key: cord-0059403-14s7dyc7
authors: Heimo, Olli I.; Kimppa, Kai K.
title: Public Fora and Freedom of Expression
date: 2020-10-09
journal: Human-Centric Computing in a Data-Driven Society
DOI: 10.1007/978-3-030-62803-1_27
sha: 9a1996dbe775dad9784203ee0364a71d0451a699
doc_id: 59403
cord_uid: 14s7dyc7

This paper focuses on freedom of expression in the age of digitalization specifically on social media public fora. As the current political climate worldwide is polarising and both democracies and dictatorships are tightening their grip on ‘hate speech’ and other ‘unwanted’ discourse, we as people should focus on what is being done with our freedom of expression. As it is obvious that not all expression can be free, e.g. child pornography, we should focus on both what expression we should limit, but also how: who should have the power to limit the expression and how should they do it? The argument of the paper is that the current corporation-lead black-box censorship in social media creates a danger for the freedom of expression as the corporations, often foreign for most people, seem to use the platform to affect the functioning of the democracies’ legislation processes. Thus, if we value our democracies, the censorship should be conducted in a justified and as open as possible manner and never without a sound reason.

In this paper we discuss freedom of expression. Freedom of expression is a crucial part of a working democratic society as the citizen's require the ability to share and receive information, express their needs and wants, and discuss on how to develop the society to be a better place for them and everyone else. (C.f. e.g. [1] ) In this paper freedom of expression is examined from the viewpoint of political philosophy with the implications on who is limiting it, how, and why, not from the perspective of the law, as we are proposing basis for new law, not trying to explain how the current legislation functions. It is important to notice that the discussion presented here is not only about freedom of speech, but of expression. An expression of artfor example a modern dance against the totalitarian regime of the People's Republic of Chinacan and must be seen as valuable as verbally pronounced critique against the said 'republic'. When this ability for expression, or receiving the expression is limited, it should be done with utmost care due the consequences. As the society can be seen as a struggle between tyranny and liberty (cf. [1] ), freedom of expression is a tool to keep the tyrannical tendencies in governance at bay.

In the current era of digitalization, expression of opinion has been turning increasingly towards social media and Internet in general. Internet has become a hotbed for political discussion. Whereas the President of the United States used to declare information through press conferences, the trend has been turning more and more towards Twitter. Since 2017 Tweets made by The President of the United States have been considered official statements [2] . And the President of the United States is not alone in this, but government actors and politicians all around the world, as well as corporations, celebrities, and private citizens announce their actions, express their opinions, and communicate with each other through social media. Digitalization has made an era where a child from a developing country can comment US president's post and be heard; possibly. And a child from Sweden is already having an on-going discussion (of sorts) with US president; and is heard.

In social media though, not all the participants are equal, centralization is a clear issue. When checking the statistics in Fig. 1 , it is clear that social media platforms differ in their user-base.

The impact of video platform sizes is somewhat hard due the issue to calculate, but as an estimate Statista, a German online portal for statistics, declares the seven biggest video sharing sites in United States to be YouTube (90% of Internet users reached), Facebook (60%), Instagram (by Facebook) (35%), Twitter (21%), Snapchat (18%), CNN (17%) and Fox News (16%) [4] , the last two being the only sites to have an [3] editor-in-chief. Globally, compared to the second best video-only sharing site Vimeo, YouTube has over 2 billion users log in and watch videos on YouTube every month, whereas Vimeo has just 973,000 registered users in total [5] .

Hence it is clear that in both video sharing and overall social media, centralization of the Internet is true. Moreover, one could clearly argue that with that amount of centralized information sharing, it would make quite an impact socially if someone were to decide on what gets to be shared and watched, and by whom.

Censorship has traditionally been connected to state censorship. This is due to the state having been the only power capable of ordering large scale pre-censorship and punishment for publishing materials seen to be worth stopping from circulation amongst the people. Of course that has led to self-censorship (or publishing under a pen name or entirely anonymously) already in these situations, as the potential authors (or painters or musicians) have typically been quite aware of what might happen were they found creatingor even in possessionof forbidden materials; this ranging from fines through prison sentences or deportation all the way to being burned as a heretic. These days, however, a new form of major censorship has become an even more pressing problem, as the power of the state as a censor has waned, especially in liberal democracies; namely censorship by the corporations.

The Internet opened the possibilities for us to express ourselves enormously; anyone could build a web page, send messages to message boards, talk in chat software and as of late, do all these in one platform or another such as Facebook, Twitter, LinkedIn, Instagram or similar, either internationally or regionally. However, as these platforms have centralized, and in many cases there are no optional platforms, but rather a monopoly on specific areas has become the norm, Twitter for short messages, Facebook for longer discussions, Instagram for pictures (owned by Facebook) or WhatsApp for replacing text messages and small groups (also owned by Facebook), YouTube for video (owned by Google) etc., their community standards and wild interpretations of laws on different countries or regions are threatening freedom of expression. Not always the freedom to say what one thinks, of course you can go on your own web page and say things, but rather the freedom to receive communications, as more often than not, those aforementioned webpages are just not visited by many, if any. The same is true if one tries to migrate to a smaller, less used platform, as most of ones friends or followers are unlikely to do so. Thus, the point of censoring has also shifted from the state (although it is still indirectly there) to the corporations, which are practically fully outside of our democratic control.

Merriam-Webster dictionary defines censoring being "to examine in order to suppress or delete anything considered objectionable". Censorship itself however can be done in several different ways. In this paper we examine three simple ways to limit the freedom of expression: 1. to prevent or suppress the expression to reach the audience, 2. to delete the expression after the expression has reached the audience, and 3. to prevent or suppress the publication of the expression before it is expressed.

Prevention or suppression of the expression to reach the audience requires active censoring and control over the media the expression can be delivered upon. Hence, only big actorse.g. governments and large corporationscan effectively do it. Typically there is also co-operation (voluntary or involuntary) with the public and the private sector.

Deletion of the expression is of course the most inefficient way to censor because the expression, when published, can be copied easily and to fully censor an expression all the copies must be gathered to the deletion (e.g. late music video Knebel by Lindemann which was censored at YouTube, of which the uncensored version is available at e.g. Vimeo). This, however often leads to less access to the content, as it must be presented on alternative media platforms, which are typically not as easily censored, but also less used in general and also less visible in search engines; "if it is not on the first page of a Google search, it does not exist", as the saying goes.

Third and most efficient method of censorship, preventing or suppressing the publication of the expression in the before it is even expressed works with the fear of consequences tied with the act of publishing an expression deemed 'unpublishable'. These consequences include (but are not limited to) legal punishment, social stigmatization, or denying certain services or privileges.

Whereas all these alone can be effective by themselves, the most efficient method seems to be to combine these methods, for example actively censoring certain content, deleting all copies mentioning that content, and retaliating against those who have a copy or spread it. Destroying most copies of a book is not nearly as effective as destroying most copies and making the possession of the book a crime.

To emphasize the unwillingness to publish an expression even more effective is to make it unclear where the limits of acceptable expression are. When the boundaries are clear, it is easy to understand whether the publication of the said expression legal. As an example, it can be made quite explicitly clear where an expression is child pornography, as Interpol report [6] states: "Child pornography is the consequence of the exploitation or sexual abuse perpetrated against a child. It can be defined as any means of depicting or promoting sexual abuse of a child, including print and/or audio, centred on sex acts or the genital organs of children" 1 Moreover, if one is not sure is one's piece of expression child pornography when compared to this statement, one should clearly not publishor make!it in the first place; if not for the sake of legal consequences, then for the sake of the children and their rights. But when the definition lacks clarity and there are no clear precedents available, the act of self-censorship called 'chilling effect' is likely. In these cases the publisher of the expression does not have a clear idea whether or not a certain 1 It is worth to note however that there is discussion on virtually created child pornography and whether it should be banned or not, as there are no directly harmed parties if it is virtually created. Indirect harm may be strong enough for a case, but there is no consensus on this as of yet, and some countries (e.g. Japan) allow creation and possession of such material, where others (e.g. Sweden) do not. This discussion, and especially the ethical implications around it, however is one the authors do not wish to pursue further due the social stigmatisation and the chilling effect created around the subject.

expression is illegal and thereforejust for the safety of itrefrains from publishing it. Or, as Open Net Initiative [7] web pages explain:

Another common and effective strategy to limit exposure to Internet content is by encouraging self-censorship both in browsing habits and in choosing content to post online. This may take place through the threat of legal action, the promotion of social norms, or informal methods of intimidation. Arrest and detention related to Internet offenses, or on unrelated charges, have been used in many instances to induce compliance with Internet content restrictions. In many cases, the content restrictions are neither spoken nor written. The perception that the government is engaged in the surveillance and monitoring of Internet activity, whether accurate or not, provides another strong incentive to avoid posting material or visiting sites that might draw the attention of authorities.

In the current climate of 'hate speech'either legally mandated or limited by the platform ownersmore and more people self-sensor even thoughts and ideas that either are not 'hate speech', but can understood as such by fringe elements, or in fear of being a target of hate speech, such as doxing or threats (see e.g. [8] ). At times these fears are of course also justified. As a banal example either being extremely frustrated or under the influence of mind altering substances it is likely not a good idea to publish ones thoughts, as it can too easily end up being hate speech, such as inciting or even promising to participate in actual violence (see e.g. 2 . This, however is not a justification to censor oneself from opinions which only differ from those of other discussants, yet, the chilling effect causes people to self-censor these comments as well, especially in environments where "wrong think" is punished either directly at the site or even outside the site by attempts to dox the person with "wrong" views, or affect their private or professional life.

Historical expressions quite acceptable at the time are also in the line of attack now that the legislation has changed; especially since the use of language has changed to add previously more-or-less neutral words to the category of ones now unacceptable.

(See e.g. [8] .) Therefore it seems that as the morals shift with the time, we are expected to start censoring our historyor at least warn people about the history which they should be aware. E.g. in Finland, a long-standing member of parliament and exminister from Christian Democrat Party is at the moment being investigated for publishing bible quotations in 2004 [9] according to a law which came into force 2011 [10] .

As the legislation is often after crude language, chilling effect affects mainly people with lower education since they lack the linguistic tools to express themselves in formalized manner. This results in a problem with equality before the court as the critique towards different policies and groups of people often coincide with criminal activities or moral questions [8] . Whereas an educated person can argue the increased immorality, criminality, etc. of a certain group in an eloquent and thus also in a legal manner, the same idea delivered with lesser skills in wording will lead to blatant and illegal conclusion. For example:

1. "The increased amount of immigration will very probably raise the levels of criminal activity in the area due to socio-economical differences in the population groups." 2. "The immigrants will soon be robbing and raping us!"

Mill [1] points out that:

The peculiar evil of silencing the expression of an opinion is, that it is robbing the human race; posterity as well as the existing generation; those who dissent from the opinion, still more than those who hold it. If the opinion is right, they are deprived of the opportunity of exchanging error for truth: if wrong, they lose, what is almost as great a benefit, the clearer perception and livelier impression of truth, produced by its collision with error.

Therefore, according to Mill [1] , freedom of expression is also about being able to receive dissenting views and other expressions deemed uncomfortable by some.

Reference withheld for blind review [8] also point out that:

When people start censoring themselves to not conduct "hate speech", which is defined differently in different societies, anything can be determined to be hate speech. In Turkey critiquing government policies or supporting the Kurds can be considered hate speech; in Finland critiquing current immigration practicesor defending themcan be hate speech; in Russia promoting gay rights is hate speechin the Netherlands critiquing them is hate speech.

Therefore it seems that the legislationnot ethicsdefines hate speech and legislation is observably culturally dependent, whereas ethics at least ought not be.

wrong, and laws ought to reflect this, we know from empirical observation this is not always the case. Many laws in many countries are clearly unethical, examples abound from cruel and unusual punishment to punishing people about choices that affect only themselves; and of course, in the history of the humankind there is plethora of unjust laws. Let us see to it, that hate speech legislation does not become one of those laws.

While the most efficient censorship done so far is by the government, private sector censorship has gained ground within the growth of the Internet due to the centralization of the Internet as discussed in chapter 1 (see also [11] ). With this centralization the 'big players' have taken the key roles of delivering information, some by generating a lot of content, and others by creating a public platforma public forumfor the usergenerated content. Whereas content-creators such as Springer, Netflix or Washington Post thrive from the expressions they produce or publish, the public forum providers rely on the users to express themselves and then benefiting from the advertising revenue, monthly fees, and private information provided by the large user-base to be sold for various purposes, including but not limited to advertisers. This is a key difference. Some of these platforms, e.g. YouTube, reward the users from generating the revenue whereas others, e.g. Twitter or Facebook, do not. This also plays a role in the discussion.

As it is seen as normal in Western society that a traditional mediaradio, newspapers, other print media, television, video renting companies etc. can choose what expressions they wish to publish and promote, the new media and the Internet has been seen as a 'wild west' of expressionanything goes, as long as it is not illegal. With the centralization and perhaps 'stabilization' of the Internet, many of the new media platforms are mainly benefiting from user-generated content to their various fora. The publisher has an editor-in-chief who is responsible for the content published by the publisher, in the public forum the responsibility currently lies with the users doing the publishing. Some limitations do apply however, as the forum upkeep is in some countries required to remove illegal content from the server within a reasonable time.

This difference has been noted by the US Appeals court, which in decision Knight First Amendment Institute vs. Trump 2017 3 noted that "[d]efendants' blocking of the Individual Plaintiffs from the @realDonaldTrump account violates the First Amendment because it imposes a viewpoint-based restriction on the Individual Plaintiffs' participation in a public forum." In this case the president (a public sector actor) used Twitter's (a private sector actor) platform (a public forum) function to limit the freedom of expression.

Various issues arise from the concentration of social media in very few hands and the polarisation of the users within these fora. These include (but are likely not limited in) demonetization, ending in bubbles which strengthen one's world view without ever challenging it, hiding content from searches and not promoting it even though the user has expressed their wish to see all the content by the provider.

Demonetization is a complex phenomenon. First, it cuts the publishers' income, when they are acquiring income via new media content monetization. Most of the money one receives from the content comes within the first few days. Thus, if one is demonetized, one practically always loses the main part of the potential income with the content even if it is remonetized later [12] . Secondly, in many cases (e.g. You-Tube), one's content is also 'unlisted', when demonetized so that the user must know where to find the content, and the content is not promoted [13] . From the perspective of censorship, this is a form of light-censorship, as unlisted media does not produce information about new content from the producer to the audience, neither to those subscribing to the videos or those hoping to get them as suggestions, and especially not to those who want to value their privacy by not logging into the media platform, for example, those using browser privacy-modes or deleting their browser history and cookies, or those using publicly accessible computers, such as library computers or computers in Internet cafes. Therefore, only those seeking new content from the producers by clicking their content directly or searching with exact keywords will find the unlisted content, while other, competing content is made easily visible before, during, and after viewing videos.

If it is altogether too easy to choose one content over another as it is clearly the case here, it will bias the information received and thus might bias the world-view of at least some portion of people and therefore produces not only Internet bubbles where people do not see certain views, or only those accepted by the platform provider, but also previously mentioned 'light censorship' where the competing opinions cannot compete within "the marketplace of ideas", as it was called by Justice Wendell already in 1919. Moreover, when those committed or acquainted with certain perspectives notice that information supporting their world view is being even slightly censored, it creates frustration, anger, and in the end, polarization, pushing them more deeply in their bubble and others against them. If the idea is to not polarize people but to get people to work together and find compromises, this kind of behaviour is contradictory to the intent. In the end it only strengthens the defensive lines and suppresses the will of the groups to understand each other.

In a public forum one should not use (at least closed) AI, or a black box version of an algorithm. We need to have public officials who verify which algorithms are used and how they actually workthey need to produce the same result with the same input; and it needs to be understood how. This is relevant to the democratic development of society as a whole, and the freedom of different ideas to compete in society, so that the best ideas can actually be found. If the algorithm is indeed a 'black box' or the rules apply differently to different people, there is a concern of treating people differently and thus promoting racism, sexism, or other forms of oppression. If the results change over time or due to the algorithm 'learning' what the user searches, the results will indeed bias the view of the viewer. Manythe authors includedtry to avoid this in various ways, for example pressing "like" on interesting, rather than agreed with content, using aforementioned privacy modes and even other computers and computers at different places to see content that could raise new ideas; but unfortunately most people are unlikely to follow such practices to get a wider understanding of the topics, and rather end up in the bubbles the algorithms push them into. It is also questionable whether this even works in the long run, especially if the algorithms use IP address as an identifier; in this case even privacy mode might not helpor the content offered could be related to whatever those who use the same IP are interested in, not what the person searching for information is interested in.

One could argue that these are private entrepreneurs not the public sector. So why cannot a private actor choose what is published in their fora, with their money, with their name? The answer is twofold: first, it is about the authorship and its responsibility. Whereas a publisher of a magazine, television broadcast, video rental, or a WordPress web page is responsible on what is published, in a public forum the responsibility and accountability lies with those creating the content and not those providing the forum for publishing it. Hence when on a discussion board, whether it be an Internet forum or a public space in the physical world, the responsibility of what people publish is by those who upload the content, not those who supply the space for discussion. We should treat an owner of a YouTube account, Facebook account, Twitter account or any other new media account as the editor in chief. Interestingly, the US legislation might just support this (see e.g. [13] ), but this is yet to be seen. Secondly, as the centralization of the Internet becomes more prevalent [14] , the control over discussion will be moving more and more to the hands of the few when looking from the perspective of content producers. If only those few get a say on what is discussed, the consequences may be quite dystopian. As the control of the expression falls in the hands of few and if there are little rules dictating on how that control over others can be used, the one in control has quite free hands on what ideas and expressions are being discussed and promoted in the society as a whole. The idea of freedom of expression does not specify the participant ruling the consequences or applying the censorship, quite the contrary.

As the world is clearly moving to a state where wealth is more and more in the hands of the few, those few, even if they cannot now, will soon be able to acquire all the media platforms and thus dominate what we are discussing. And it can be argued they already can, as especially Google and Facebook (or their respective parent organizations) are concentrating this power in their hands. As Mill [1] already noted, to protect our freedom of expression now and in the future we must allow the freedom of expression forms we are not comfortable with and not allow anyone, neither the governments nor the corporations to use our differences of opinion in matters of immigration, religion, political views, economics nor anything else similar to justify their control over the public opinion. The stakes are just too high. And to protect us from tyranny of the few, we need freedom of expression.

If one has received substandard service in a government office, there are typically ways to find out why. In the Internet platforms, the censored typically get either no explanation, or such a vague explanation that it means nothing in relation to the perceived offence (see e.g. [15] ). There are numerous examples on how public forum providers have censored political discussion. For a current example a Finnish Facebook group mensroom (miestenhuone) with over 50 000 members (*1% of all Finns) was censored probably due the promotion of a legal initiative to legalise cannabis [16] without any warning nor given information to the administrators of said group 4 . (see also [17] ) Neither the users nor the moderators of the group got any information that they have been removed from the group, that the group had been dissolved, or any explanation why. The legislation initiative (also requiring over 50 000 supporters, just by coincidence) was accepted for government processing [17] and will be delivered for discussion in the Parliament of Finland. The initiative did not support the use of said drug, but to remove punishment of using it due to increased harm to the users caused by the current legislation. Facebook authorities seem to have interpreted this however as a pro-drug campaign even though it clearly was a discussion about the social consequences and the functioning of a justice system in a modern society. Hence it is clear that the policies of Facebook do not follow the legal system and juridis-ethical discourse in Finland, but instead try to influence that discourse with a moralistic mind-set. The censorship of course did not concern only the discourse on cannabis and its state of legality, but of all the discussions in the said group, which all were permanently deleted from the public due to this one legally sound 'mistake [16] . It should be clear that legislation initiatives should be open for discussion and if they are denied by the reason that what is promoted to be legalised is illegal and hence should be not discussed it is hard point on the legal discussion. Moreover if the discussion is about the ethics it is clear where the platform controllers -in this case facebook -has a moral stand. As the discussion in this case -the case of legalising recreational use of marijuana -is indeed legal in many countries and areas, should this be legal discussion is clearly denying should this be a legal discussion is clearly a moral standpoint.

It seems, then, that Facebook acts in the legislative process by hindering (and possibly in other cases promoting?) the legislations it chooses to be morally sound. The question is why Finns or any other group of people accepts this? Moreover, why is this not an issue in more general sense on who gets to choose what is discussed, promoted, and censored?

As stated above, the Facebook group was possibly completely censored due to a single link to a government web-page where citizens were able to influence the legislation of their country. However, because Facebook, YouTube, Twitter, or any other public forum providers are responsible in answering the accusations or questions arising from their censorship, this is not certain. The first step towards more open society would be to make them accountable on the methods and decisions behind their censorship.

Whereas clear rules, open decision making, and public discussion are methods on developing an open society where people can without fear publish their opinion, learn from each other, and participate in the decision-making process concerning them and people around them, obfuscated rules, black-box (perhaps AI-based?) decision making processes, and fear of arbitrary punishment by powerful corporations, government actors, and other members of society alike, are all likely to lead to a society where only some opinions are accepted.

Freedom of expression is a tool that upholds democracy and civil liberties. Without freedom of expression we cannot function as a democracy but are limited by those limiting our expression. We should be careful about the government limiting our expression, and we should not give that power to private corporations either. These are powerful tools in the wrong hands. The public fora should remain public spaces for discussion and development of the society, not private spaces where someone with capital, political power, or loud enough supporters can choose what is discussed and how.

On liberty

Trump's tweets are 'official statements

Global Social Networks Ranked by Number of Users 2020

Statista: Leading online video platforms in the United States in

UNESCO: Sexual abuse of children on the Internet: a new challenge for Interpol

OpenNet Initiative, About Filtering

Hate speech recognition AI -a new method for censorship

Hallituksen esitys eduskunnalle Euroopan neuvoston tietoverkkorikollisuutta koskevan yleissopimuksen lisäpöytäkirjan, joka koskee tietojärjestelmien välityksellä tehtyjen luonteeltaan rasististen ja muukalaisvihamielisten tekojen kriminalisointia, hyväksymisestä ja laiksi sen lainsäädännön alaan kuuluvien määräysten voimaansaattamisesta sekä laeiksi rikoslain ja tietoyhteiskunnan palvelujen tarjoamisesta annetun lain 15 §:n muuttamisesta

https:// hackernoon.com/the-evolution-of-the-Internet-from-decentralized-to-centralized-3e2fa65898f5

Business Insider 'Like you've been fired from your job': YouTubers have lost thousands of dollars after their channels were mistakenly demonetized for months

Case 5:17-cv-06064-LHK Document 1, United States District Court

P2P and the promise of internet equality

YouTube vs Grey: a ballad of accidental suspension

Iltalehti: Suosittu Facebook-ryhmä Miestenhuone poistui yllättäen -yleisö ymmällään

Kansalaisaloite: Kansalaisaloite Kannabiksen käytön rangaistavuuden poistamiseksi