FREEDOM OF SPEECH AND FIRESTORMS:
The marginalisation of women on social media
Hannah Storey
31
stAugust 2018
Contents
Abstract ... 2
Introduction ... 3
1. Firestorms and online norm-enforcement ... 7
1.1 What are firestorms? ... 7
1.2 The question of anonymity, social identity and norm enforcement ... 10
1.3 Online norm-enforcement ... 16
2. Frustrating truth seeking ... 19
2.1 The balancing of harms ... 19
2.2 The Tyranny of the Minority ... 23
3. Misogyny and the silencing of women ... 26
3.1 Silencing women ... 26
3.2 Misogyny ... 31
3.3 Categories of misogynistic abuse ... 37
4. Combatting silencing ... 39
Conclusion ... 44
Bibliography ... 46
Abstract
Online firestorms, massive outbursts of negative and abusive language directed at an individual, have become an increasingly frequent social media phenomenon. In this thesis, we shall see that firestorms targeted at women are an expression of the sexist attitudes of some online communities and that they are therefore misogynistic actions which enforce these sexist norms online. By silencing and marginalising the views of women on social media these firestorms disrupt truth seeking and thus undermine the purpose of free speech according to Mill, although, because of this, firestorms warrant regulation, in practice regulating firestorms is impractical and would fail to undermine their detrimental effects. Therefore, instead of championing regulation, this thesis will dissect the causes of firestorms and the circumstances in which they are facilitated so that they can be disrupted. Social media should actively break up clustered networks and encourage healthy debate across different groups in order to disempower firestorms.
Introduction
Social media has allowed millions of users to connect across the world. People who may have once struggled to find like-minded people in their local community are able to reach into cyberspace and find kindred spirits with whom they can share their interests. Even more amazingly, those living in oppressive societies can connect with people to explore their mutual dissent in relative safety. During the Arab Spring, for example, social media allowed citizens to exercise their freedom of speech1 as well as to organize and coordinate protests. Even more than this, participants in the
protests were able to communicate their perspective of events to the outside world without media or state influence.
The connectivity of social media means that it has been heralded as an enhancer of freedom of speech. Anyone, as long as they have access to the technology – which of course not all people do but that is for a separate discussion – can now join Twitter, Facebook, Instagram etc. and share their thoughts, feelings and opinions. People whose views have previously been sidelined because they were from marginalized communities or did not have the opportunities to publish their ideas in the traditional media, now have the opportunity to speak to a global audience. Normal people can comment alongside those with power, as well as communicate with those in power, with relative ease and in unprecedented numbers.
Although there is no doubt that social media has facilitated conversations which may never have occurred or reached such a wide audience if it wasn’t for these platforms, social media is
unfortunately not the egalitarian, utopic vision many expected or hoped it would be. Although social media appears democratic in structure, powerful voices still dominate and the way in which we interact online can often lead to the sidelining or silencing of views.
The clustering of online networks and the creation of cyber echo chambers on social media has been the subject of much recent debate. Echo chambers have been blamed as the cause of Brexit,2 the
1 Ghannam, Jeffrey, Digital Media in the Arab World One Year After the Revolutions, Center for International Media Assistance, the
National Endowment for Democracy, 2012
2 Chater, James, What the EU referendum result teaches us about the dangers of the echo chamber (2016),
https://www.newstatesman.com/2016/07/what-eu-referendum-result-teaches-us-about-dangers-echo-chamber accessed on: 10/08/2018
reason for Trump’s presidency,3 and even the destroyer of democracy itself4 and it is certainly true
that the EU referendum result and Trump’s victory came as great surprises to the liberal media, and that echo chambers could well be the explanation for this.
Another social media phenomenon which has piqued the interest of society is the deluge of abuse some unfortunate social media users receive. Shocking examples of horrific abuse in enormous volumes have become increasingly common.5 In 2014 for example #gamergate saw several women
targeted with harassment campaigns. Feminist and media critic Anita Sarkeesian sustained some of the nastiest abuse. At its worst, Sarkeesian received death and rape threats and she had her personal information shared online (doxing). She was forced to flee her home.
It struck me that these online phenomena may not be separate entities; that the prevalence of online abuse might make social media users wary to interact outside of their echo chambers for fear of retaliation, thus, limiting interactions between those with differing views online. Although there has been significant research into the impact of hate speech both online and offline, as well as research into the legitimacy of regulating hate speech, there has been less focus on these online mobs – also called firestorms. Although they might be considered a subset of hate speech due to the horrific language used by some, these firestorms are distinctive. The sheer number of participants in a firestorm, as well as their speed, can mean that the impact of a firestorm on the target is
significant. Throughout the research for this thesis it became clear that not only do these firestorms cause people to retreat further into echo chambers, they can also silence targets completely by causing them to leave social media in order to stem the abuse. These firestorms are therefore not only harmful to the individual, they are also detrimental to their right to free speech.
Mill argues clearly in ‘On Liberty’ that the value of freedom of speech is as a mechanism for seeking the truth; that truth was best sought by constantly questioning perceived wisdom, so that it may be replaced if a better truth is identified or it may be strengthened by its ability to ward off other ideas attempting to replace it. In Mill’s account, a society which ideally facilitates freedom of speech must support the speech of a variety of actors. Supporting speech from all types of people with different experiences would allow the greatest range of ideas to be discussed. This is similarly true when
3 Hooton, Christopher, Social media echo chambers gifted Donald Trump the Presidency (2016),
https://www.independent.co.uk/voices/donald-trump-president-social-media-echo-chamber-hypernormalisation-adam-curtis-protests-blame-a7409481.html accessed on: 10/08/2018
4 El-Bermawy, Mostafa, Your Filter Bubble is destroying democracy (2016),
https://www.wired.com/2016/11/filter-bubble-destroying-democracy/ accessed on: 10/08/2018
5 Bosker, Bianca, Why we should all fear the righteous online mob (2013), https://www.huffingtonpost.co.uk/entry/justin-sacco-online-vigilantism_n_4505452 accessed on: 20/08/2018
those of differing opinions retreat into groups to discuss their ideas. It is not enough that a range of views are being expressed, if those views are only being expressed to those who already hold the same (or similar) views and they are not shared between groups who hold different views. Sharing between groups allows debate, and debate is at the core of Mill’s theory on truth seeking and freedom of speech. Firestorms, which can silence individuals as well as encourage individuals to retreat into echo chambers, are therefore detrimental to truth seeking itself.
Contrary to popular belief this thesis will show that firestorms are generally not the work of anonymous online users. Instead this thesis will argue that firestorms function as a norm-enforcing action similar to Mill’s tyranny of the majority. However, given that firestorms often do not enforce society-wide norms they cannot be considered the tyranny of the majority. Instead firestorms enforce the norms of different online communities, therefore, I propose that firestorms are termed the tyranny of the minority. As norm-enforcing sanctions, firestorms punish individuals and thus discourage them and others to continue to speak out on certain topics. Just as the tyranny of the majority discourages individuality, the tyranny of the minority discourages the expression of certain views on social media. Firestorms therefore have the ability to cause the homogeneity of views expressed online. Some online communities are so successful, and their firestorms so ubiquitous, that the views which they target have already become marginalized.
One type of firestorm that is particularly common is that which is targeted at women, particularly women who attempt to occupy a position of power, criticize a male dominated space or speak out on the topic of feminism. The firestorms targeted at women also differ from other types of
firestorms. Research has shown that when women are targeted by firestorms the abuse they receive is uniquely gendered. This means that the firestorm is more likely to contain elements such as sexist language, threats of sexual violence and gendered stereotypes. 6 I argue in this thesis that firestorms
targeted at these women are a particularly vicious example of the tyranny of the minority. I argue that these firestorms are in fact a form of misogyny, meaning that they aim to punish women for stepping outside of gendered social norms. These firestorms enforce sexist views online and, although the sexist views which inspire these firestorms are not mainstream views, the power of these firestorms is causing the marginalization of women’s views online. Given that social media is now such an important aspect of engaging with debate, side lining the views of women online is hugely detrimental to free speech.
Many groups are disproportionately targeted by firestorms including people of colour, people with disabilities and minorities, but for the purposes of this essay I will be focusing on women. This is for many reasons, including brevity, but also experience. It is not for me to speak for others when I do not fully understand their experience and may misrepresent it. For this reason, I acknowledge that there are aspects of prejudice which I will fail to include here. I do not intend to state that women are unique in facing a hostile environment on social media, or even that they have it the worst, but they certainly do face heightened levels of hostility and this is worth investigating. It must also be noted that identities are intersectional, meaning that not all women will experience hate speech on social media in the same way. Women of colour for example are doubly targeted due to their gender and race, and women with disabilities will similarly experience multiple levels of discrimination. This is based on different deep-set prejudices which I will not have the space to fully dissect here but might change the way in which they experience or react to firestorms.
Given their negative impact on truth-seeking, we will see that there is a clear free speech argument for the limitation of firestorms. However, pragmatically speaking the regulation of firestorms is very challenging. Governments could place restrictions on instigating firestorms but determining
culpability is almost impossible. Alternatively, they could regulate the worst of the speech in a firestorm e.g. hate speech, but it remains unclear whether or not removing this speech from a firestorm would undermine its impact on the targets and therefore remove the silencing effect. For this reason, I will not be championing the regulation of firestorms in this thesis.
In order to discover how to disempower firestorms throughout this thesis we will dissect how social media facilitates firestorms and therefore how we might address them. In the real world, the most closely aligned action to a firestorm is the angry mob and yet the angry mob is uncommon and the firestorm prevalent. The reasons for this are the key to discouraging firestorms. We shall see that firestorms occur more frequently because participating in firestorms is low-cost and high reward, there is a lack of non-verbal cues, very little chance of negative retaliation and participants are encouraged to join firestorms due to echo chambers. I suggest that the most effective way to discourage firestorms would be for social media networks to undermine echo chambers. Given that echo chambers are caused by the clustering of online networks, social media sites should make a concerted effort to break up this clustering. Although it occurs naturally online, as it does offline, social media sites also exacerbate clustering through their algorithms. These algorithms mean that social media users are shown more of the same content thereby intensifying the clustering of networks. In order for social media to truly be an egalitarian space for free speech, social media
needs to rewrite their algorithms as well as taking steps to actively break up these clustered networks. This is the only way to halt and hopefully reverse the increase of firestorms on social media.
1. Firestorms and online norm-enforcement
In this chapter I endeavour to explain in detail what a social media firestorm is and its identifying features. Namely the rapid dissemination of vitriolic speech in large volumes on social media. By examining social identities and how they are constructed we shall then see how firestorms function as enforcement mechanisms of the norms of online communities. This is contrary to popular opinion which views firestorms as anti-normative behaviour. Finally, we shall see why norm enforcing actions are more frequent online than offline.
1.1 What are firestorms?
In August 2014 Eron Gjoni published the “Zoe post,” a 10,000-word manifesto about his ex-girlfriend Zoe Quinn. In the post, he attempted to damage Quinn’s reputation by accusing her, an independent games developer, of trading sex for positive reviews of her game.7 This was an unfounded
accusation, but it didn’t stop the internet exploding with criticism of Zoe and the “ethics of games journalism.” The firestorm of abuse Quinn and many other women received in her wake has come to be known as Gamergate after the hashtag #GamerGate was used on Twitter. Quinn herself faced rape and death threats, her personal details, including her home address and phone number, were released (doxing) and someone even changed the information on her Wikipedia page to say she died on 13th October 2014, the date of her next public appearance.
If I ever see you are doing a pannel [sic] at an event I am going to, I will literally kill you. You are lower than shit and deserve to be hurt, maimed, killed, and finally, graced with my piss on your rotting corpse a thousand times over.8
In response, fearing for her physical safety, Quinn fled her home.9
7 Lewis, Helen, Gamergate: a brief history of a computer-age war,
https://www.theguardian.com/technology/2015/jan/11/gamergate-a-brief-history-of-a-computer-age-war accessed on: 1/07/2018
8 Malone, Noreen, Zoe and the Trolls (2017), http://nymag.com/selectall/2017/07/zoe-quinn-surviving-gamergate.html accessed on:
11/08/2018
9 Stuart, Keith, Interview: Zoe Quinn: ‘All Gamergate has done is ruin people’s lives’ (2014),
The sheer number of comments directed at Quinn was astounding. The hashtag #GamerGate was used 244,000 times on Twitter in the first week alone10 and although Quinn’s experience was
extreme, she is not alone. The proliferation of hate speech towards social media users at great volume is a social media phenomenon that has become increasingly common. The phenomenon is called a firestorm:
In online firestorms, large amounts of critique, insulting comments, and swearwords against a person, organization, or group may be formed by, and propagated via, thousands or millions of people within hours11
The high speed and extreme volume of interactions are key aspects of firestorms. This is what differentiates firestorms from hate speech. Although firestorms do include hateful comments, hate speech can come in many forms including as a standalone comment, firestorms on the other hand are formed when a very large number of people interact with the same topic all at once. This is why the term firestorm best conveys the intensity of these types of attacks on social media, as well as their ability to sustain themselves. Firestorms start with a few comments from influencers (or sparks) and then expand exponentially as more people join the frenzy (building up like a fire) until eventually the peak is reached, and the fire burns itself out. Firestorms at their peak can dominate social media channels, such as Twitter, with hundreds or thousands of unique users commenting or engaging.
Chiefly, firestorms occur when large numbers of people express anger towards a company, organization or individual. The more detailed reasons firestorms occur will be covered later in this chapter however, it is important to note here that anger is a key component; firestormers, at least the initial commenters, usually feel wronged by the target. This can be for any number of reasons. In 2012 McDonald’s asked users to share their experiences of visiting McDonalds on Twitter with the hashtag #McDStories. The hashtag exploded but instead of sharing positive experiences, Twitter users took the opportunity to share their horror stories of visiting the fast food chain instead. Although McDonald’s pulled the promotion in under 2 hours, at its peak there were over 1600 conversations on the topic.
10 The star, After the storm of Gamergate (2014), https://knowyourmeme.com/memes/events/gamergate#fn13 accessed on: 20/08/2018 11 Rost, Stahel and Frey, ‘Digital Social Norm Enforcement: Online Firestorms in Social Media’, Social Media. PLoS ONE, 2016, 11:6
Often firestorms occur in response to an action or comment made by a person, company or organization which social media users deem unacceptable. Instigators of the firestorm, or ‘sparks’, might start with legitimate criticisms of the initial action or comment, but commenters will quickly descend into solely offensive comments void of content or arguments;insulting, rude or outright hateful speech for example.
Vitriol is another key feature of firestorms. The definition of firestorms given here doesn’t quite capture the malice of a firestorm but a key component of a firestorm is the sheer volume of hatred: threats, abusive language and hate speech, which are directed at the target. This hatred often appears vastly disproportionate to the action of the target to which the firestorm is responding. For instance, Mary Beard, a Cambridge University don, scholar and classicist, was the target of a
firestorm after she appeared on British weekly political debate show, Question Time in 2013. After casting doubt that public services were under strain due to immigration, Beard received instant backlash, which quickly turned in her words, “truly vile.”12 Soon Beard was the subject of aggressive
and sexualized comments. She was called “a vile, spiteful excuse for a woman, who eats too much cabbage and has cheese straws for teeth” and “an ignorant c**t,” she was subject to questioning about the size of her vagina and humiliated by a discussion about her pubic hair.13 She also faced
threats of sexual violence. It is clear that, no matter the validity or invalidity of her argument, the harassment she received was unwarranted, and some of it surely illegal.
Firestorms usually occur out of the blue and are extremely difficult to predict. Although celebrities might reasonably expect their chances of being targeted by firestorms to be higher, due to their visibility, anyone can be targeted. Catalysts are varied. An ill-thought out comment, or a
controversial statement might ignite a firestorm, but so might a fallacious rumour. On top of this, attempts to stem the flow of vitriol or turn the tide, particularly by the target, are mostly futile,14
meaning the victim has very little influence on the outcome, even if the firestorm started over a fabrication.
Thus far we have identified the three key features of firestorms: speed, volume and vitriol. This is not to say that explosions of positive comments cannot occur and self-perpetuate in the way that
12 Dowell, Ben, Mary Beard suffers ‘truly vile’ online abuse after Question Time (2013),
https://www.theguardian.com/media/2013/jan/21/mary-beard-suffers-Twitter-abuse accessed on: 1/07/2018
13Beard, Mary, A Don’s Life, https://genius.com/Professor-mary-beard-a-dons-life-blog-annotated accessed on: 1/07/2018
14 Although marketers are attempting this see Conick, Hal, How to Leverage a Social Media Firestorm to Increase Brand Value (2017),
https://www.ama.org/publications/eNewsletters/Marketing-News-Weekly/Pages/social-firestorm-increase-brand-value.aspx accessed on: 1/07/2018
negative comments do in firestorms, but these positive versions of firestorms do not detrimentally impact social media’s ability to offer an egalitarian space for free speech, and therefore I will not consider them here.
1.2 The question of anonymity, social identity and norm enforcement
The cause and purpose of firestorms has caused much debate and confusion, particularly the rapid descent into vitriolic and hateful speech. Many theorists have assumed that anonymity is a key factor in online aggression, and therefore in firestorms, in line with Le Bon’s classic deindividuation theory.15 Le Bon theorized that people in crowds lost their sense of personal identity, therefore
losing their societal constraints and becoming more likely to act hedonistically or primitively.16 Le
Bon suggested that the state of deindividuation (when personal identity is overtaken by social identity) was brought on by the lack of accountability in crowd situations. In other words, classical deindividuation theory suggests that when people feel unidentifiable in a crowd (or anonymous) they are more likely to act antisocially due to the lack of consequences.
Many theorists have linked classic deindividuation theory to anonymity online, suggesting that people who are anonymous, or feel anonymous as part of a large online community, are unlikely to suffer consequences for their actions. They are therefore uninhibited in participating in anti-social behavior such as firestorms. Surprisingly however, research contradicts this and has shown that “online anonymity does not promote online aggression in the context of online firestorms.”17 If
decreased inhibitions due to online anonymity were crucial to social media aggression then introducing policies which undermine anonymity online would surely weaken this aggression, yet Rost, Stahel and Frey have shown that the reverse can be true.18 Firestormers are not likely to be
anonymous.
Rost et al suggest that social norm theory provides a more satisfying explanation for aggression in firestorms than classic deindividuation theory. Although firestorms are an example of antisocial behavior we should not consider them to be anti-normative. The vitriol of the speech in firestorms
15 For example, see Suler, J.,’ The online disinhibition effect’, Cyberpsychol Behav., 2004, 7(3):321–6, Hollenbaugh EE, Everett MK, ‘The
effects of anonymity on self-disclosure in blogs: An application of the online disinhibition effect’, Journal of Computer-Mediated Communication, 2013, 18(3):283–302 and Moore MJ, Nakano T, Enomoto A, Suda T., ‘Anonymity and roles associated with aggressive
posts in an online forum’, Computers in Human Behavior, 2012, 28(3):861–7
16 Festinger, L et al., ‘Some Consequences of Deindividuation in a Group’, Journal of Abnormal and Social Psychology, 1952, 47:382-390 17 Rost et al, p17
has made observers assume that firestorms are anti-normative but in fact firestorms are
norm-enforcing. Norms guide the behaviour of individuals as they internalise societal expectations or the
expectations of the group. Firestorms arise when someone (or a company) has acted contrary to those norms. The firestorm is therefore a type of group sanction targeted at the norm-breaker aiming to punish them for their transgression. The norms being sanctioned may be contrary to general social norms and this is likely why outsider observers have struggled to see firestorms as norm-enforcing actions.
It should be noted here that firestorms are distinct from trolling. A troll is someone who purposefully starts arguments and spreads discord on the internet.19 They post inflammatory
comments with the express purpose of being controversial and eliciting an emotional response. Although trolling may occur during a firestorm, or indeed spark one, participants in firestorms are generally not trolls. The actions of trolls are planned to create a stir and as such they may well act anonymously to prevent punishment. Firestorms on the other hand are the result of anger and are predicated on firm beliefs. Unlike trolls, firestormers are less likely to act anonymously.
Social norms are the unwritten societal rules which dictate how we behave and general adherence to social norms is, I suggest, uncontroversial. Social norms prescribe how individuals should act in different circumstances, but they differ from person to person, group to group. For example, there are norms which dictate how one should act as a student, as a daughter or son or as an employee. There might be even more specific norms about how one should act at a particular university, or in a particular workplace. There are norms dictated by religion, or the lack of religion and there are cultural norms, dictated by where the individual lives. People act differently at work than they do at home in recognition of these varying norms. Understanding of and conformity to norms can even predict how successful people are in different environments.
Deindividuation occurs with the immersion of the self into the group; when personal identity is superseded by social identity. Actions therefore become regulated by group norms, rather than personal norms. The lack of self-regulation may lead to anti-social behaviours, but only – and this is crucial – if the group encourages it. Deindividuation can also lead to positive outcomes. When a group tends towards positive behaviours, these are imitated by others in the group. Alternatively,
19 Oxford Dictionary defines a troll as “a person who makes a deliberately offensive or provocative online post” at
when a group tends towards anti-social behaviours, these anti-social behaviours are imitated instead.
The so-called “antinormative” behavior found in the crowd, according to deindividuation theory, is actually a display of what is normative within the crowd20
Deindividuation causes what Turner called the “cogitative redefinition of the self”21 from the
personal to the collective, thus turning individual behavior into collective behavior.22 In essence,
deindividuation is a reduction of self-awareness which leads to a reduction of self-regulation; group members are instead regulated by the group norms. Deindividuation can therefore only cause antisocial behavior to occur if that antisocial behavior is in line with the needs and norms of the group.
Social norms are regulated and proliferated through the use of sanctions. Some norms are enforced by the law and are therefore subject to legal sanctions, such as prison sentences or fines, but sanctions do not have to be official; communities also self-regulate through social sanctions. Elster argues that external sanctions are not always necessary, he argues that internalized norms can trigger feelings of guilt or shame, or at least the anticipation of guilt or shame, which can be enough to enforce the norm.23 He convincingly demonstrates this by musing that he does not throw litter,
even when there is nobody to observe him.
Turner suggests that self-categorization with a group or community is the root of their social influence over their members. Therefore, the groups which people choose to identify with are likely to hold more influence over their actions.24 Self-categorization theory proposes that each person has
many identities: a personal identity e.g. ‘I Hannah Storey’ and various social identities e.g. “we Europeans”, “we British”, “we feminists”. People define themselves with different identities depending on the situation and which identity becomes salient.25 In those situations where people
define themselves through social identities, personal identity becomes less relevant, and the norms
20 Reicher, S.D., ‘Social influence in the crowd: Attitudinal and behaviour effects of de-individuation in conditions of high and low group
salience’, British Journal of Social Psychology, 1984, 23: 341-350
21 Turner, J. C., ‘Social identification and psychological group formation’, in The social dimension: European developments in social
psychology, 2nd ed., by Taifel, H. (Cambridge: Cambridge University Press, 1984), pp. 518-538.
22 Trepte, Sabine, ‘Social Identity Theory and Self-Categorization Theory’, The International Encyclopedia of Media Effects, 2017 23 Elster, J., Social Norms and economic theory, Jounral of Economic Perspectives, 1989, 3(4): 99-117, p104
24 See Turner, J. C., ‘The analysis of social influence’, in Social identity and intergroup relations, eds., by Turner, J. C. Hogg, M. A., Oakes, P.
J., Reicher, S. D., and Wetherell, M. S., (Cambridge: Cambridge University Press, 1987) and Turner, J. C., Social Influence, (Milton Keynes: Open University Press, 1991)
25 Turner, J.C., ‘Some current issues in research on social identity and self-categorization theories’, in Social identity: Context, commitment,
of the social group identified with become increasingly important. In other words when an individual identifies as a feminist rather than as an individual, they are more likely to act according to feminist norms. Alternatively, when they define themselves as British, they are more likely to act according to British norms. The fluidity of social identities means that group members are not in a constant state of deindividuation. The influence of group norms on individuals clearly waxes and wanes as social identities become more or less salient to the individual.
Turner’s theory also suggests that group influence is minimized if the individual does not categorize themselves as a group member. For example, if an individual joins a local political party just to see what’s going on in the meetings, but distinctly does not identify with the party, they would be unlikely to be influenced by their conversations or join in their actions. In other words, by
self-differentiating themselves, they would be unaffected by the norms of the group.26
The influence of social identities offline is uncontroversial. For example, in the 1980s British mining towns had a strong sense of community in response to their livelihoods being under threat (as the mines were being closed down). All members of the communities would have voted for the British Labour party unquestioningly, in line with the norms of the community. To do otherwise would have been seen as an affront to the group. A community member who dared to vote Conservative would have likely been shunned or even attacked. Clearly the influence of group identity can be very strong. However, some view communities on social media as less valid, and therefore less influential, than communities in the real world. Someone who holds this view might assume then that social media communities are less likely than real world communities to regulate the behaviour of their members.
However, the influence of online communities on their members can in fact be very strong. There are several reasons for this. Firstly, participants have actively chosen to join that online community and have therefore chosen to strongly align themselves with the group (self-categorization). This is not to say that there aren’t outliers in online communities but given that outliers are unlikely to participate in firestorms we will not concern ourselves with them here. Secondly, the vastness of cyberspace means that the choice of communities to join is extensive, and users can consequently find communities whose views strongly align with their own more easily than they would in the real
26 Nadler, Arie and Goldberg, Marta, ‘Effect of Self-Differentiation and Anonymity in Group on Deindividuation’, Journal of Personality and
world, where communities to join are limited geographically. Finally, social identity on social media is in fact strengthened by our ability to construct our identity online, what I call relative obscurity.27
Relative obscurity is achievable on social media because we have a high level of control over how we display ourselves online. Users can curate their online identity. Trolls clearly use this tactic to achieve anonymity online by creating a completely fake persona, but many users simply create a profile which highlights their preferred qualities, meaning their online identity is not completely fake, but it is not completely transparent either. Instead of relative obscurity reducing our ability to have meaningful connections online, the social identity model of deindividuation effects (SIDE model)28
suggests that relative obscurity actually reinforces our group bonds. This is because relative obscurity highlights our similarities whilst diminishing any differences which might weaken our feelings of unity.29 Social media users might therefore – rightly or wrongly - feel that their online
friends are more likeminded than their real-world counterparts. This is exemplified by online dating. When people meet online it is easy to think they have everything in common. Whereas it is much more difficult when they meet in person. In other words, the SIDE model suggests that after people self-categorize as a member of a group, relative obscurity of online communities can enhance group identification which heightens the influence of the group.30
I should clarify here that by relative obscurity I do not mean anonymity where no information about the user is conveyed at all, instead I mean obscurity in the sense that the user only displays the information which they wish to convey. Online anonymity does not facilitate deindividuation effectively because deindividuation relies on genuine connections between group members. It is difficult to trust or identify with people who are anonymous and as such, groups of anonymous users are less likely to influence behavior.
As with all communities, online communities have and enforce their own sets of social norms or ‘rules’. However, these norms can be obscure. For example, there is a very large online community called Incels (Involuntary Celibates)31 who have hit headlines for their misogynistic views. Incels are
a community of men who are not in, and unable to get into, a relationship. The community is deeply sexist and blames feminism for their predicament – being unable to find relationships with women.
27 Postmes, Tom, Spears, Russell, Sakhel, Khaled and de Groot, Daphne,’ Social Influence in Computer-Mediated Communication: The
Effects of Anonymity on Group Behaviour’, Personality and Social Psychology bulletin, 2001, 53: 1243- 1254, p 1244
28 Reicher, S. D., Spears, R. and Postmes, T., ‘A Social Identity Model of Deindividuation Phenomena’, European Review of Social
Psychology, 1995, 6(1): 161-198
29 Postmes et al 30 Ibid
The rules, which allow threatening and abusing language toward women, but don’t allow bragging about past relationships, are for most of us unfathomable but clearly dictate their behaviour within the group.
Although norms are infinitely changeable (think back 50 years to what acceptable behavior was then), they are so embedded that when we look at communities who follow contrasting norms - as Incels do in the previous example - we can react in disbelief or horror. In 2014 Elliot Rodger
committed a mass shooting in Isla Vista California. Before he carried out the shooting, Rodger posted his manifesto “Elliot Rodger’s Retribution” online where he outlined his motives for the upcoming attack. Rodger made it clear that he wanted to punish women for sexually rejecting him, “I don't know why you girls aren't attracted to me, but I will punish you all for it.”32 Rodger was a
member of the Incel community and is now widely celebrated by them. His actions in Isla Vista clearly affirmed the norms of the Incel community even though his horrifying actions appear anti-normative to those of us outside that community.
In order to understand how obscure norms can become so embedded we must remember how much time people spend online interacting with these communities. The average adult in the United Kingdom now spends nearly 9 hours per day online33 and this is just the average, many spend much
longer. Some might even spend longer online than they spend interacting in the real world. Embedding norms which are so contradictory to general societal norms cannot happen overnight. These norms are entrenched over hours and hours of interaction with the group.
Unsurprisingly firestorms are often targeted at individuals the perpetrators view as a threat. Henri Tajfel’s social identity theory may provide a reason for this.34 Tajfel argues that once we have
identified with a social group, we will sort the world into insiders and outsiders through social categorization. The presence of an outsider causes a heightened sense of similarity to those in the group in contrast to the outsider, and further embeds group identity. Similarly, when social identity is challenged by an outsider, group members more strongly align themselves with the group. Therefore, when an outsider’s actions are seen to attack the group, it is unsurprising that the group members come together to sanction them. This might suggest that groups with obscure norms
32 Elliot Rodger’s Retribution, https://genius.com/Elliot-rodger-elliot-rodgers-retribution-annotated accessed on: 1/07/2018
33 Woods, Ben, UK now spends more time online each day than sleeping (2016),
https://www.wired.co.uk/article/uk-spends-more-time-online-sleeping accessed on: 20/08/2019
34 Tajfel, H., & Turner, J. C., ‘An integrative theory of intergroup conflict’, in The social psychology of intergroup relations, ed. By W. G.
would actually be more likely to participate in firestorms, as their views are more likely to be threatened.
In summary, firestorms are anti-social but it is wrong to assume they are anti-normative. Firestorms on social media do not occur because users are anonymous and are therefore unafraid of
consequences; firestorms are the response to “(perceived) violating behaviours of public actors”35
and can therefore be seen as sanctions targeted at norm breakers. Group members will only enforce aggressive sanctions when experiencing deindividuation – a loss of self-awareness in the group – and contrary to common belief, deindividuation can occur in online communities as well as offline. In fact, the SIDE model suggests online communities might be even more influential to individuals than offline communities. Firestorms reinforce group-norms, however niche, and can therefore often appear contrary to common societal norms.
1.3 Online norm-enforcement
In terms of magnitude and aggression, the real-world norm-enforcing action which is most comparable to the firestorm is the ‘angry mob’. Yet angry mobs are extremely uncommon and firestorms frequent. If firestorms are norm-enforcing actions as we saw in the previous section, then it is not immediately clear why they are more common than offline norm-enforcing actions.
Firestorms rely on the numbers participating to increase rapidly and exponentially in ways that are irreplaceable offline. This phenomenon is achievable on social media due to several factors
including: clustered networks, echo chambers,36 low cost interactions, lack of non-verbal cues, large
online networks and a lack of negative feedback.
The term network clustering refers to the creation of transitive links between people in groups, meaning that if a is connected with b and c, b is probably also connected with c.37 Social media
shows significant network clustering. Clustering occurs because social media users choose who to follow, often following like-minded people or people similar to them in terms of age, background or interests. This is called homophily.38 Social media users also generally avoid following those they
35 Rost et al, p3
36 Pfeffer, J., Zorbach, T. and Carley, K. M., ‘Understanding online firestorms: Negative word-of-mouth dynamics in social media networks’,
Journal or Marketing Communications, 2014, 20(1-2): 117-128
37 Groeger, L. and Buttle, F., ‘Word-of-mouth marketing influence on offline and online communications: Evidence from case study
research’, Journal or Marketing Communications, 2014, 20(1-2): 21-41
38 McPherson, Miller, Smith-Lovin, Lynn, and Cook, James M., ‘Birds of a Feather: Homophily in Social Networks’, Annu. Rev. Sociol., 2001,
disagree with. Many social media platforms further exacerbate clustering by recommending connections to us. Those recommendations tend to be connections of connections or those who create similar content to people in our current network. Our network therefore becomes even more homogenous.
When a topic becomes popular within a clustered network, the topic can appear to echo at a user from all directions making it appear as if everyone is talking about it. This is called an echo chamber39
and they work as follows. On Twitter, Lucy follows Matt. Matt has seen an interesting piece of news about Brexit and tweets his disgust that there will be no second EU referendum in the UK. Due to homophily Matt’s followers are also likely to be disgusted with the news, and some tweet similar sentiments. Due to clustered networks Lucy is likely to also follow many of Matt’s followers and she therefore sees several people tweeting their disgust in a short space of time. Given that she is also highly likely to agree with the opinions of those she follows, she tweets her agreement.
Network clustering can mean one section of social media is completely consumed with a topic which other sections are entirely unaware of. Even if a social media user did want to discover which topics other social media networks were discussing, it has become increasingly difficult to do so. Echo chambers are exacerbated by social media, not only in their recommendations, but also because social media sites show their users limited content. Social media is acutely designed to show
favourable content to users. On Facebook for example, the news feed algorithm was created to keep us reading for as long as possible, as such it tends to only show us articles, status updates or
advertisements which it believes we will like as this will keep us reading. Generally, users like content they agree with so Facebook shows users agreeable content. The more users browse Facebook, the more it knows about them, and the more specific the content becomes. Disagreeable content is thus difficult to access.
Unlike in the real world on social media people can connect with hundreds, thousands or millions of users. The echo chambers of which they are a part are consequently vast and it is therefore
understandable that users start to believe that their views are supported by the majority of people. When a firestorm is sparked within a community, echo chambers mean that the same views are repeated across the users’ social media and the ideas communicated in the firestorm are thus strongly reinforced.
Engaging in a firestorm differs drastically from engaging in public norm-enforcement, like joining a public protest for example because engaging in a firestorm is very low cost. This means it is very easy to do; It is inexpensive and non time-consuming. Social media commenters can go online, make a comment, and leave straight away to make a cup of tea. In fact, Twitter, which limits the
characters of messages, actively encourages short and quick communications. The effort is so minimal, social media users don’t even have to write post, they can just ‘like’ or ‘share’ someone else’s. Joining a firestorm can be done at any time and in any location with ease and firestormers can communicate globally without any extra effort. Conversely joining offline norm-enforcing actions, like joining a public protest, can require significant planning or travel and can be very time
consuming. Given the ease of joining in a firestorm, it is unsurprising that we are more likely to engage in ‘norm-enforcing’ behaviour online than offline.
Another aspect of online interactions which vary from the real world, is the state of being one step removed. Being one step removed from the target e.g. behind a computer screen means there is a lack of nonverbal cues in online interactions, such as facial expressions or body language. Without nonverbal cues empathy is reduced, which might go some way towards explaining the sheer levels of aggression in firestorms (or on social media more generally for that matter).40 The context and
meaning of a statement or action can also be stripped or twisted online which can lead to hostile misinterpretations, exaggerations or outright lies.
Joining a firestorm is a way for users to gain kudos in the online community as well as affirming their identity within it. Online communities are often nebulous and vast meaning that it is very difficult to be noticed. Popular online communities will have a relatively constant stream of content being created with people posting to try and gain ‘likes’ or ‘upvotes’. Creating popular content is a
challenge, and consequently users will often post something which very clearly aligns with the group mentality i.e. something they know will be popular. Participating in a firestorm is an easy way to ensure popularity. Users gain approval by participating but for those trying to gain extra approval, posting ever more shocking content is a way to gain attention. The more shocking a post the more attention it elicits. For this reason, users constantly try to outdo one another, creating content that is evermore appalling and thus quickly escalating the level of abuse which victims of firestorms receive.
40 Rost et al
The chance of immediate negative feedback for online norm-enforcement is slim compared to similar actions in the real world and the chances of a retaliatory physical attack are essentially nil. Even if a firestormer is targeted for retaliation, which is unlikely given the sheer number of them, they can easily ignore the responses by leaving the platform (e.g. turning the computer off) or blocking people for example. Online firestormers can participate in an aggressive attack, and then effectively “sneak off” to safety.41 Firestormers are also usually safe from social sanctions, as their
family and friends are often unaware of their actions. Unless they have done something illegal, firestormers can therefore easily avoid consequences.
In summary, given that firestorms are norm-enforcing actions there is a clear benefit of engaging in them for group members as it asserts their alignment with the group and may gain them approval. Given that joining a firestorm is extremely low cost and users are unlikely to receive negative feedback it would appear that the benefits of joining clearly outweigh the costs. Furthermore, the lack of non-verbal cues mean that firestormers are less likely to confront the harm they have caused to the victim and are therefore unlikely to be dissuaded from participating by this.
2. Frustrating truth seeking
In the previous chapter, we saw how and why firestorms occur. In this chapter, we shall see how firestorms negatively impact the free speech of their targets. Mill’s theory on the freedom of expression will therefore provide the grounding for an argument on why firestorms warrant regulation. Theorists who have discussed restrictions on hate speech42 tend to focus on comparing
the harm in restricting versus not restricting speech but this approach has been unsuccessful. Returning to Mill’s original argument will show that he argues determinedly for free speech on the grounds that it is a method for truth seeking. Speech which upsets truth seeking by silencing others thus contravenes the purpose of free speech. Just as people require protection from the tyranny of the majority in order to protect individuality, they also require protection from firestorms, which I argue function as a tyranny of the minority by stifling marginalized opinions online.
2.1 The balancing of harms
41 Ibid
In the second chapter of ‘On Liberty’, ‘Of the Liberty of Thought and Discussion’ Mill defends his theory for total and complete freedom of expression. Mill exerts that freedom of expression is not only of value to the individual but also to the masses. Mill of course grounds his argument in Utilitarianism, suggesting that if free expression were only valuable to the individual then the silencing of one individual would be less harmful than the silencing of many, but the silencing of even one individual’s expression of their opinion is robbing the human race. This is because Mill’s theory is predicated on the value of truth and he argues that silencing even one opinion removes an opportunity to gain a stronger understanding of the truth.
Mill suggests that to silence the expression of an opinion is wrong whether or not it is correct. If the silenced opinion is true, then society has been deprived of the opportunity to replace a falsehood with the truth. If the silenced opinion is false, then society loses a clearer and more robust view of what the truth is. Either way, society has the opportunity to strengthen their understanding of the truth. This is not to say that the true opinion will always win. Society might hear a false opinion and wrongly think it to be true. However, silencing an opinion will mean society has lost the opportunity to have a stronger understanding of the truth.
Mill clearly rejects the regulation of speech even if the opinion expressed is vile and abhorrent.
if the arguments of the present chapter are of any validity, there ought to exist the fullest liberty of professing and discussing, as a matter of ethical conviction, any doctrine, however immoral it may be considered.43
Mill argues that we cannot interfere with speech for any reason even if it is immoral. This is not because Mill is suggesting that the immoral speech may be correct but because in order to ban speech, an entity – presumably the government – would need to decide which speech to ban. Making the decision to ban speech on the basis of its content would be to assume their infallibility.44
To refuse a hearing to an opinion, because they are sure that it is false, is to assume that their certainty is the same thing as absolute certainty.45
43 Mill, J. S., On Liberty, in On Liberty: John Stuart Mill, edited by Alexander, E. (Ontario, Canada: Broadview, 1999) original work published
1859
44 “All silencing of discussion is an assumption of infallibility” Mill, J. S., p60 45 Ibid
It is tempting to restrict firestorms on the basis that it is devoid of all possible argument that could be true. However, it seems unlikely that Mill would support this regulation, no matter how
contemptable the speech. Nevertheless, he does propose one caveat to his theory, the Harm Principle.
The only purpose for which power can be rightfully exercised over any member of a civilized community, against his will, is to prevent harm to others.46
Mill suggests that the only legitimate reason for which we might regulate speech is in order to prevent harm to others.
This is not to say that Mill advocated that all speech which ultimately causes someone to be harmed should be restricted. For example, he states that “an opinion that corn-dealers are starvers of the poor, or that private property is robbery, ought to be unmolested when simply circulated through the press.”47 In contrast Mill states that the same sentiment “delivered orally to an excited mob
assembled before the house of a corn dealer”48 would “justly incur punishment.”49 His argument
being that when spoken to the assembled mob, the expression incites the crowd to violence. Nevertheless, even if the view is only expressed in print, the corn-dealer might still suffer some harm. For example, the damage to the reputation of the corn dealer might mean customers are less eager to buy from him and the corn-dealer might therefore suffer financial hardship. Mill does not address these types of harms suggesting that he was not advocating for all harmful speech to be restricted. Therefore, although the harm principle is a necessary condition for the regulation of speech, it is not a sufficient condition.
Mill does not fully explore what harm would legitimately warrant the restriction of speech except to say that instigating harmful actions was a legitimate reason for restriction. Here is the full sentence about corn-dealers which is partially quoted in the previous paragraph:
Even opinions lose their immunity when the circumstances in which they are expressed are such as to constitute their expression a positive instigation to some mischievous act. An opinion that corn-dealers are starvers of the poor, or that private property is robbery, ought
46 Ibid, p52 47 Ibid, p101 48 Ibid 49 Ibid
to be unmolested when simply circulated through the press, but may justly incur
punishment when delivered orally to an excited mob assembled before the house of a corn dealer, or when handed about among the same mob in the form of a placard.50
It is clear that Mill is suggesting this speech should be restricted not by virtue of its content, which criticizes the corn dealers, but because it directly incites violence toward the corn-dealers. By making this statement in front of the angry mob, a reasonable person would expect the mob to react and therefore he should not do so. Context is clearly important as is the direct nature of the harm caused.
In balancing the harms of hate speech,51 a similar action to firestorms, commentators have struggled
to reach a consensus. Theorists have argued that speech can cause (or constitute) harm in many ways; it can cause (or constitute) subordination,52 it can legitimize discrimination and propagate
prejudice, it can enact discriminatory norms,53 it can cause psychological harm and of course it can
incite violence. On the opposing side theorists have contested whether or not this harm is done
directly by the speaker, or argued that the harm to one individual targeted by hate speech is minimal
compared to the harm to society if speech is restricted, however negligibly. Many of these arguments remain unconvincing either because it remains subjective whether or not the harm caused to society by the speech outweighs the harm caused to society by allowing the speech, or because the theorists are attempting to convince the reader that the harm caused by the speech is direct not indirect. Langton for example argues that pornography is (or could be at least) the subordination of women.54 Many remain unconvinced.
Mill’s argument for the liberty of thought and discussion is dependent upon the value of seeking truth, as well as the argument that truth seeking is best facilitated by the expression of a range of opinions, including the false ones. Mill argues that what we hold to be the truth needs to be constantly questioned, even if it is true. Our view of the truth will only be strengthened by questioning it.
50 Ibid
51 I focus here on hate speech, not firestorms as much of the literature focuses on individual cases of hate speech rather than the full
firestorm.
52 Maitra, I., ‘Subordinating Speech’, in Speech and Harm by Maitra, I. and McGowan M. K. (Oxford: Oxford University Press, 2012) pp
94-120
53 McGowan, ‘On ‘whites only’ signs and racist hate speech: verbal acts of racial discrimination’, in Speech and Harm, by Maitra, I. and
McGowan M. K. (Oxford: Oxford University Press, 2012) pp 121-147
That the only way in which a human being can make some approach to knowing the whole of a subject, is by hearing what can be said about it by persons of every variety of opinion, and studying all modes in which it can be looked at by every character of mind.55
In other words, in order to fully understand something, we must truly know the grounds for our opinions, and we must know at least the most common objections to it. One could therefore argue that firestorms in fact strengthen our view of the truth by giving victims an opportunity to respond and repudiate the claims made by firestormers. Unfortunately, the victims of firestorms are not often not capable of doing such a thing – as we shall see in more detail later – as firestorms often intimidate their victims in silence.
In order to evaluate firestorms, the balancing of harms approach is not appropriate. As with hate speech this approach is unlikely to produce a consensus. Given Mill's preoccupation was with truth rather than harm, the focus here should be on which actions detrimentally impact truth-seeking. Firestorms are clearly harmful, as we shall see, and thus satisfy the harm principle’s condition for restriction but it is because firestorms have the ability to silence their victims, and thus silence whole sections of online communities, that means they warrant regulation. This silencing means that firestorms have a negative impact on truth seeking.
2.2 The Tyranny of the Minority
In ‘On Liberty’ Mill discusses what he terms ‘the tyranny of the majority’. Mill argues that democratic societies should be wary of the will of the most numerous or the will of the most active part of the society as this will can be exerted as a kind of ‘social tyranny’. Although other prominent thinkers before him had recognised the ability of the will of the majority to be imposed through democratic instruments, Mill recognised this tyranny could also be imposed socially. Mill argued that public opinion can be disastrous for individuality as the majority tend to impose their views and values on others, leaving those with marginalized views and opinions unwilling to express themselves.
The tyranny of the majority is thus a term to express how social norms are imposed on individuals. Those who deviate from the norm, are punished through social punishments and learn to keep quiet about their dissenting thoughts and opinions. Mill argued that social tyranny, although much less
extreme than political oppression, can have a much deeper impact on individuals as it manages to suppress the formation or development of individuality when it strays too far from the mainstream.
Though not usually upheld by such extreme penalties, (the tyranny of the majority) leaves fewer means of escape, penetrating much more deeply into the details of life, and enslaving the soul itself.56
Although social punishments, such as social isolation and disapproval, are targeted at individuals or groups, any onlooker who similarly deviates from the norm will also be discouraged from displaying individuality.
Chapter 1 showed that firestorms are often a method for enforcing social norms. As such they could be a technique through which the tyranny of the majority operates. However, chapter 1 also
highlighted that the norms which are enforced by firestorms are not always held by the majority, sometimes they are the norms of a niche group. Whereas in Mill’s day niche groups would have generally held little power, social media has allowed geographically disparate likeminded people to form communities of staggering size and influence. Firestorms, which enforce norms outside of the mainstream, therefore function as a tyranny of the minority.
The tyranny of the minority functions in much the same way as the tyranny of the majority, in that it can regulate and mold the behaviour of targets (and onlookers) by distributing social sanctions to those who contravene social norms. However, contra to the tyranny of the majority, the tyranny of the minority originates outside of the mainstream and therefore enforcing group norms, not general social norms. This does not mean it is any less harmful.
Before she became a victim of Gamergate, feminist blogger Anita Sarkeesian was subject to a firestorm after launching a Kickstarter campaign57 to raise funds to create a series of videos
examining sexist tropes in video games.58 Sarkeesian had stepped into very dangerous territory by
criticizing gaming. She faced what she called a ‘cyber mob’59 in a classic example of the tyranny of
the minority. Video gamers are by no means a homogenous group but there is a vocal subgroup of gaming, generally young, white, male, gamers who often express anger at people who critique
56 Mill, J.S. p47
57 An online crowd-funding website
58 Sarkeesian, A., TEDxWomen Talk about Online Harassment & Cyber Mobs, 2012,
https://feministfrequency.com/video/tedxwomen-talk-on-sexist-harassment-cyber-mobs/ accessed on: 20/08/2018
gaming culture. The vitriol Sarkeesian received was graphic and explicit in its detail.60 A game where
users were invited to “beat the bitch up” was created where an image of Sarkeesian become increasingly injured as users clicked on the screen. Pornographic images doctored to look as if Sarkeesian was being raped by video games characters were distributed online and sent to her repeatedly. In a talk she gave about the incident, Sarkeesian describes the firestorm as a game. Firestormers gained praise as rewards for producing ever worse ways to abuse her.
The attack on Sarkeesian was a horrific example of the tyranny of the minority. Although the views of the firestormers are hopefully not mainstream, they managed to cause a significant amount of harm not only to Sarkeesian, but also to other people who might critique gaming. The attack aimed to punish Sarkeesian for criticizing gaming culture as a vocal subsection of the gaming community clearly saw her as a threat and yet it also served as a warning to those who might follow in her footsteps. The treatment Sarkeesian received was looked at with trepidation by all women involved in the gaming industry concerned that they might be next, indeed some of them were as the firestorm went on to attack many other women.
The impact of the tyranny of the minority goes well beyond the initial victim, as onlookers also absorb the lesson which the firestorm conveys. It may be that some women have been put off gaming by the incident, or no longer dream of working in the industry. It is likely many women have hesitated before critiquing the gaming industry or stopped doing so altogether. It is impossible to know the true extent of the damage.
Mill praised eccentricity in society, arguing that it was proportional to “the amount of genius, mental vigour, and moral courage it contained.”61 He lamented the tyranny of the majority because it
prevented the expression of eccentricity and alternative opinions. The tyranny of the minority online similar represses alternative views. Just as Mill suggested the public needed protecting from the tyranny of the majority
There is a limit to the legitimate interference of collective opinion with individual independence; and to find that limit, and maintain it against encroachment, is as
60 Ibid
indispensable to a good condition of human affairs, as protection against political despotism.62
It is clear that in an online environment, protection from the tyranny of the majority is not enough. There also needs to be protection from the tyranny of the minority.
3. Misogyny and the silencing of women
Although all firestorms by definition include vile and insulting speech, women who are targeted by firestorms are more likely to experience speech which attacks or insults them on the basis of their gender. In other words, women are more likely to face hate speech:
Speech that attacks, threatens, or insults a person or group on the basis of national origin, ethnicity, color, religion, gender, gender identity, sexual orientation, or disability.63
In the previous chapter, through studying a firestorm targeted at Anita Sarkeesian we saw an example of how firestorms target individuals in order to punish them for breaking the norms of online communities. We also saw how firestorms can function as the tyranny of the minority i.e. niche communities imposing their views on their victims and onlookers. In this chapter, we shall discuss firestorms targeted at women and how they often enforce sexist norms of repugnant online communities. We will also see how these misogynistic firestorms can silence and marginalise women’s views online.
3.1 Silencing women
Communication is the main reason we value speech. If speech did not serve as a method for communicating our thoughts and opinions, then it would surely not be valuable to us as individuals. Returning to Mill, without communication it seems unlikely that speech would be valuable to society more generally either. Silencing is thus problematic because it frustrates a speaker’s ability to communicate. This in turn must mean that a speaker is silenced when their ability to communicate is frustrated.
62 Mill, J. S., p47
In ‘Words that Silence? Freedom of Expression and Racist Hate Speech’ Caroline West argues that silencing can manifest itself in many ways. One can be silenced in being prevented from making words, by having one’s tongue chopped off for example, or by being kept in solitary confinement. In the first example, the victim has been prevented from producing words altogether, in the second they were able to produce words but unable to distribute those words. In neither of these scenarios would we recognise that the speaker has the right to true freedom of expression. Production of speech and distribution of speech are thus important elements of free speech.
There are of course ways in which an individual might overcome production and distribution failure. By learning sign language for example, or by sneaking notes out from under the locked door in the previous example. However, these obstacles to speaking are arduous. Although the speaker may persevere, given the discriminatory obstacles placed in their way they may not, and the world would therefore be lacking their opinion.
Freedom of speech therefore requires the opportunity for meaningful production and distribution of speech. In other words, we must have the opportunity to make sounds (or actions) which can be heard (or seen) by others in order to meaningfully be able to say we have free speech. This is not to argue that some logical and fair restrictions (legal or practical) cannot legitimately impact the production and distribution of our speech without negating that we have free speech. For example, we don’t all have the opportunity to speak before parliament on a whim, for practical reasons we have to follow bureaucratic procedures in order to do so. This is not an unfair restriction on freedom of speech, unless those bureaucratic hurdles are too complicated as to make it effectively impossible to speak in front of parliament.
Production and distribution are only two key elements of free speech. In order for speakers to be able to communicate, speech also needs to be considered and comprehended by its audience. In other words, there needs to be consideration, meaning that speech is not dismissed out of hand, and
comprehension meaning that speech can be wholly understood by the listener. For instance, if a
speaker is able to make speech by producing sounds and she is unrestricted in her movements and therefore able to share that speech with others, but the audience, having seen that she is a woman, dismisses her as ‘silly’ and therefore shelves her ideas without fully engaging with them, then the speaker has faced consideration failure. Alternatively, imagine if the same woman were speaking, and the audience did not dismiss her views. In fact, they were listening intently and very much wanted to understand her. This time however she was standing behind a glass screen and her voice
was distorted such that the audience could not understand her. This is an example of comprehension failure. In both examples the speaker has not failed to produce or distribute speech, but she has still failed to communicate.
Of course, when people speak they cannot always expect listeners to engage with their speech. There are many legitimate reasons why audiences disengage. For instance, perhaps because they are prejudiced and do not think the speaker is capable of having valid views – as in the previous example - but also perhaps because they are boring, and the audience is struggling to concentrate on what they are saying. In this later scenario, the speaker has failed to communicate but through something over which they have control. However, imagine the speaker in the early example, where the audience disregarded her speech due to prejudice. Again, it could be argued that the speaker was not silenced, as she has still produced and distributed words, and the words have been understood by the listener, but they have been disregarded without consideration. This could happen to anyone yet imagine that this speaker encounters the same response whenever she speaks, so much so that she is unable to have her views considered in any circumstances. If this were the case, then she would be rendered unable to communicate and she would be effectively silenced.
It is clear that arguing for the totality of these four aspects of free speech: production and distribution, consideration and comprehension is incorrect. Sometimes speakers fail to produce speech, distribute it to whom they want to hear it, or the intended audience fails to listen or understand it. In these instances, the speaker has failed to communicate but it is not the case that they have been wholly silenced. There is no guarantee that every time people speak they will be heard or understood. However, West argues that true free speech is a negative liberty. Free speech doesn’t provide a guarantee of comprehension by the audience, but it does require that
comprehension not be unfairly prevented by the actions of others. Therefore, free speech requires that agents do not act in ways which unduly prevents the comprehension, consideration,
distribution and production of the speech of others.64 Consequently a firestorm which prevents
women from communicating could therefore be considered detrimental to free speech.
In the UK, feminist campaigner Caroline Criado-Perez became the target of a firestorm after starting a campaign for a woman to appear on one of the British banknotes. She faced vile and malicious attacks and threats. One tweeter asked another if he would join him in raping Criado-Perez, one
64 For a full explanation see West, C, ‘Words that Silence: Freedom of Expression and Racist Hate Speech’, in Speech and Harm, by Maitra,
