The riots in Capitol Hill are a wake-up call for American society-it stems from the fact that some right-wingers no longer believe in the legitimacy of the entire democratic system, especially the election results. This situation is not only caused by Trump’s speech on the day of the riots in the Capitol, but also by his tens of thousands of false or misleading statements in the past four years and social media platforms that have contributed to this.
American political scientists Jacob S. Hacker and Paul Pierson pointed out in their book "Governing the Country by Twitter" that the right-wing media has made "extraordinary achievements" in squeezing conservative audiences into the narrow, extreme and exclusive information space and leading the Republican Party to extremism and tribalism. "From the beginning, a prominent feature of the right-wing media was that it invested huge energy in attacking other channels of information and knowledge sources, especially traditional media," the authors believe, which played a crucial role in creating a loyal audience isolated from the outside world and strengthening political polarization. The rise of social networks has accelerated this process. According to Article 230 of the Communication Regulation Act of the United States, the Internet platform does not need to bear any responsibility for the comments made by users or third parties on the platform, and there is no longer a "gatekeeper" to screen the complicated network information, which puts forward strict requirements for people to distinguish the truth from the false, and also greatly amplifies the voice of the extreme right wing.
After the riots on Capitol Hill, major social platforms and technology companies in the United States quickly took measures to ban the accounts of Trump himself and his supporters. This incident has caused different reactions in the United States: the left thinks that technology companies have done "too little too late" in regulating online speech-during the 2016 US presidential election, Facebook was criticized for allowing the proliferation of fake news reports, and Franklin Foer, former editor-in-chief of New Republic magazine, criticized that "the right-wing conspiracies that were created by force were endless, which pushed up Donald Trump’s candidate public opinion". At the same time, the right accused technology companies of being held hostage by left-wing ideology.
"Banning Trump" has also caused controversy internationally. German Chancellor Angela Merkel stated that Twitter’s practice was "problematic" and stressed that the relevant laws to ban social media accounts should be formulated by the government and should not be fully controlled by private companies. European leaders such as Merkel’s criticism of technology giants reflects the different contexts and ideologies that Europe and the United States face on the issues of speech censorship and platform governance. However, an unavoidable fact is that transnational social platforms have become one of the indispensable "infrastructures" in contemporary life. How to keep a balance between freedom of public expression, breaking the information cocoon and reducing conflict damage will be a long-term global challenge.
Recently, Interface Culture (ID: Booksandfun) invited Xu Xibai, a doctoral candidate in politics and international relations at Oxford University, and Liu Lizhi, an assistant professor at Georgetown University Business School in the United States, to discuss the above issues in the form of a round table. There are four main topics in this round-table discussion: first, we discussed whether the riots on Capitol Hill will have an impact on the Republican Party’s future campaign strategy and political policy; Secondly, from the perspective of "banning Trump", this paper discusses the boundary issue of freedom of speech, especially the influence of internet monopoly giants on it; In the second half of the article, the two scholars discussed the role of social networks in promoting political polarization and various disputes on global Internet platform governance.
Interface culture: During last year’s general election, xu teacher talked about that the Republican Party will further slide to right-wing populism, and the Democratic Party will be very passive in the next decade, predicting that Trump will not admit defeat. However, the riots on Capitol Hill probably exceeded many of our expectations. After this incident, how will it affect the campaign strategy and political policy of the Republican Party? Will the Republican Party give up with the "right-wing populist forces" represented by Trump?
Xu Xibai: The unanimous condemnation of the riots on Capitol Hill by the mainstream media in the United States has obscured the real public opinion to some extent. According to a poll conducted by YouGov on January 7th, among the Republican respondents, 68% thought that the riots on Capitol Hill would not pose a threat to American democracy, 45% supported demonstrators to break into Congress, and 69% thought that it should not be blamed on Trump. Among the Democratic respondents, these three figures are only 4%, 2% and 4%. It can be seen that the evaluation of the riots on Capitol Hill is still highly polarized by political parties, and condemnation is far from the consensus of the American people. Axios-Ipsos poll on January 13th showed that among Republican respondents, 67% supported Trump’s recent move, and 57% thought Trump should be the presidential candidate in 2024. These figures show that Trump is still the undisputed number one politician in the Republican Party.
Factions within the Republican Party that opposed Trump and adhered to moderate conservatism finally had a chance to speak out after the riots, but they may not be able to influence the party’s line. The defeat of the Senate election in Georgia has shown that once the Republican Party is infighting and loses Trump’s "blessing", then the assured seat will also be lost. Alienating Trump supporters is undoubtedly a political suicide in some constituencies. The American electoral system can only create bipartisan politics, and small parties have little room for survival. If the "moderates" resolutely give up their seats and form a party independently, then they will be marginalized, not the "populists."
The same is true in practice. As many as 138 (more than 65%) of the 211 Republican congressmen voted against the certification of the election results, which is very close to the proportion that 75% of Republican supporters in the poll thought Biden was illegally elected. In the second impeachment vote of Trump in the House of Representatives, only 10 Republican congressmen voted in favor. Except for a few people who have long opposed Trump, such as Romney, most Republican politicians are still excusing him. In calling for "unity" these days, they just want to "roll with the punches and roll with the punches". The right-wing media continued to advocate conspiracy theories, claiming that the mole who broke into Congress was the left-wing organization Antifa, and the Twitter title was that the technology giant and the left conspired to persecute Trump. Even if Trump himself fades out of politics, these conspiracy theories and "Biden stole the election" are still part of the Big Lie "he spread, and will remain an important public opinion tool to mobilize Trump supporters in the next few years. It is impossible for the Republican Party to give up these voters, and its trend of continuing to slide to the right-wing populism will not change significantly.
Liu Lizhi: After the election, the Washington Post had a commentary entitled "Trump will go, but Trumpism will stay". This sentence can be used to answer the question of "Will you leave?" It also applies to the riots on Capitol Hill.
As Xi Bai’s analysis above shows, Trump doctrine will haunt the Republican Party for many years due to election pressure and the composition of voters. The background is still the growing political polarization since the mid-1990s. Before the political polarization, because there were a large number of middle voters, both parties needed to please and attract middle voters in order to win the general election, which made the policy propositions of the two parties similar and the moderate middle politicians had more opportunities. However, in recent years, due to political polarization, there are fewer and fewer middle voters, and the two camps, the Democratic Party and the Republican Party, have become more homogeneous and heterogeneous. In addition, the voter turnout rate in the American general election has been low in western democracies (66.7% in 2020), and both parties have begun to pay more attention to encouraging voters in their own camps to vote, rather than attracting voters from the other camp or the middle.
Trump’s strategy in the past four years is to give up uniting with the masses, not be afraid of division, and mobilize and consolidate his voter base through controversial remarks; In fact, if it is not because of the epidemic, Trump’s chances of re-election by this strategy are not small. This time, the reason why the Democratic Party was able to turn the "traditional red state" of Georgia blue (supporting Biden and winning two key seats in the Senate) was also due to Stacey Abrams’s grass-roots strategy, that is, to help the bottom voters solve their voting difficulties and let more people from their own camp come out to vote to enrich their votes, instead of focusing on attracting middle voters. With the promotion of S. abrams’s position in the Democratic Party, this strategy is likely to be used as the future strategy of the Democratic Party in the whole South. This strategy of focusing on one’s own camp and fighting for voter turnout will make Trumpism continue to exist in the Republican Party, because moderate slogans may not be able to mobilize voters.
Of course, if the timeline is long enough, the trend of the Republican Party sliding to the right-wing populism may not be irreversible. In fact, there are many cycles in American history, and there have been many political polarization phenomena (such as the end of the 19th century and the 1950s and 1960s), and there have been many torn periods. Populism is not the first time to emerge, and even there has been a civil war, but it has since healed itself. American society has its own resilience. Only this time, I don’t know how long the repair process will take. If the new Biden administration can handle the epidemic and economic problems well, the cracks may gradually begin to heal.
Xu Xibai: Lizhi’s analysis is very reasonable. I would like to add that the Republican Party has realized that Trumpism can attract some middle-class voters, such as blue-collar workers in 2016. Many of them voted for the Democratic Party all their lives, but they were captured overnight by Trump’s "Make America Great Again" discourse. Therefore, the polarization we are talking about is not that the supporters of the two parties turn left and right at the same pace, but that many people jump directly from the center left to the far right. For example, after the outbreak of the epidemic last year, the search volume of the conspiracy theory "Q Anonymous" increased several times, attracting many middle-class voters who were not interested in politics before. They just tried to find answers and comfort through the Internet under the anxiety caused by the epidemic, and the result was bewitched.
Interface culture: In the book "Ruling the Country by Twitter", American political scientists Jacob Huck and Paul Pearson pointed out that European and American democracies faced the dilemma of the Conservative Party from the beginning of their establishment, that is, how the elites defended their political and economic interests in front of the people who gradually gained political power, and once they found that they could not realize their demands through normal democratic procedures, they would destroy democracy. Since left-wing parties usually win on economic issues, the Conservative Party must introduce or emphasize other social differences if it wants to turn over. This cultural differences strategy is exactly what the Republican Party has relied on in the past 40 years. The riots on Capitol Hill may be an event that completely exposed the current social tear in the United States. What will it leave in American history?
Xu Xibai: It is still too early to make a historical evaluation. However, this is indeed the closest moment for the United States to overthrow legitimate elections and subvert the democratic system since the Civil War. Schwarzenegger, a former governor of California and a Republican, compared it to the "night of broken glass" in Germany in a recent video (note: Nazi party member and the SS raided Jews all over Germany from November 9 to 10, 1938, which was regarded as the beginning of the Nazi organized massacre of Jews). In the past four years, Trump has constantly established a cult of personality, attacked media supervision, and weakened people’s belief in the legitimacy of the democratic system. Fortunately, his incitement and organizational ability are far inferior to those of the great dictators in the 20th century, while the balance mechanism of power in the United States can still play a role and avoid a greater crisis. Otherwise, this incident may really become the American version of "Night of Broken Glass".
Liu Lizhi: Yes, it is too early to make a historical evaluation, but the riots on Capitol Hill are different from ordinary demonstrations and riots, because they are not aimed at a specific policy or a politician, but originated from some Trump supporters questioning the legitimacy of the entire democratic system, especially the election results. Although after Trump was elected in 2016, large-scale demonstrations broke out in various parts of the United States. Many protesters used the slogan "Trump is not my president", but they opposed Trump himself, not thinking that the election was fraudulent and unfair. What they hoped for was to send Trump out of the White House through votes four years later, rather than seeking means other than elections. It is precisely because of the challenge of the riots on Capitol Hill to the democratic system itself, as Josep Borrell, vice-president of the European Commission, said, the Capitol Hill incident is a "wake-up call" for all democratic countries, reminding people of the dangers caused by letting democratic values slide.
Interface culture: A few days ago, Trump’s social account was banned by social media platforms such as Twitter and Facebook, which caused a debate about whether this move violated Trump’s freedom of speech. Earlier, it has been pointed out that the First Amendment of the US Constitution protects citizens’ freedom of speech that is not restricted by the government, while social media platforms are private enterprises, not the government, so the First Amendment does not apply to this situation, so this is not a question of freedom of speech. What do you think of this argument?
Xu Xibai: There are several issues that need to be clarified: First, Trump’s live speech on the day of the riots is not a statement protected by the First Amendment. The First Amendment stipulates that Congress shall not legislate to restrict citizens’ freedom of speech, but it is vague about what kind of speech is protected. Historically, it has always relied on the precedent of the Supreme Court as a criterion. In terms of hate and violent speech, the standard of "immediate illegal behavior" established in the case of Brandenburg v. Ohio in 1969 is to evaluate whether the relevant speech "aims at inciting or producing immediate illegal behavior" and whether the speech "is likely to cause such behavior". Some people may say that Trump did not directly instruct his supporters to attack Congress. We can refer to john mill’s "injury principle". In On Freedom, Mill gave an example: publishing articles in newspapers criticizing grain merchants for starving the poor should not be restricted. However, if a large number of angry people have gathered in front of the grain merchants, it is very problematic to make these remarks to them at this time, because it is "actively inciting some kind of disaster" and may lead to personal safety of the grain merchants. Trump’s call for supporters to "fight like hell" at the scene is in line with Mill’s "injury principle" and the Supreme Court’s "immediate illegal behavior" standard.
Liu Lizhi: Yes, by the way, freedom of speech does not mean that all speech will be protected. The reason is simple: in a social group, if everyone does whatever they want and talks endlessly, there will be endless conflicts and society will not be organized, so freedom also has boundaries. The freedom advocated by Mill is actually quite broad. "As long as it does not involve the interests of others, others have no right to interfere", but even the freedom defined by Mill has the premise of "not hurting others". As Oliver Wendell Holmes Jr, Justice of the Supreme Court of the United States, famously said, "Even the strictest protection of freedom of speech does not protect those who lie about fire and create chaos in the theater." Therefore, when we judge whether freedom of speech is infringed, we must first look at whether a certain speech is within the scope of protection.
Xu Xibai: The second question is: The First Amendment does not apply to private enterprises such as Twitter. Their actions to ban Trump accounts are legal and in line with Article 230 of the US Communications Regulation Act. This provision stipulates that the Internet platform does not need to bear any responsibility for the comments made by users or third parties on the platform. At the same time, out of goodwill, the platform can restrict "obscene, obscene, dirty, excessively violent, harassing others or other unpleasant remarks", regardless of whether these remarks are protected by the Constitution. This can be regarded as a two-way exemption clause, which gives the internet platform great censorship power.
The third question is: I have seen many comments that since the First Amendment does not apply, this is a question of whether Twitter is illegal, not a question of freedom of speech. This accords with a popular view that freedom of speech is aimed at public power, which means that the government cannot restrict citizens’ speech. I call this view "the doctrine of the First Amendment", and its interpretation of freedom of speech is too one-sided. In addition to public power, all kinds of private power, such as private enterprises, religious authorities, non-governmental organizations, family forces and even terrorists, may infringe on people’s freedom of speech. The terrorist attack on Charlie Hebdo in France in 2015 is an example (forcing others to silence with violence or threats of violence). Twitter banned Trump, of course, is a question involving freedom of speech, and it involves the core issue of freedom of speech, that is, under what circumstances can private power reasonably restrict people’s speech. As far as Trump’s case is concerned, his remarks are not protected in the legal sense or other senses. Twitter banned his account and did not infringe on his freedom of speech.
Liu Lizhi: I agree with the above views of Xi Bai. Let’s break away from Trump’s case and further discuss the relationship between platform ban and freedom of speech in a general sense.
Xi Bai said that we should be alert to the "doctrine of the First Amendment", which I deeply agree with. It is not only public power that will damage freedom of speech; Just because private power is not illegal does not mean that it has no impact on freedom of speech. But I also want to add that another dogma is not desirable, that is, any normal market behavior will be deemed to damage freedom of speech.
General non-monopoly private enterprises are actually more difficult to infringe on individual freedom of speech. For a simple example, if a left-winger in the United States applies to the right-wing media to speak, the right-wing media is likely to reject him, but this is a normal market choice behavior, and it is hard to say that this move hurts this person’s freedom of speech. General non-monopoly private enterprises have effective control over the market and fierce competition among them. If one company refuses to provide services, people can try another and often find a substitute.
But big social platforms like Twitter and Facebook are different from ordinary private enterprises. These large technology platforms are special not only because of their large scale and high market share, but also because of the particularity of the business form of "platform": because of the network effect of platforms, in the digital economy, we often see a phenomenon of "winner takes all", and finally there are one or several super-large platforms left in the market. It is precisely because users are clustered on the "super platform" and users are highly sticky that it is difficult to find substitutes in the field of social media. Once someone is expelled from a super platform, the communication power of his speech will be greatly reduced, and the freedom of speech in a broad sense will be affected. For example, in theory, a person who is blocked on Twitter and Facebook can switch to other small platforms, but the audience will be much less, and the spread of speech will be limited.
However, it should also be noted that public power and social media platforms have different restrictions on speech. In essence, the platform ban is to take back the megaphone for personal information release (in the absence of a platform, most people are just the recipients of information), and the blocked individuals generally do not have to bear legal responsibility, and the platform’s restrictions on speech are limited to specific platforms. That is to say, the same speech can still be said in another place, but the power of speech communication is often limited. Therefore, although we really need to discuss the ban of platforms under the framework of freedom of speech, we should also pay attention to the differences.
Interface culture: As Teacher Liu said, when these platforms decided to expel some people, it did affect people’s freedom of speech to some extent. Is this an unprecedented new situation at the legal level, and is the current law out of date?
Xu Xibai: An important function of freedom of speech is to enable people to participate in the discussion of public affairs. If cyberspace is a "digital commons" where people gather and discuss public affairs, then private platforms are private public spaces (POPS) drawn from this commons. Facebook has 2.7 billion active users around the world and more than 300 million Twitter users. It can be called a virtual country that spans countless physical countries. Their management is like a virtual government in these virtual countries. Their terms of service and algorithms have a greater impact on freedom of speech than the laws of any country. In view of the transnational monopoly of Internet platforms, once users are banned by large platforms, they lose the right to participate in public discussions in the digital commons, which is also an unprecedented new phenomenon. How to regulate the influence of these Internet monopoly giants on freedom of speech is a new challenge for legislators.
Liu Lizhi: This phenomenon does pose some new challenges. What is new is not that public power is not as good as that of private power. Since ancient times, human beings have had a lot of private rights that have influenced and restricted personal behavior and speech; Especially where the state power and law are not as good as each other, many private powers can provide an alternative governance system (an extreme example is that the East India Company once ruled the Indian mainland under the authorization of the British royal family, owned the army and could also conduct diplomacy). What is new is that the private governance system of digital platform relies heavily on data and algorithms. Data has always been one of the factors of production in human history, but it has never been so important as in the era of digital economy. The biggest impact on the current legal framework here is the property rights and sovereignty of data.
For example, the property rights of data. The speech on the platform is also a kind of personal data in essence. But when this data is produced, does it belong to individuals or platforms? The reason why the platform has the power to delete comments is because the platform now owns this data (this part of the property right has been transferred when the user registers and signs the consent form). In addition, the platform will also collect users’ personal information, and make profits through targeted advertising or reselling user information to third parties. In other words, the "freedom of speech" of all users on the platform is not a free right, but is exchanged by users with personal information. But is this property right design reasonable? Can it really promote social welfare? Some scholars believe that data property rights should be owned by users/consumers rather than companies. At present, the great power of platform companies is based on almost unlimited property rights of personal data, but if the property rights of data change in the future, the power of the platform will also be restricted.
The other is the issue of data sovereignty. For example, the ban on American users by the US platform does not involve sovereignty issues. However, because many American platforms are global, if foreign users are banned, diplomatic disputes may be involved. Previously, Xi Bai mentioned that platforms such as Facebook are "virtual countries". The challenge to the current legal system here mainly lies in the fact that the territory of these "virtual countries" does not coincide with the territory of the entity countries. However, this state will not last forever, because many countries are now establishing digital border, striving to establish sovereignty and control over platform data, cloud services and algorithms within the country.
I recently wrote an article "The Rise of Data Politics: Digital China and the World", in which I made a preliminary study on data sovereignty and China. At present, China is also actively establishing data sovereignty, but China’s data sovereignty is incompatible with some countries’ extraterritorial jurisdiction over data (such as CLOUD Act in the United States), which may also lead to conflicts in the future. In addition, the rise of "data sovereignty" in other countries will also create uncertainty for the overseas expansion of China’s science and technology enterprises.
Interface culture: Intellectuals have long discussed the negative effects of social networks in the post-truth era on the democratic political ecology, which should be an important prerequisite for this ban on Trump. One question I am interested in is, to what extent will social media affect people’s political views, especially the extreme right-wing political views?
Xu Xibai: In 2006, when social media was still in its infancy, American scholar Cass Sunstein pointed out in his book Infotopia that people are more concerned about what they like and are more willing to communicate with people with similar intellectual interests. The same is true of online media, forums and blogs. Web links generally point to similar websites, which makes the communication within the community smoother, but blocks the communication between communities with different opinions, forming an "information cocoon", which even exceeds the traditional media era. This magnifies the "groupthink" that people tend to "follow the crowd" and are unwilling to question the mainstream views of people around them, which intensifies the polarization. Pariser put forward a similar "filter bubble" mechanism in 2011, that is, Internet platforms use algorithms to filter and reorganize information, and guess and push what users want to see, such as personalized search results of Google, personalized dynamic news of Facebook, and personalized product recommendation of Amazon. As a result, users can’t see different views and information more and more, and they are isolated in their own cultural and ideological bubbles, which intensifies. The development of history has largely confirmed their predictions.
Different social media have different effects, which is highly related to the structure and algorithm of social media. A recent study by the University of Virginia found that after long-term use of Facebook, right-wing users read more right-wing news than before, while right-wing users who use Reddit for a long time will read more neutral news. The reason is that Facebook is an "acquaintance network", and users can only see the sharing of friends or friends of friends. These people’s views are usually similar, and channel subscription is basically one-to-many one-way information dissemination, which is easy to cause the "information cocoon room" effect. Reddit is a "stranger network". The sections with the largest number of users on the platform are divided according to interests and topics, and the views of posts in the sections tend to be diversified. The praise mechanisms of the two platforms are also different. Facebook can only like it, but can’t object to it. As a result, most of the high praise posts are extreme, and these posts are highly likely to be shared and recommended. Reddit’s praise is mainly the recognition of the post’s quality and topic interest, and users can like it or object to it. Posts with extreme views are more likely to be checked and balanced by negative votes, which helps to appear high-quality posts with diverse views in the user’s information flow.
There are also some studies that the Internet and social media are not the main reasons for polarization. A questionnaire survey of 36 countries in the 2017 Digital News Report of Reuters School of Journalism, Oxford University shows that even in countries with similar technological development and Internet popularity, the degree of polarization is quite different. For example, Norway’s social media usage rate is close to that of the United States, slightly higher than that of the United Kingdom, but the polarization of Norwegian news media is very low, basically in the middle, and the news reading of Norwegian netizens is also concentrated in these media with middle views. British news is more partisan and polarized, but the most visited news website is still the BBC with a middle political view. The views of the left and right media in the United States are the farthest apart, and the degree of polarization is the highest. There are almost no large-scale centrist media, and the news reading of American netizens is also concentrated on the left and right ends of polarization.
The report does not further elaborate the reasons behind this phenomenon, but I think we can make some reasonable assumptions: First, polarization may be related to economic and ethnic inequality. The emergence of polarization requires one or more issues with fierce conflicts and huge controversies, tearing people into two groups with opposite views and worldviews. Norway’s society is more equal, people’s dissatisfaction and anxiety are lower, and there are fewer conflicts between different ethnic groups, so there is no social soil for polarization. Secondly, polarization may be related to the media systems of various countries. Norway adopts the corporatism mode of deep government intervention and strong supervision, and it is difficult for the media with extreme views to survive. Britain and the United States also implement the free market system, but the United States abolished the principle of "fair reporting" in the Reagan era, and did not restrict media mergers and acquisitions and monopoly; In Britain, there is also a BBC funded by the government, which is independent of party influence, neutral in view, and can be called a giant in scale and influence, which has suppressed the polarization of the media to some extent. In this way, social media may just add fuel to the fire on the basis of the polarization that has already formed.
Interface culture: How should we respond to the right/Trump supporters’ view that banning Trump is a technology company being held hostage by left-wing ideology?
Xu Xibai: The right-wing conspiracy theory of "persecution" has been popular for decades. The Federal Communications Commission of the United States adopted the principle of "fair reporting" in 1949, requiring the media to be honest, fair and balanced when reporting controversial events, and to show the opinions of both sides. Conservatives have been fiercely criticizing this as discrimination and suppressing their views, which eventually led to the abolition of the principle of "fair reporting" in 1987. The accusation of "bias" in the Internet age is very popular, and another reason is that Internet giants have never disclosed their own algorithms and censorship records. So far, however, this accusation has only been proved by a few cases. There are also many examples of left-wing speeches and accounts being banned, and there is no evidence that social media systematically suppresses right-wing views. If the research conclusion quoted above holds, then the situation is just the opposite. Facebook is a paradise for right-wingers, and their voices have not been suppressed, but have been amplified.
This controversy is also reflected in the attitude of Republicans and Democrats towards Section 230. Both parties advocate amending this clause, but the reasons are diametrically opposite. Republicans believe that technology companies abuse the second half of the clause and impose too many restrictions on right-wing speech; The Democratic Party believes that technology companies have abused the first half of the clause, using the excuse that they cannot be responsible for users’ comments and failing to fulfill their censorship responsibilities, which has led to the spread of false news and conspiracy theories on the right.
Liu Lizhi: The current situation of the science and technology platform can be summarized by a common saying: "Left and right are not people". As Xi Bai said before, many right-wingers think that "banning Trump" is to suppress freedom of speech, while many left-wingers think that "banning Trump" and rumoring have come too late and not done enough. In fact, for the science and technology platform, it is not in its commercial interests to be biased towards any party; Too strict and too loose platform audit scale will lead to the loss of users.
The reason for "banning Trump" this time is beyond the left and right ideology. The reason given by social media platforms is not that Trump’s speech violates a specific ideology, but that his speech is in danger of inciting violence. After all, not only Democrats but also Republicans are besieged on Capitol Hill. Some Trump supporters entered Congress shouting "Hang Burns". In fact, every society needs to maintain some "common bottom line" that transcends the left and right ideologies, otherwise the society cannot function. One of the common bottom lines is to oppose violence.
Interface culture: For a long time, technology companies were reluctant to implement "speech censorship". Trump has spread many rumors and conspiracy theories since he became a Republican presidential candidate and was elected president, and has not encountered any obstruction or clarification from the platform. It was not until May 27, 2020 that Trump claimed on Twitter that "mail voting is undoubtedly fraud", and Twitter began to verify Trump’s remarks; After the riots on Capitol Hill, Twitter finally decided to permanently ban Trump’s account. Is it too late to ban Trump at this time, and where is the scale for evaluating his crossing the line?
Xu Xibai: Let me start with the question of whether technology companies should conduct censorship. A traditional view holds that the best way to deal with false and malicious speech is not to censor it, but to counterspeech, putting everything in the sun and refuting it with facts and evidence. In the era of traditional media, this "the more truth is debated, the more it becomes clear" is convincing, because at that time, information was scarce, the number of media was limited, and the entry threshold was high, which helped to start a debate. But in the internet age, information is not too little, but too much, and what is scarce is the attention of the audience. The makers of fake news and conspiracy theories just take advantage of this. They don’t need to confront the real information at all, but only need to stimulate people’s emotions and use strategies such as water army, sprayer and social bots to drown the real information with massive false and low-quality information. Therefore, it is indeed necessary for social media platforms to conduct censorship and intervention.
Specific to Trump, his account has repeatedly violated the terms of service such as "beautifying violence", "hate speech" and "civic integrity" (spreading false news about the epidemic and unsubstantiated accusations against elections), which should have been banned long ago. However, government officials enjoy immunity on Twitter, and tweets that violate the terms are generally marked and reserved for public interest. Jack Dorsey, CEO of Twitter, has always opposed the ban on Trump. He believes that world leaders should be allowed to tweet to encourage public discussion, even if some remarks are disgusting. The decision to permanently ban this time was made after assessing that Trump’s follow-up remarks may continue to incite violence. Dorsey also issued a review on January 14th, saying: "This ban shows that we finally failed and failed to promote a healthy dialogue.".
I think the reason for the ban is reasonable, but it is really difficult to calm people’s doubts at this time. Why is the policy so inconsistent before and after? Did you see that Trump didn’t use value before deciding the title? It can be said that this case embodies all kinds of problems in regulating social media speech: the censorship standards vary from person to person, can not be completely fair, and are influenced by the personal preferences of platform executives; The formulation and implementation of rules are completely opaque, and users cannot be held accountable and can only passively accept them; Users with restricted speech lack effective appeal and relief channels.
Liu Lizhi: Shi Cai Xi Bai said that the scale of social media supervision is opaque and can’t be accountable. In fact, it can be abstracted into a theoretical question: Who supervises? This question used to be directed at government power: who will supervise the government as a supervisor? The conclusion is that the government itself needs to be bound by the system (such as public opinion supervision, internal and external checks and balances); Only by putting power in a cage can corruption and dereliction of duty be reduced. Similarly, when the social media platform has huge supervision power over the accounts on the platform, a natural question is: Who will supervise the platform as the supervisor? How to restrict the power of the platform? What kind of scale is appropriate?
There are several possible directions.
First, rely on the platform for self-discipline and self-supervision. Some people may ask: Why should private enterprises manage themselves? In fact, there are many documents in management that prove that private enterprises have the motivation to self-regulate, because self-regulation of enterprises can reduce the possibility of direct supervision by the state, and once the state takes the initiative, the supervision is often stronger and the impact on private enterprises will be greater. But it is precisely because of this that even if enterprises can seriously self-regulate, their regulatory standards will be much looser than those directly regulated by the state. For example, the Internet in the United States has now issued some supervision and marking for rumors, but it can be seen that this degree of supervision is far from enough.
Second, determine the regulatory scale of social platforms through government legislation. Because the "First Amendment" is deeply rooted in the hearts of the people in the United States, I am afraid that it is difficult to have such direct legislative restrictions on speech as the Hate Speech Law of European countries in the near future. It is more likely to cancel Article 230 (the exemption right of Internet companies). Once cancelled, Internet companies need to take responsibility for the content published by users on the platform, which will prompt the platform to adopt stricter censorship standards.
Third, I think the interesting new direction is to establish a private legal system on the platform, and hand over the difficult matter of censorship to independent institutions outside the platform, such as randomly selected user juries or independent legal experts. The advantage of this approach to the platform is that because the scale is not drawn by the platform, the platform is not responsible regardless of the scale.
China is the first place where this kind of online jury practice appeared. For example, I have done research on Taobao’s public jury (Liu and Weingast 2018). This mechanism was founded in 2013 and was originally inspired by Super Girl’s public review mechanism. Once there is a dispute between the buyer and the seller, Taobao will randomly select 31 public review (later the number will be reduced to 13), and then public review will vote according to the evidence presented by the disputing party, and the party with a simple majority will win. At present, this model has solved about 16 million trading disputes on Taobao. Later, this jury mechanism was used by Taobao to judge more complicated issues, such as shanzhai brand, and other platforms began to use similar mechanisms to determine the judging criteria. For example, WeChat used public review to judge whether an article was "redacted". Besides public review, the platform can also judge the scale through an independent team of experts. For example, in 2020, Facebook set up its own "Supreme Court" and nominated 20 people, including law professors and Nobel Prize winners, to judge the scale of content censorship. Recently, Facebook handed over the decision of "whether to restore Trump account" to the "Supreme Court", which will make a decision within 90 days; Once Facebook’s "Supreme Court" makes a decision, neither Zuckerberg himself nor other Facebook executives can overturn it. The development of this private online legal institutions deserves close attention.
Interface culture: Merkel believes that "it is problematic for the account of the US President (Trump) to be blocked", pointing out that there is no "procedural justice" in the social media platform. An article by Quartz argues that although Merkel’s concern is reasonable, it is not in line with the legal reality of the United States, because it is impossible to ban accounts or specific speeches by legal means in the United States at present. Therefore, American technology companies are actually managing the comments that appear on the platform in the absence of a legal framework. On the issue of platform governance, are Europe and the United States at different stages and thinking about different issues?
Xu Xibai: The focus of Internet governance in Europe and the United States is very different. Europe pays more attention to privacy and data protection. Articles 7 and 8 of the Charter of Fundamental Rights of the European Union list "respecting personal privacy" and "protecting personal data" as basic human rights. The General Data Protection Ordinance (GDPR) passed in 2016 is the most stringent data bill in the world, which requires that personal data must be collected in a legal, fair and transparent way, and the way of data collection must be explained to users in concise language, and personal data must be anonymized or anonymized. The regulations also stipulate that users have the right to ask the platform to permanently delete personal data to protect the so-called "forgotten" rights. Snowden’s "Prism Gate", which allows the government to access personal data held by Internet companies at will and monitor personal communications on a large scale, is hard to imagine in Europe.
On the other hand, there is no tradition of the First Amendment in Europe, and the removal of online content for the sake of maintaining public order and good customs or protecting vulnerable groups is generally higher than freedom of speech. It is generally acceptable in Europe to block network content first without judicial procedures or informing users. For example, in the UK, most Internet service providers (ISPs) will block the content that operators deem illegal or inappropriate, such as child pornography. In the United States, this is likely to be accused of violating the principle of network neutrality. The Digital Services Act (DSA) currently being drafted by the European Union requires the head platform to remove illegal content after being notified by the authorities, and at the same time, provide a complaint channel for users whose content has been removed. It must maintain a high degree of transparency in algorithms and online advertisements, and allow researchers to obtain key data of the platform for supervision. Platforms that violate this Act will face a maximum fine of about 10% of their global operating income.
There is also very strict hate speech legislation in Europe, which basically complies with Article 20 of the International Covenant on Civil and Political Rights (hereinafter referred to as the "Covenant"). "Any advocacy of national, racial or religious hatred that constitutes incitement to discrimination, hostility or violence shall be prohibited by law." For historical reasons, some countries have "blasphemy laws" to protect religions, and countries such as France and Germany have further legislated to prohibit denying the Nazi Holocaust or selling Nazi-related books or souvenirs. The United States, on the other hand, is a signatory to the Convention, but when it ratified the treaty, Congress attached a reservation, that is, it cannot restrict the freedom of speech protected by the US Constitution on the grounds of Article 20 of the Convention. That is to say, hate speech is protected by the US Constitution as long as it does not meet the standard of "immediate illegal behavior".
Liu Lizhi: I very much agree with Xi Bai that the focus of Internet governance in Europe and the United States is different. In terms of Internet platform governance, Europe and America are not at different stages of the same path, but taking different paths.
Why do you take different paths? What Xi Bai said above is mainly that there are differences in political culture between Europe and the United States, such as the different priorities of freedom of speech. In addition, people in western Europe are more supportive of government intervention in the market. According to a survey conducted by the Pew Research Center in 2018, most people in Western Europe believe that government regulation of enterprises is generally beneficial to society (this ratio is particularly high in Britain and Spain, at around 70%). On the same issue, the American people are more divided (50%), and different parties are completely different. Although Democrats’ support for government intervention is close to that of other western European countries, only 31% of Republicans (including those who are biased towards Republicans) in 2017 believe that government regulation of enterprises is necessary to protect public interests. Compared with Europe, it is precisely because the United States generally respects small government and is more suspicious of government intervention, so the United States prefers market problems to be solved by the market, and industry self-discipline rather than government intervention, which is also reflected in the governance of the Internet industry.
In addition to the differences in the "underlying logic" such as political culture, the Internet governance models in Europe and the United States are different, and there are geopolitical considerations because of the uneven strength of the Internet industry. There are many global Internet platforms in the United States. The government can reduce the compliance costs of these enterprises, help them expand globally and further enhance their competitiveness. For Europe, strict management of science and technology platforms is a part of establishing European "data sovereignty". The popular technology platforms and digital services in Europe are almost all "made in America", such as Facebook, Google and Amazon. In addition, TikTok’s overseas version of Tiktok is also very popular in Europe, reaching 100 million users in 2020. Since most of the popular technology platforms in Europe are imported, it is in Europe’s own interest to control the platforms more strictly, and there are fewer internal obstacles. Merkel thinks that "there is something wrong with blocking Trump account", on the one hand, because she thinks that this decision should be made by national legislation, not the platform, but another hidden logic is that the power of the private platform in the United States is too large, which will affect Europe’s "data sovereignty". Imagine that if an American company can unilaterally delete the accounts of European politicians who do not conform to its platform rules, will it interfere with European national sovereignty? (Of course, if the EU requires the US Internet platform to do the same thing in accordance with EU regulations, this situation does not count). In the same vein, in recent years, European leaders such as Merkel and Macron have been calling for reducing data dependence on China and the United States, and launched Gaia-X Project in 2019.With a view to establishing a common digital infrastructure for EU countries.
Interface culture: If we want to control extreme speech on social networking platforms, another question is whether it is necessary to trace the responsibility of right-wing media. Many scholars have pointed out that the right-wing media played an important role in leading the Republican Party to extremism and tribalism-they attacked other sources of information and knowledge, such as traditional media, thus creating information cocoons-social networks are probably just the natural development result of this trend, just as the right-wing tendency of Fox News is not as good as the extreme right-wing new media Breitbart News Network.
Liu Lizhi: For institutional media, traditional judicial channels provide a means of accountability, and the injured party can sue the media. For example, after this election, conservative media repeatedly attacked the fraud in the voting counting system and deleted the votes supporting Trump in the equipment, which made Biden win the election. Smartmatic, the affected counting company, sent a warning letter to several right-wing media including Fox News in December 2020, threatening to take legal action. However, many rumors on the platform come from anonymous accounts, which is difficult to blame.
Xu Xibai: Fox News is just the tip of the iceberg. In the past decades, a large number of local radio stations, newspapers and online media have played a key role in polarization. In recent years, creating false news to influence and manipulate people is no longer limited to the media, but has formed a complete global industrial chain, but there is still little relevant legislation. At present, only France has passed the law against fake news, and it only applies to the news related to the election in the three months before the election. In 2018, the scandal of Cambridge Analytica broke out, and the company obtained the personal information of 87 million Facebook users, which was used for the accurate delivery of promotional materials (including many false news) in the Brexit and US presidential elections in 2016. At present, the lawsuit against it is limited to illegal access to personal information and illegal marketing. The legal act of legislation mainly refers to libel litigation. Here we should also mention the differences between Europe and the United States: compared with Europe, the legal environment in the United States generally protects the media more, and is even more unfriendly to libel plaintiffs. If the plaintiffs are public figures, it will be more difficult to win the case, because in addition to proving that the defendant’s statements are untrue, it is also necessary to prove that the defendant has actual malice when making false statements. If the subject of slander is difficult to determine (for example, "stars and senior officials worship Satan and even eat children"), it is even more impossible to blame.
Interface culture: If we admit that social networks can’t be an "extra-legal place", how should the platform be responsible? How can the private sector and the state coordinate the relationship between power and responsibility? In view of the transnational nature of the Internet, my feeling is that platform governance also needs international cooperation, and the judgment of "illegal speech" needs to consider the different situations of different countries/societies, and all these will make the platform governance problem very complicated. What are the discussions and consensus in academic circles at present?
Liu Lizhi: This part of the research is still very preliminary. But at present, China, the United States and Europe have different platforms-state relations. As I have mentioned before, the local technology platform in the United States is powerful. In the past, it was mainly self-regulation and industry self-discipline, and even in this "Trump ban" incident, it restricted public power. Although the US government will definitely have more regulations on the technology platform in the future (the anti-monopoly case against Google is currently underway), the very active lobbying mechanism in the United States will also give the platform a lot of room for change. In Europe, because the main platforms are all originated from foreign countries, the regulation of public power is very strict (the European Union has fined Google $9.7 billion for three monopoly cases). The EU also actively empowers EU users on American platforms, allowing EU citizens to restrict the European business of American platforms through personal privacy. In China, the national power is very strong, and there is also a large technology platform made in China. Although there are some conflicts between regulators and platforms, there is a lot of cooperation between the state and platforms in general. According to my own research, there is a phenomenon of "institutional outsourcing", that is, the state outsourced a series of legal, political, social and other institutional functions to the digital platform (Liu, 2021). For example, the Internet platform in China actually exercised many legal functions for the state, such as anti-money laundering, curbing commercial bribery, and collecting money from Lao Lai.
Xu Xibai: The European platform governance model is actually very controversial. Criticism focuses on several aspects: first, there is a lack of data to prove that hate speech legislation can effectively reduce discrimination and harm; Second, the definitions of hate speech vary from country to country. At present, there are at least 19 different definitions, and some definitions even regard speech that "offends" or "may offend" others as hate speech, which may lead to excessive shooting and injure many normal critical speeches, especially religious criticism, thus limiting the legitimate freedom of speech; Third, the banning of "false statements" such as denying the Nazi Holocaust may be counterproductive, making the banned self-proclaimed as being persecuted by the government, but attracting the sympathy and interest of conspiracy theorists.
There are also many problems at the operational level. Germany has always adopted the compromise of "turning a blind eye". Users can’t find the content of denying the Holocaust on the German version of Google, but they can find relevant content by switching to the American version of Google. In recent years, Germany has strengthened its control. In 2017, it promulgated the Law on Network Enforcement (NetzDG), which gave the responsibility of identifying and removing illegal content to the Internet platform. In the face of limited review time and high fines, the platform will generally strengthen self-censorship, preferring to kill one thousand by mistake and not let go of one. Moreover, users whose content has been removed or whose number has been deleted have no judicial channels to appeal, which has caused great harm to freedom of speech and has been widely criticized.
Of course, the biggest problem is how to coordinate the territory of "entity country" and "virtual country". The possible future direction of the EU is to require the platform to implement EU regulations all over the world, but what if other countries also put forward such requirements? Which country’s law shall prevail? Or do they manage their own affairs, strengthen network sovereignty, and finally cut the Internet by country? Timothy Garton Ash compared the Internet to a "cyber metropolis". Although we are in a certain entity country, we can directly communicate with people all over the world, receive news from all over the world and participate in public discussions there by accessing the "metropolis" through the Internet. In this sense, we are all "digital immigrants". For the first time, the Internet has brought us so close to the "freedom to seek, receive and disseminate all kinds of news and ideas regardless of national boundaries" outlined in Article 19 of the Convention. Therefore, the internationality and independence of the Internet should be defended.
To this end, the Internet platform is also making various new attempts. As mentioned earlier, Facebook invited experts to form an independent oversight committee as the "Supreme Court" for content review. The freedom of speech research team of Oxford University, which I participated in before, recently published "Nine Ways for Facebook to Become a Platform for Democracy and Freedom of Speech Better" together with Facebook and Stanford University, including suggestions such as setting up an external content policy advisory group, improving the appeal process, setting up an external appeal agency, and increasing the transparency of decision-making. This collaborative governance model, which involves the participation of platform, government, experts and civil society and is reached on the basis of negotiation, may be a more feasible deliberation.
References:
Garton Ash, Timothy. 2016: Free Speech: Ten Principles for a Connected World. New Haven: Yale University Press.
Garton Ash, Timothy, Gorwa, Robert and Metaxa, Dana?. 2020: GLASNOST! Nine Ways Facebook Can Make Itself A Better Forum for Free Speech and Democracy. Oxford: Reuters Institute for the Study of Journalism.
Jones, C.I. and Tonetti, C. 2020: ‘Nonrivalry and the Economics of Data’. American Economic Review, vol. 110, no. 9, pp.2819-58.
Kitchens, Brent, Johnson, Steven L. and Gray, Peter. 2020: ‘Understanding Echo Chambers and Filter Bubbles: The Impact of Social Media on Diversification and Partisan Shifts in News Consumption’. The MIS Quarterly, vol. 44, no. 4.
Liu, Lizhi. 2020: The Rise of Data Politics: Digital China and the World. Georgetown McDonough School of Business Research Paper No. 3669452.
Liu, Lizhi. 2021: From Click to Boom: The Political Economy of E-Commerce in China. Book Manuscript.
Liu, Lizhi and Weingast, Barry R. 2018: ‘Taobao, Federalism, and the Emergence of Law, Chinese Style’. Minn. L. Rev., vol. 102, p.1563.
Mill, John Stuart. 1991 [1859]: On Liberty and Other Essays. Oxford: Oxford University Press. (For Chinese version, see john mill, translated by Xu Baoxun, On Freedom, Beijing: The Commercial Press, 2005).
Newman, Nic and Fletcher, Richard, etc. eds. 2017: Reuters Institute Digital News Report 2017. Oxford: Reuters Institute for the Study of Journalism.
Pariser, Eli. 2011: The Filter Bubble: What is the Internet Hiding from You. New York: Penguin. (Chinese version: by Eli Pariser, translated by Fang Shifu and Yang Yuan, Filter Bubble: The Secret Manipulation of the Internet on Us, Beijing: Renmin University of China Press, 2020).
Persily, Nathaniel and Tucker, Joshua A. eds. 2020: Social Media and Democracy: The State of the Field and Prospects for Reform. Cambridge: Cambridge University Press.
Sunstein, Cass R. 2006: Infotopia: How Many Minds Produce Knowledge. Oxford: Oxford University Press. (Chinese version: Keith Sunstein, translated by Bi Jingyue, Information Utopia: How People Generate Knowledge, Beijing: Law Press, 2008)