibility	Human interest	Conflict	Morality
Political position	F	5.78	11.34*	4.52	39.69**
	Cohen’s d	0.13	0.19	0.11	0.30
Conservative (n = 168)		0.40	0.15	0.18	0.13
Liberal (n = 216)		0.34	0.22	0.14	0.05
Attitude	F	0.17	47.56***	2.70	27.01
	Cohen’s d	0.07	0.71	0.16	0.37
Anti-Japan (n = 340)		0.37	0.20	0.15	0.06
Anti-Korea (n = 44)		0.34	0.41	0.21	0.18
Nationality	F	30.96**	7.60	8.51	0.21
	Cohen’s d	0.36	0.17	0.15	0.04
S. Korean (n = 248)		0.42	0.21	0.14	0.08
Japanese (n = 136)		0.26	0.15	0.20	0.07

Note. Scores range from 0 (not present) to 1 (absolutely present).

p < .05. **p < .01. ***p < .001.

In terms of editorial political position (conservative for the Yomiuri Shimbun and the Chosun Ilbo, and liberal for the Hankoryeh and the Asahi Shimbun), both the human interest frame (t = −1.99, df = 375.10, p = .048; Cohen’s d = 0.19) and the morality frame (t = 2.94, df = 263.76, p = .004; Cohen’s d = 0.30) were significantly different. Liberal newspapers adopted more human interest frames and less morality frames. However, all effect sizes of morality and human interest frames as measured by Cohen’s d were 0.30 or below; thus, the difference was significant but small.

In terms of attitude (anti–South Korea for the Yomiuri Shimbun and anti-Japan for the three other papers), only the human interest frame (t = −4.66, df = 93.16, p = .000; Cohen’s d = 0.71) was different: the Yomiuri Shimbun used human interest frames much more frequently than the other papers. The effect size was large enough to detect a difference between the two groups.

In terms of the division by nationality, only the responsibility frame (t = 3.38, df = 311.85, p = .001; Cohen’s d = 0.36) showed a significant difference: South Korean newspapers used more responsibility frames than Japanese newspapers. The effect size as measured by Cohen’s d was moderate.

RQ5: Frame Change

Frames changed considerably according to time (favorable or combative period), as seen in Figure 1. Responsibility frames appeared most often in all periods except Period 3 (favorable). Conflict frames rarely appeared during favorable periods (M = 0.04 for both Period 1 and Period 3) but were abundant during combative periods (M = 0.18 for Period 2 and M = 0.19 for Period 4). According to an independent-samples t test between the two groups, divided by favorable (n = 70) and combative (n = 314) periods, three frames changed significantly: human interest (t = 2.82, df = 89.42, p = .006; Cohen’s d = 0.39), conflict (t = −4.84, df = 243.24, p = .000; Cohen’s d = 0.51), and morality (t = −4.89, df = 381.09, p = .000; Cohen’s d = 0.38). Conflict and morality frames were conspicuous during combative periods, while human interest frames were abundant during favorable periods. Effect sizes for the three frames as measured by Cohen’s d were large with conflict frame and moderate with human and morality frames. Responsibility frames were also higher during the combative era, although this difference was not statistically significant.

Figure 1.

Frame changing according to period (favorable/combative).

Conclusions and Suggestions

This study examined South Korean and Japanese newspapers’ reports on the comfort women issue by analyzing news content, with the goal of advancing frame theory and ultimately furthering our understanding of desirable ways of reporting on international controversies. The analysis obtained several meaningful findings.

First, this research expands framing theory from the perspective of Scheufele (1999)’s assertion that journalist, organizational, and external levels may influence frame building and thus cause frame difference. The responsibility frame was most abundant regardless of the paper’s nationality (South Korea/Japan), editorial political position (conservative/liberal), or the period of relations between the two countries (favorable/combative). In addition, none of the four frames—responsibility, conflict, human interest, and morality—varied much over these changing conditions. Only in a few cases were any of the four frames different with statistical significance: the human interest and the morality frames by editorial political position; the human interest frame with attitude (anti–South Korea/anti-Japan); and the responsibility frame by nationality. In addition, effect sizes for differences by editorial political position and nationality as evaluated by Cohen’s d ranged from 0.19 to 0.36, levels that are not high. These results suggest a new concept in framing, which we can call the stereotyped frame and raises a possibility that highly controversial and sensitive issues may be reported in a similar way regardless of personal, organizational, external, or even national inclinations.

Second, this study shows repetition of same frames according to the period of relations (favorable/combative), which has been rarely found in the previous frame theories. When the relationship has been combative, conflict and morality frames have been abundant (as have responsibility frames, even though this finding was not statistically significant). In contrast, when the relationship has been favorable, human interest frames have been ample. As noted by Chyi and McCombs (2004), Downs (1972), Houston et al. (2012), and Dimitrova (2006), public attention to and significance attached to issues change over time, and thus frames set on issues by media change as well. However, the results of the current study reflect that there is also a tendency of the same frames to persist, especially frames reflecting a favorable or combative national environment, which constitutes a meaningful addition to framing theory.

Third, South Korean newspapers showed a very high incidence of negative tone toward Japan as well as of episodic frames. In the present data, 60 out of 124 articles (48.39%) in the Chosun Ilbo showed both an episodic frame and a negative tone, as did 50 out of 124 articles (40.32%) in the Hankyoreh. However, both episodic frame and negative tone were very uncommon in the Japanese newspapers; they were present in only six out of 92 articles (6.52%) in the Asahi Shimbun and zero out of 44 articles (0%) in the Yomiuri Shimbun. As Niner et al. (2013) and Iyengar and Simon (1994) indicate, articles using a simple episodic frame and presenting not necessarily representative cases have the risk of leading readers to misunderstand the situation or to support radical and not necessarily helpful solutions. In addition, negative-toned news arouses cynicism or negative perceptions against the object of the news (Cappella & Jamieson, 1997; Moon & Sung, 2006). Therefore, it seems plausible that the prevalence of episodic frame and negative tone in comfort women articles in South Korean newspapers may lead readers to have a more negative perception of the Japanese government and politicians.

As of October 2015, efforts to resolve the comfort woman issue are still in a state of deadlock. Regular talks have been held at a high level between the foreign affairs ministries of the two countries since May 2014 to solve the problem, but they have not produced any meaningful results. Thus, on the basis of this study’s findings, the author would like to suggest some practical implications by which news coverage may be able to help foster solutions and not exacerbate the problem.

All the newspapers considered here, but especially the South Korean newspapers, need to report on delicate or controversial issues both more broadly and more in depth. In other words, South Korean newspapers need to write more articles using thematic frames, not just deliver simple episodic descriptions of events. As South Korea and Japan have repeatedly experienced political and diplomatic conflict on many issues, including not only comfort women but also territorial disputes, Japanese history textbooks, the visits of Japanese prime ministers to Yasukuni Shrine, and so on, simply reporting politicians’ volatile comments will continue to severely inflame nationalist sentiment among South Koreans.

In addition, it can be recommended that South Korean newspapers write comfort women articles with a less sensationalistic tone. In some cases, readers may be motivated negatively toward Japan simply by reading a title; for example, “Japanese Cabinet Denied Comfort Women, Neglected Korean People” (Chosun Ilbo, July 31, 2013), and “Abe’s Lie Revealed by Evidence of Forcible Mobilization of Comfort Women” (Hankyoreh, October 7, 2013).

As for Japanese newspapers, they need to be careful not to produce provocative comfort women articles under the conflict frames that they frequently adopt. Conflict frames deal with disagreements between parties and usually contain reproaches, which can lead the public to mistrust the object of those approaches (Cappella & Jamieson, 1997). As noted above, conflict frames were abundant during combative periods in this study.

The significance of the current study is in its analysis of comfort women articles from a media studies perspective. As the comfort women problem is one of the core ongoing points of conflict between South Korea and Japan and is widely reported by media in the two countries as well as abroad, it is meaningful to analyze how media have reproduced, framed, and delivered the reality of the issue. By doing so, this study could contribute to deepen framing theories.

However, the current study has some limitations, too, of which the most important is that all coders were South Korean students. As the comfort women issue is highly charged, the results might have differed if coding had been done by Japanese coders, with their different cultural background and perspectives (on the issue and in general). The author expects future studies to be able to address this lack of complementary coding. For example, comparing coding of the same data done by Japanese versus coding done by South Koreans might be an interesting study in its own right.

In addition, the current study has not presented empirical evidence demonstrating that a simple episodic frame and negative-toned news arouse negative perceptions; this remains a plausible but unproven assumption. Furthermore, the study examined only newspaper coverage; it would have been more comprehensive and perhaps yielded a different perspective if analysis of television news coverage had been included as well. The author expects that these drawbacks can be addressed in future studies. In particular, empirical studies on the relationship between news frames and people’s emotions and perceptions, across various media outlets, on highly controversial issues like that of the comfort women are desirable.

Footnotes

Appendix

Factor Analysis of All Newspaper Articles (N = 384).

Framing items	Factor 1Responsibility	Factor 2Human interest	Factor 3Conflict	Factor 4Morality
Attribution of responsibility frame
1. Does the story suggest that some level of government has the ability to alleviate the problem/issue?	.91	−.23	−.21	−.12
2. Does the story suggest that some level of government is responsible for the problem/issue?	.83	−.29	−.29	−.19
3. Does the story suggest solution(s) to the problem/issue?	.92	−.20	−.18	−.10
4. Does the story suggest the problem/issue requires urgent action?	.90	−.20	−.18	−.09
Human interest frame
5. Does the story provide a human example or “human face” on the issue?	−.21	.90	−.13	−.10
6. Does the story employ adjectives or personal vignettes that generate feelings of outrage, empathy/caring, sympathy, or compassion?	−.22	.93	−.13	−.10
7. Does the story emphasize how individuals and groups are affected by the issue/problem?	−.18	.90	−.11	−.07
8. Does the story go into the private or personal lives of the actors?	−.10	.86	−.04	−.02
9. Does the story contain visual information that might generate feelings of outrage, empathy/caring, sympathy, or compassion?	−.22	.92	−.11	−.09
Conflict frame
10. Does the story reflect disagreement between parties/individuals/groups/countries?	−.25	−.16	.94	−.10
11. Does one party/individual/group/country reproach another?	−.20	−.12	.92	−.08
12. Does the story refer to two or more sides of the problem or issue?	−.26	−.17	.94	−.10
Morality frame
13. Does the story contain any moral message?	−.17	−.14	−.13	.95
14. Does the story make reference to morality, God, or other religious tenets?	−.17	−.13	−.12	.96

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Author Biography

Hyeong-Jun Pak has been a journalist for the Donga Ilbo in the Republic of Korea (South Korea) since 2001 and been dispatched to its Tokyo branch as a newspaper correspondent since June 2012 for 3 years. His 2013 PhD dissertation completed at Hanyang University, South Korea, was titled “The Effect of Newsroom Convergence on Job Satisfaction—Focusing on News Coverage Environment, Organizational Atmosphere and Quality of Journalism.” Research interests include factors pertaining to the media industry, journalism, and content analysis of South Korean and Japanese newspapers.

References

Bedingfield

Anshari

(2014). Thinking about Romney: Frame building in a battleground state in the 2012 presidential election. Journalism & Mass Communication Quarterly, 91, 78-97.

Bohle

R. H.

(1986). Negativism as news selection predictor. Journalism Quarterly, 63, 789-796.

Boling

(1995). Mass rape, enforced prostitution, and the Japanese imperial Army: Japan eschews international legal responsibility? Columbia Journal of Transnational Law, 32, 533-590.

Bosman

d’Haenens

(2008). News reporting on Pim Fortuyn: Framing in two Dutch newspapers. Media, Culture & Society, 30, 735-748.

Cappella

Jamieson

(1997). Spiral of cynicism. New York, NY: Oxford University Press.

Carpenter

(2007). U.S. elite and non-elite newspapers’ portrayal of the Iraq War: A comparison of frames and source use. Journalism & Mass Communication Quarterly, 84, 761-776.

Cho

S.-H.

(2012). Is the Japanese military “comfort women” issue a humanitarian problem? On the recent positions of the Korean and the Japanese Governments. Democratic Law, 49, 165-195. (In Korean)

Choi

H.-S.

(2013). 20th and 21st century women’s bodies on the Korean Peninsula: Military sexual slavery by Japan, female partisans, and North Korean women defectors. Korean National Culture, 46, 277-311. (In Korean)

Chyi

H. I.

McCombs

M. E.

(2004). Media salience and the process of framing: Coverage of the Columbine school shooting. Journalism & Mass Communication Quarterly, 81, 22-35.

10.

Dimitrova

D. V.

(2006). Episodic frames dominate early coverage of Iraq war in the NYTimes.com. Newspaper Research Journal, 27(4), 79-83.

11.

Doh

S.-H.

(2008). A new perspective on “comfort women” in international human rights law. Korean Journal of International Law, 53(3), 41-68. (In Korean)

12.

Downs

(1972). Up and down with ecology: The “issue-attention cycle.” Public Interest, 28, 38-50.

13.

d’Haenens

Lange

(2001). Framing of asylum seekers in Dutch regional newspapers. Media, Culture & Society, 23, 847-860.

14.

Entman

R. M.

(1993). Framing: Toward clarification of a fractured paradigm. Journal of Communication, 43(4), 51-58.

15.

Gitlin

(1980). The whole world is watching: Mass media in the making and unmaking of the new left. Berkeley: University of California Press.

16.

Goffman

(1974). Frame analysis: An essay on the organization of experience. New York, NY: Harper & Row.

17.

Hallin

D. C.

Mancini

(1985). Speaking of the president: Political structure and representational form in US and Italian TV news. Theory and Society, 13, 829-850.

18.

Houston

J. B.

Pfefferbaum

Rosenholtz

C. E.

(2012). Disaster news: Framing and frame changing in coverage of major U.S. natural disasters, 2000-2010. Journalism & Mass Communication Quarterly, 89, 606-623.

19.

Iyengar

Kinder

D. R.

(1991). Is anyone responsible? How television frames political issues. London, England: University of Chicago Press.

20.

Iyengar

Simon

(1994). News coverage of the Gulf crisis and public opinion: A study of agenda-setting, priming, and framing. In Bennett

W. L.

Paletz

D. L.

(Eds.), Taken by storm: The media, public opinion, and U.S. foreign policy in the Gulf War (pp. 167-185). Chicago, IL: University of Chicago Press.

21.

Kang

J.-S.

(2012). Controversy about the Japanese military’s sexual slavery (comfort women). Gender Review, 26, 46-51. (In Korean)

22.

Kim

H.-K.

(2010). Japanese “comfort women” and responsibility. Asia Studies, 53(3), 79-108. (In Korean)

23.

Korea to raise sex slavery at U.N. (2012, September 3). Korea Herald. Retrieved from http://www.koreaherald.com/view.php?ud=20120903001366

24.

Lee

Y.-M.

Min

(2011). Social movements and media frame: Collective action frames of the Japanese “comfort women” issue. Media, Gender & Culture, 19, 39-70. (In Korean)

25.

Lee strongly criticizes Japan for lack of efforts to resolve colonial-era issues. (2012, August 13). Korea Herald. Retrieved from http://www.koreaherald.com/view.php?ud=20120813000221

26.

Matthes

(2009). What’s in a frame? A content analysis of media framing studies in the world’s leading communication journals, 1990-2005. Journalism & Mass Communication Quarterly, 86, 349-367.

27.

McCombs

M. E.

Shaw

D. L.

(1972). The agenda-setting function of mass media. Public Opinion Quarterly, 36, 176-187.

28.

Moon

S.-J.

Sung

J.-Y.

(2006). Exposure to negative media issue and anti-Japanese sentiment. Korean Journal of Broadcasting and Telecommunication Studies, 20(5), 70-105. (In Korean)

29.

Niksch

(2007). Japanese military’s “comfort women” system (Congress Research Service report to the U.S. congress). Retrieved from http://japanfocus.org/data/CRS%20CW%20Report%20April%2007.pdf

30.

Niner

Ahmad

Cuthbert

(2013). The “social tsunami”: Media coverage of child abuse in Malaysia’s English-language newspapers in 2010. Media, Culture & Society, 35, 435-453.

31.

Nishino

(2014, October 18). Japan-Korea relationship: Diagnosis and prescription. Paper presented at the 2014 Japan-Korea Journalists Forum, Tokyo. (In Japanese)

32.

D.-Y.

(2011). Frame difference analysis on the news of Korean and Japanese newspapers regarding the 100th anniversary of Japan’s colonization of Korea. Korean Journal of Journalism and Communication Studies, 55(1), 140-168. (In Korean)

33.

Parry

(2011). Images of liberation? Visual framing, humanitarianism and British press photography during the 2003 Iraq invasion. Media, Culture & Society, 33, 1185-1201.

34.

Pfau

Haigh

Gettle

Donnelly

Scott

Warr

Wittenberg

(2004). Embedding journalists in military combat units: Impact on newspaper story frames and tone. Journalism & Mass Communication Quarterly, 81, 74-88.

35.

Rattliff

J. R.

(2001). Economic news and public perceptions of the U.S. economy (Unpublished doctoral dissertation). Duke University, Durham, NC.

36.

Scheufele

(1999). Framing as a theory of media effects. Journal of Communication, 49(1), 103-122.

37.

Schudson

(2003). The sociology of news. New York, NY: W.W. Norton.

38.

Semetko

H. A.

Valkenburg

P. M.

(2000). Framing European politics: A content analysis of press and television news. Journal of Communication, 50(2), 93-109.

39.

Soh

S. C.

(2009). The comfort women: Sexual violence and postcolonial memory in Korea and Japan. Chicago, IL: University of Chicago Press.

40.

Tadaki

Akizuki

Arima

Kawano

Hata

(2014). Details of exchanges between Japan and the Republic of Korea (ROK) regarding the comfort women issue: From the drafting of the Kono statement to the Asian Women’s Fund (Japanese government report). Retrieved from http://www.mofa.go.jp/mofaj/ (In Japanese)

41.

Takekawa

(2014, August 23). Limitation of editorial journalism; “comfort women issue.” Paper presented at the Japan-Korea International Symposium, Tokyo. (In Japanese)