Seeking the True State of Online Opinion


Online discourse is often disproportionately represented by a repetitive stream of extremist comments. To determine what online opinion really is, it is necessary to take a closer look.

Twitter Big in Japan

In Japan, Twitter plays a major role in both social sharing and the mass flaming attacks known as enjō. According to data from June 2014, there were an average of more than 500 million tweets worldwide each day. Japanese tweets accounted for 87 million, or 16% of this total; more than half of all Japan’s digital natives in their teens and twenties use the site. Twitter is also a significant hub for encountering news stories. Some 43% of all Japanese people aged 16–24 click on news tweets they see in their timeline to view the full stories. If we want to know the reactions of mainly young users to topics and events they see on the net—in other words, online public opinion among Japanese digital natives—Twitter is a key venue to examine.

In the 2010s, there has been continuing empirical research on enjō and nationalist sentiment on Twitter. In their 2016 book Netto enjō no kenkyū (Research on Enjō Flaming), Tanaka Tatsuo and Yamaguchi Shin’ichi presented estimates on enjō on Twitter based on a large-scale web survey with more than 20,000 respondents. At the time of the survey, 1 in 200 Japanese Internet users were estimated to be active participants in enjō, with about 2,000 of them taking part in each attack. Around 90% of participants made only one comment, while the number of users who wrote several times during one incident or made direct personal attacks was limited to the single or double digits.

Taka Fumiaki’s 2015 work Reishizumu o kaibō suru (Dissecting Racism) is based on the analysis of more than 100,000 Japanese tweets related to South Korea expressing nationalist sentiment. Of the 43,000 accounts these came from, nearly 80% made only one tweet in the collection. By contrast, 471, or just over 1%, made over 100 tweets; the top 50 accounts were responsible for one in eight of the total, most of which were clearly discriminatory. Some tweets expressing strong discrimination tend to be posted and retweeted over and over again, thus having considerable impact on those who have a tendency to see such tweets. Research in the field indicates the distinctive feature of Internet discourse that continued repetition of extreme views from a small number of users can give the impression that they are a far greater online presence than their actual numbers.

Listening to All Online Users

The predominantly expressed views are generally held to represent online opinion, but these views can be swayed by a tiny minority of users. Does this mean that online opinion is biased and far removed from public opinion in wider society?

Most people post on social media from time to time, talking about their thoughts and feelings in their own words. Rather than limiting the definition of online opinion to include only extreme views, I think it is necessary to carefully survey people’s individual posts, focus attention on their wide variety of opinions and feelings, and take into account the actions they take, including retweets and likes. This establishes a fuller, more complex view of online opinion.

As evidenced by Brexit, the election of Donald Trump, and the 35% share of the French presidential election vote for far-right leader Marine Le Pen, nationalism is rising in developed countries. At the same time, traditional opinion polls are not functioning accurately in the offline world. Some have pointed at the way opinions are formed online as a factor behind this disconnect. But if only a small percentage of net participants are responsible for the bulk of the extreme views posted, we should be able to write them off as a mere fraction of the whole and expect the majority of people’s opinions and voting preferences in the offline world to remain more in line with general societal norms.

A dispassionate observation of recent phenomena, however, shows that online opinion reflects social trends to a certain extent. At the same time, though, we must not overlook the imbalance created by repetition of violent and insulting language.

Nor must we overlook the fact that Japanese social media includes numerous posts related to China and South Korea, history, race, territory, and nationalism. There is a powerful tendency here of seeking a social identity for Japan and Japanese people. While putting neighboring countries in an outside group, these netizens seek to clarify and fortify the in-group mentality.

Irritation Aimed at Minorities

When I look at large numbers of comments posted on Twitter, Facebook, and other social media, I gain a strong impression of irritation toward the vulnerable and social minorities. Rather than showing concern for their difficulties, many posters express annoyance at those who identify as minorities, seeing this as just a way to gain additional rights and compensation.

Criticism is aimed at a broad range of topics, including welfare recipients, the public annoyance posed by baby strollers, lenient treatment of young offenders under the Juvenile Act, the rights of LGBT and disabled people, and historical and territorial issues related to Okinawa, China, and South Korea. The intolerance evident in these critical posts is not isolated; it is, indeed, a manifestation of social attitudes that continue to grow harsher and more broadly held in Japan, as in other developed countries.

I would like to call this attitude “nonminority politics.” It is practiced by members of a nonminority who, though being part of the majority, feel they do not enjoy sufficient benefits despite that status. They are vehemently disparaging of liberal political views that support minorities, and demand to receive “fair treatment” in the form of more benefits for the majority to counterbalance those they see as overwhelmingly accruing to a minority. This spirit is greatly active in the venues where online opinion is formed.

I am currently conducting research based on the cultural psychological method of moral foundations theory to consider the social movements behind nonminority politics. In an Internet usage survey that I carried out, I included questions to determine the political outlook of respondents. My results showed that 68% could be classified as conservative, 24% as liberal, and 9% as libertarian. The proportion of liberals dropped away sharply among people in their twenties and thirties.

The table below shows that at least 60% of both conservatives and liberals think that Japan should always reflect on its actions in World War II. I would like to note that at the same time, around 80% of both groups believe that it is not necessary to keep apologizing in the third and fourth generations and that continual expectation of apologies is going too far.

Attitudes to World War II by Political Stance

Conservative (%) Liberal (%) Libertarian (%) Total (%)
Japan should always reflect on its actions in World War II. 60.8 70.2 43.0 61.4
There is no need to continue apologizing for Japan’s wartime actions three or four generations later. 78.4 78.3 60.0 76.7
Countries that expect Japan to continue apologizing forever for its actions in World War II are going too far. 81.3 81.4 61.0 79.5

Percentages include those respondents who stated they “strongly agree,” “agree,” and “basically agree” with the given statements. Political stances were established based on moral foundations theory.

In the undercurrents of online opinion, some see a strong resistance to continued demands for apologies as a basis for negative Japanese feelings about China and South Korea. The survey results support this view. Through moving analysis forward, it is possible to see how online opinion reflects society and to confirm that the Internet is not a venue that merely presents the extreme views of a small percentage of people. Keeping one’s eyes and ears open to online opinion offers a great opportunity for rethinking society.

(Originally published in Japanese on July 5, 2017. Banner photo © Aflo.)

media Internet