MAGISTERARBEIT / MASTER’S THESIS 

Titel der Magisterarbeit / Title of the Master‘s Thesis 

„Freedom of expression and hate speech 
 in the case law of the European Court of Human Rights  

and in the practice of social media“ 
 

verfasst von / submitted by 

Darya Novatorova, BA 
 

angestrebter akademischer Grad / in partial fulfilment of the requirements for the degree of 

Magistra der Philosophie (Mag.phil.) 

Wien, 2019 / Vienna 2019  

Studienkennzahl lt. Studienblatt / 
degree programme code as it appears on 
the student record sheet: 

UA 066 841 

Studienrichtung  lt. Studienblatt / 
degree programme as it appears on 
the student record sheet: 

   Magisterstudium Publizistik-u. 

   Kommunikationswissenschaft  

Betreut von / Supervisor: 

 

Univ.-Prof. Dr. Katharine Sarikakis  

 



 2 

Table of contents 

 

 
1. Introduction………………………………………………………………………..4 

2. Theoretical background …………………………………………………………...8 

           2.1. Freedom of expression………………………………………………………...8  

           2.2. Defining hate speech…………………………………..….………………….10 

           2.3. Hate speech and hate crimes……………………………………………...….14 

           2.4. Hate speech: contextual typology……………………………………………16 

           2.5. Other categorisations of hate speech…………………………………………22  

           2.6. Philosophical and practical reasons to ban hate speech……………………...25  

           2.7. International legal base for freedom of expression and hate speech’ 

regulation...........................................................................................................................28 

           2.8. Combating hate speech: minimal and extensive regulation approaches……..29  

           2.9. European and American approaches towards hate speech………………...…33  

           2.10. Online hate speech: general overview……………………………….……..35   

           2.11. Alternative ways to combat hate speech…………………………….……...37   

             

            

       3.   Research questions and methodology…………………………...………………40 

       

       4.  European Court of Human Rights’ approach towards hate speech……………...45 

           4.1. Definition and types of hate speech according to the ECtHR………………..45 

           4.2. ECtHR’s algorithm of establishing hate speech……………………………..48 

           4.3. Overview of ECtHR’s hate speech practice in 1976-2012…………………..50 

           4.4. ECtHR’s case law concerning hate speech in 2012-2019................................57  

                  4.4.1. Political, racial, ethnic hate speech …………………………………...58 

                  4.4.2. Religious hate speech………………………………………………….68 

                  4.4.3. Hate speech based on sexual orientation……………………………....71 

                  4.4.4. Online hate speech cases in the practice of the ECtHR.........................74 

           4.5. Summary of the ECtHR’s practice concerning hate speech……………..…..79  

 



 3 

       5.   Social media and online hate speech…………………...……………………..…85 

            5.1. Hate speech definitions in social media’s guidances.……………….….…...86 

            5.2. Regulation of hate speech on social media………………………….………93  

            5.3. Alternative ways of combating online hate speech: attempts of automatic 

regulation………………………………………………………………………………...96  

             5.4. Other initiatives to counteract online hate speech………………………….97                  

        

       6. Conclusion…………….……………………………………………………….…99 

                     

       Table 1. Argumentation of the ECtHR in selected hate speech-related 

 cases 1976-2012………………………………………………………………………...55 

       Table 2. Argumentation of the ECtHR in selected hate speech-related  

cases 2012-2019………………………………………………………………….………80 

       Table 3. Hate speech definitions by Facebook, YouTube and Twitter……………...92  

 

 

       7. References…………………………………………………………………….....107  

 

       8. Abstract……………………………………………………………………….…122 

       9. Zusammenfassung…………………………………………………………….....123  

 

 

 

 

 

 

 

 

 

 

 



 4 

1. Introduction 

 

 

The collision between freedom of expression and hate speech is one of the most essential 

problems of the modern media law. On the one hand, freedom of expression is a 

fundamental human right and one of the main conditions for the existence of a truly 

democratic society. In the western countries intellectuals have emphasized this fact 

starting with the Modern history. After the Second World War the right for freedom of 

expression was repeatedly reiterated and secured in the major Conventions and 

Declarations of the 20th century, including the European Convention on Human Rights 

(1950). In the 21st century freedom of expression remains a crucial and an absolute 

component of human freedom and the bravest of us continue to fight for it all over the 

world.    

 

Hate speech, on the other hand, is a destructive and offensive content – words, images or 

videos – that aims to encourage hatred and discrimination in the society.  While rapid 

development of the Internet and new media, without doubt, has considerably contributed 

to the realization of freedom of expression worldwide, it has also facilitated the 

dissemination of hate speech, making it a global issue of our time. The key question is 

how to establish control over the latest, without violating the former: who and how 

determines, what hate speech is and how to combat it? Do different concerned parties – in 

particular, traditional courts and social media – have equal approaches towards hate 

speech and if not, how do they differ? What are the ways for combating hate speech that 



 5 

social media have already developed, are they effective and what other alternatives could 

be employed?  These are the underlying research questions of the current thesis.  

 

As the European Court of Human Rights has acknowledged, no universally accepted 

definition of hate speech exists in modern media law, which, of course, further 

complicates combating this harmful phenomena (ECtHR, Factsheet hate speech, 2019). 

To our knowledge, one of the most precise definition of hate speech was given by 

Council of Europe who in 1997 defined it as “all forms of expression which spread, 

incite, promote or justify racial hatred, xenophobia, anti-Semitism or other forms of 

hatred based on intolerance, including: intolerance expressed by aggressive nationalism 

and ethnocentrism, discrimination and hostility against minorities, migrants and people of 

immigrant origin” (Recommendation No. R 97(20), Appendix).  

 

As will be demonstrated later, by no means this definition is the only one or universally 

accepted. Though it must be emphasized that already in this definition of Council of 

Europe it is possible to establish two necessary components that together constitute hate 

speech: (1) promotion/justification/dissemination of hatred based on (2) one of the many 

possible causes, whether it is nationality, ethnicity, migrant origin or affiliation with a 

minority group of any kind. Two further “causes” that for some reasons were not 

included in the definition of Council of Europe in 1997, are gender and religion. In fact, 

gendered hate speech and hate speech based on religious beliefs are a widespread issue of 

Internet era and will be discussed later in the first part of the study.      

 



 6 

In an attempt to answer the research questions specified in the respective paragraph and 

taking into account the significance of the European Court of Human Rights for the 

European Convention on Human Rights1 and, in general, for the European law practice, it 

was decided to focus in the present thesis on the recent case law of the European Court of 

Human Rights related to hate speech and to compare the Court’s approach with such of 

the major social media – namely, Facebook, Twitter, and YouTube. The choice of these 

social media was dictated by their evident popularity, wide spreading in the society and 

influence over people’s lives.      

 

After the theoretical overview of the related issues in the first part of the study, in order to 

determine how the European Court of Human Rights separates protected speech that can 

“offend, shock or disturb” and hate speech that needs to be forbidden, a content analysis 

of the selected Court’s cases was conducted in the second part of the thesis. The Court’s 

definition of hate speech and proposed typology are also provided in this part of the 

study.   

 

Accordingly, the third part of the thesis is dedicated to the practice of the above 

mentioned social media. Their current policies regarding hate speech, as well as the 

respective guidance for the users are analysed. Attempts of combating hate speech in the 

alternative ways are discussed in the paragraphs 5.3-5.4.   

  

                                                 
1 The European Court of Human Rights was established in 1959 for protecting the fundamental human 

rights enshrined in the Convention. See: 

http://www.echr.coe.int/Pages/home.aspx?p=basictexts&c=#n1359128122487_pointer  
 

http://www.echr.coe.int/Pages/home.aspx?p=basictexts&c=#n1359128122487_pointer


 7 

Finally, key observations regarding the comparison between European Court of Human 

Rights’ approach towards hate speech and that of the social media are provided in the 

Conclusion.       

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 



 8 

2. Theoretical background 

 

2.1. Freedom of expression  

 

There is no need to prove what essential role freedom of expression plays in any 

democratic society: it is its keystone and an indispensable condition. The right for 

freedom of expression is regarded as one of the most significant human rights in many 

countries, especially – in the states-members of the Council of Europe2 whose main 

normative act – the European Convention on Human Rights – proclaims and guarantees 

freedom of expression in the well-known Article 10.  

 

According to this Article, freedom of expression consists of two principal parts: (1) 

freedom to hold opinions and (2) to receive or to share information. Significantly, in the 

second part of the same Article 10 it is emphasised that freedom of expression is not 

absolute and may go hand in hand with responsibilities. The latest – possible restrictions 

– must be “prescribed by law” in the interest of a democratic society, protecting various 

concepts, but among others – public safety, order, health and morals, and even “the 

reputation or rights of others” (European Convention on Human Rights, Art. 10.2). Thus, 

it is evident, that the European Convention on Human Rights codifies freedom of 

expression while at the same time acknowledging its possible limits and leaving 

sufficient legal space for the prohibition of hate speech.       

 

                                                 
2 Current countries-members of the Council of Europe can be found here 

http://www.coe.int/en/web/portal/47-members-states 

http://www.coe.int/en/web/portal/47-members-states


 9 

The debates around freedom of expression versus hate speech mainly circles around one 

issue: how to maintain diversity of thoughts and views, while at the same time preserving 

dignity of all groups of citizens (Maitra, McGowan, 2018, p.305). The second part of the 

Article 10 of the European Convention on Human Rights – important and indispensable 

as it is – is very broad and  therefore has caused many questions and interpretations by 

courts, as well as became an object of analysis for generations of media law researches 

(among many others, Korn, 2014; Holoubek, Kassai, Traimer, 2014). Provided by the 

Convention list of conditions, when the restrictions of freedom of expression are 

necessary in a democratic society, makes it impossible for the courts to use the Article 

10.2 automatically. On the contrary, in each case courts have to consider all the 

circumstances and the details. 

 

To make things even more complicated, when discussing freedom of expression in 

addition to the Convention it is necessary to take into account the case-law of the 

European Court of Human Rights.  In the famous and historical case Handyside v. United 

Kingdom, 1976, the Court has stated that not only that information should be free that is 

considered inoffensive or indifferent, but also any ideas that can “offend, shock or 

disturb” any particular group of people (Handyside v. United Kingdom, 1976, par. 49). 

The Court motived it by “pluralism, tolerance and broadmindedness” that are the pillars 

for any democratic society (ibid).  

 

For decades, this statement about the information and ideas that can “offend, shock or 

disturb” was a keystone of all modern media law practice. It establishes the broad 



 10 

understanding of freedom of expression and also implies tolerance for some speech as a 

price for living in “a democratic society”. As was summarized by McGonagle, 

democracy, too, has “rough edges” and “tough talk” should be regarded as a part of 

public debate (2013, p.5).  

 

At the same time, the decision in Handyside v. United Kingdom does not abolish or 

contradict the second part of the Article 10 quoted earlier: there is still information that 

must be restricted when prescribed by law and “in the interests of a democratic society”. 

In particular, it concerns hate speech.  

 

 

2.2. Defining hate speech 

 

As was mentioned in the Introduction, there is no universally accepted definition of the 

term “hate speech” (ECtHR, Factsheet hate speech, 2019). In fact, hate speech definitions 

vary by country, jurisdictions and particular type of law (D'Souza, Griffin, et al, 2019, 

p.943). One of the definitions that were found during the current research was provided 

by the European Union Agency for fundamental rights (FRA) which defines hate speech 

as “the incitement and encouragement of hatred” as well as “discrimination or hostility 

towards an individual” based on prejudice connected with any particular characteristic, 

whether it is “sexual orientation or gender identity” (Hate Speech and Hate Crimes 

Against LGBT Persons).   

 



 11 

In comparison to the definition of hate speech given by the Council of Europe and quoted 

in the Introduction, the definition of the European Union Agency for fundamental rights 

can be regarded as incomplete, as it clearly excludes racial hate speech, or hate speech 

based on nationality, ethnicity, migrant origin etc. On the other hand, it is worse noticing 

the presence of the same two components that were highlighted in the definition of the 

Council of Europe: (1) encouragement of hatred (2) based on one or another 

characteristic of a victim.  

 

Unfortunately, in practice defining hate speech is not that straightforward. As was 

observed by Parekh, in different countries different speech is considered to be hate 

speech (Parekh, 2012, p.40). In some instances this speech expresses certain views, but 

does not call for actions; in another the speech under consideration is abusive but not 

threatening; in some situations the speaker expresses dislike of a certain group of people, 

but not hatred, and wishes no harm (ibid). Parekh emphasizes that this confusion is a 

direct result of the tendency to summarize under the concept of “hate speech” all kinds of 

uncivil or hurtful speech (ibid). 

 

Parekh claims that it is more reasonable to define hate speech as incitement of hatred 

against a group of people, based on their nationality, sexual orientation, religion, gender, 

etc. In other words, the researcher supports the definition proposed by the Council of 

Europe and quoted in the Introduction. However, according to him, the concept of hate 

speech consists primary of three, not two components: 



 12 

 (1) It is directed against concrete individual or group of people based on a particular 

feature; 

(2) It stigmatizes this individual or group of people by prescribing them highly 

undesirable qualities and  

(3)  It utilises these negative qualities to make this individual or group “a legitimate 

object of hostility” (Parekh, 2012, p. 41). Hate speech does not necessarily results in 

violence and must not necessarily be expressed with offensive of abusive language (ibid).    

 

Gelber and McNamara also admit that, whereas the term “hate speech” is widely used, it 

does not have a single meaning (2016, p.324). For their own study they adopted the three-

component definition of Parekh.  

 

Richardson-Self, while reiterating the fact that there is no one accepted definition of hate 

speech among scholars, underlines that it is speech that a) is hostile and b)silences, 

humiliates, threatens or discriminates and c) targets group of people based on common 

traits, such as race, religion beliefs, disability, sexual orientation or gender (2017, p.256).  

Hate speech per se in an oppressive act, argues Richardson-Self (ibid, p.257).    

 

Yong, too, defines hate speech as speech that attacks people based on their particular 

characters, whether it is race, nationality, gender, sexual orientation, religion believes or 

any other identity marker (2011, p.386). Yet, the researcher shares the view that the term 

“hate speech” is “unsatisfactory” (ibid). Yong debates that different kinds of speech are 



 13 

understood under this term and that appropriate response to them should differ from case 

to case (ibid).  

 

Moreover, providing a definition to hate speech in practice becomes even more difficult 

in the countries which experience great social tensions or had especially troubled 

historical past. For example, hate speech is a relatively new term in South Africa, where 

racism and discrimination until recently were institutionalized in the apartheid policy 

(Davids, 2018, p.297). It is worth noticing that now hate speech is criminalized in the 

country, but nevertheless it experiences a significant increase in both hate speech and 

hate crimes (ibid).     

 

Finally, there is a view among some scholars that to create one final definition to hate 

speech is difficult or even not possible because hate speech is a continuum (D'Souza, 

Griffin, et al, 2019, p.958). Whereas in this case the researchers initially concluded it in 

relation to gendered hate speech, the idea of hate speech as continuum or spectrum 

(McGonagle, 2013, p.4) seems to reflect the nature of this phenomenon. In particular, it 

could explain not only the problem of definition, but also different approaches and 

debates regarding hate speech regulation.  

 

Nevertheless, despite the lack of one single definition, it is possible to observe certain 

similarities in the proposed ones. To sum it up, we propose to understand under hate 

speech, utterances that (1) encourage hatred (2) towards individuals or groups of people 

(3) based on various particular group characteristics (4) which are stigmatized and (5) are 



 14 

employed to legitimize hostility. Furthermore, hate speech (6) may be understood as a 

continuum that (7) does not necessarily result in violence, but (8) silences, discriminates 

and threatens targeted individuals or groups.  

 

 

2.3. Hate speech and hate crimes 

 

Scholars often acknowledge that hate speech involves incitement of hatred (D'Souza, 

Griffin, et al, 2019, p.945), and thus the concept of “hate speech” is sometimes 

misunderstood as a synonym to the notion of “hate crime”. Whereas these concepts are 

related, it is significant for the purpose of the current thesis to underline the difference 

between them.  

 

Hate crimes can be regarded as violent manifestations of intolerance (OSCE, 2009, p.11) 

that usually have an impact not only on the victim, but also on the whole group to which 

the victim belongs. From other crimes hate crimes can be distinguished by bias motive. It 

is important to understand that hate crime can be manifested as an intimidation, threats, 

assaults, murder or similar criminal offence  (ibid, p.16).  

 

The bias motive and criminal offence which have occurred are two principal features of 

the hate crimes. It means that a perpetrator selects a victim not randomly, but on the basis 

of his or her belonging to a special group. Thus, the key issue is what the victim 



 15 

represents, and not who he or she is. (ibid, p.17).  Race, language, religion, gender and 

similar aspects are possible examples of basis of the hate crimes (ibid, p.9).  

 

On the other hand, hate speech will still be penalized in some countries even if no 

criminal base offence have occurred, but only speech that promotes hatred or insults 

certain group of people (ibid, p.25). For example, if someone violates a person’s or 

nation’s honor or dignity, it will be hate speech, but not a hate crime.  

 

In other words, the term “hate crime” is used for acts, whereas the concept of “hate 

speech” is related to discriminatory views/speech (ibid, p.17). Thus, specific content – 

words, images, videos – can be regarded as a principal future of hate speech, but not that 

of hate crime.  At the same time, hate speech can easily lead to hate crime. For instance, 

Davids argues that this is what happens in Charlottesville in August 2017 when a rally of 

supremacists and nationalists evolved into violence (2018, p. 298). On the other side of 

the globe, in Australia gendered hate speech fuels gender-based violence, which is a view 

of government bodies and some scholars (D'Souza, Griffin, et al., 2019, p.940). Yet, there 

is a notable absence of laws regulating gendered hate speech in the country (ibid).  

  

OSCE believes that legislation is only a part of the possible answer to the problem of hate 

crimes (OSCE, 2009, p.11). In order to combat hate crimes, there is a need in a 

comprehensive national programs which should also include education, special training, 

accurate data collecting, etc. (ibid, p.12).  On the other hand, there are extreme variations 



 16 

between hate speech laws in different countries, which can be explained by different 

constitutional and even philosophical approaches (ibid, p.26).  

 

 

2.4. Hate speech: contextual typology 

 

Whereas one universally accepted typology of hate speech does not exist, it is possible to 

establish such contextually based on the most common themes that are to be found in hate 

speech examples, namely:  

 

1) Hate speech based on nationality, race or ethnicity, including hate speech against 

people with migrant origin and incitement to genocide 

 

Racist hate speech exists in all parts of the world and constitutes a critical challenge for 

human rights (UN, General Recommendation 35, 2013, p.10). Racist hate speech does 

not necessarily equals to explicitly racial comments but may use indirect language (ibid, 

p.3). Among others, common targeted groups of this type of hate speech are indigenous 

peoples and people with migrant background including refugees and migrant workers 

(ibid, p.2).  Davids debates that racist hate speech not only results in violence, but also 

leads to discrimination of a particular group of people (2018, p.298).  

 

According to Maitra and McGowan, regulation of racist hate speech should not involve 

any concerns related to freedom of speech (2010, p.370). In turn, while discussing 



 17 

approaches towards racist hate speech and its qualification as an offence punishable by 

law, UN Committee on the Elimination of Racial Discrimination proposed to take into 

account several contextual factors, such as  

- content and form of a particular commentary (for instance, whether it is provocative or 

not);  

- existent economic, political and social climate (whether there is already a discrimination 

against a targeted group or not);  

- potential influence of the speaker and his/her audience;  

- potential reach of hate speech;   

- the objectives of it (UN, General Recommendation 35, 2013, p.5). 

 

Although Committee on the Elimination of Racial Discrimination introduced this 

contextualization to the racist hate speech only, analysis of contextual factors might be 

useful for approaching other types of hate speech as well.       

 

 

2) Political hate speech, including anti-Semitist speeches or Holocaust denial 

 

Political expression and hate speech related to political issues should enjoy different legal 

protection, but in practice it is often difficult to distinguish between them, claims 

McGonagle (2013, p.19). In case of ambiguity, the researcher proposes to analyse not 

only the content of the speech itself, but also its form, context, position of a speaker in a 



 18 

society and aim of speech (ibid). For instance, a clear separation should be made between 

speech that aims at contributing to public debate and speech that incites hatred (ibid).  

 

Understandably, freedom of political expression should be protected in a democratic 

society. However, this freedom is not absolute, emphasises McGonagle, but involves 

responsibilities for the speaker (ibid, p.20).  As will be demonstrated in the second part of 

this thesis, protecting political expressions while simultaneously prohibiting political hate 

speech is a difficult challenge even for the European Court on Human Rights. This 

collision between political expression and political hate speech may become even more 

acute online (ibid). 

 

From the legal perspective, instances of Holocaust denial are a more clear issue. 

European Court of Human Rights determined in the case Garaudy v. France that the mere 

questioning of the Holocaust constitute an incitement to hatred against Jews and as such, 

contradicts the values of the Convention (2003, par. 23).  

 

 

3) Hate speech based on religious beliefs or lack of such 

 

One could assume that religious hate speech involves discrimination based on religious 

believes, but Bonotti emphasises that the relation between hate speech and religion is not 

that straightforward (2017, p.259). Firstly, there is hate speech directed against religious 

minorities. Secondly, there are religious people who attempt to justify their hate speech 



 19 

against particular groups of people based on, for instance, their sexual orientation, with 

an argument of religious freedom (ibid). In fact, they utilize hate speech as a means of 

propaganda of their faith (ibid, p.272).  

 

In this case, maintains Bonotti, religious believers spreading hate speech must be treated 

as any other citizens (ibid). The argument about religious freedom cannot be taken into 

account because there are other ways to spread their religious message, as well as express 

their disapproval of other people’s lifestyles (ibid).  

 

 

4) Gendered hate speech 

 

Analysis of gendered hate speech is rare, argues Richardson-Self who believes that 

separate study of this kind of hate speech is required (2017, p.257). Interestingly enough, 

the researcher makes a distinction among (a) misogynistic speech that she considers to be 

hate speech, and (b) oppressive sexist speech that in her opinion does not amount to hate 

speech, but can be, nevertheless, violent (ibid). On the other hand, speaking about 

misogynistic speech versus sexist speech, D'Souza, Griffin, et al. defend their choice of 

the term “gendered hate speech”, emphasizing that it is an ongoing construction of gender 

that involves using gendered norms for attacks on people or groups of people (2019, 

p.955). Yet, D'Souza, Griffin, et al. agree that language maintains stereotypes that exist in   

society about gender, and thus gendered hate speech can be regarded as a means of 



 20 

spreading misogynist hostility in order to support patriarchal structures in a society (ibid, 

p.967). 

 

Explaining the difference between misogynistic speech and oppressive sexist speech, 

Richardson-Self clarifies that misogyny involves hostility against women, whereas 

sexism is an instrument of reinforcement or justification of patriarchy (2017, p.261). 

Speech that belittles or patronizing women – for instance, calling grown women “girls” –  

can be regarded as morally unaccepted and sexist, though it is not hate speech (ibid, 

p.262). On the other hand, misogynistic hate speech usually contains direct hatred 

towards women which often manifests itself through rape-threats or comments (ibid, 

p.265).  

 

In contrast to other types of hate speech, misogynistic utterances usually target not a 

single group – that is, not all women – but only women who do not comply with 

patriarchal standards. Richardson-Self calls this phenomena “intradivisional speech” 

(ibid, p.267). Yet, this kind of speech should still be regarded as hate speech as it 

degrades and dehumanizes targeted people as much as any other type of hate speech 

(ibid, p.268).     

 

Other scholars maintain that gendered hate speech must be analyzed in socio-political 

context where it enforces patriarchal structures and norms (D'Souza, Griffin, et al, 2019, 

p.940). Gendered hate speech is inextricably connected with traditional gender 

boundaries which results in women being target of it, for instance, for voicing their ideas 



 21 

or participating in society’s life (ibid, p. 956). Nevertheless, whereas women and girls are 

much more likely to become victims of gendered hate speech, it should be noted that it is 

not necessarily directed at females only: men, too, can become victims of it if they do not 

comply with traditional roles (ibid).  

 

Interestingly enough, some scholars further distinguish so-called “conservative” 

definition of gendered hate speech and “progressive” one (ibid, p. 958). They argue that 

the former focuses on protecting public order, whereas the latest – on protecting the 

victims (ibid). In other words, supporters of the “conservative” view on gendered hate 

speech prohibit it because it may cause criminal actions, whereas supporters of the 

“progressive” view simply ban gendered hate speech because of its harm to targeted 

individuals (ibid, p.959).  

 

 

5) Hate speech based on sexual orientation 

  

European Court of Human Rights first recognised homophobic hate speech in 2012, in 

the case Vejdeland & Others v. Sweden (McGonagle, 2013, p.12). The Court did not 

provide a full definition to it, but stated that homophobic hate speech is directed against a 

person of group of people based on their sexual oritnation (Vejdeland & Others v. 

Sweden, 2012, par.42). The Court debated that despite widespread of homophobic hate 

speech, there were no established standards for approaching the problem in 2012 (ibid). 

 



 22 

As it is often the case with hate speech, the boundaries between hate speech based on 

sexual orientation and gendered hate speech can be blurred. Thus, lesbians are often 

attacked on both grounds simultaneously which supports the assumption that hate speech 

is employed to control gendered boundaries and behaviour (D'Souza, Griffin, et al, 2019, 

p.955).  

 

Overall trend however is that homosexual or bisexual people are more likely to become 

targets of hate speech involving death or rape threats than their heterosexual 

contemporaries (ibid, p.957).    

 

 

 2.5. Other categorisations of hate speech  

 

 Of course, it is necessary to underline that contextual categorisation presented above is 

not absolute. As will be demonstrated in the case law analysis later, quite often these 

categories intermingle and hate speech instance may be attributed to more than one of 

them.  

 

An alternative typology of hate speech can be based not on the content of speech itself, 

but on the way it is disseminated. The first category of hate speech then would be “face-

to-face encounters”, the second – generally circulated hate speech (Gelber, McNamara, 

2016, p.325). Whereas the distinction between these two types of hate speech is not 

always obvious, regulation of the second is much more controversial than the latest (ibid, 



 23 

p. 325). Yet, based on the interviews with people, targeted by hate speech, scholars 

Gelber and McNamara concluded that these two mentioned types were not experienced 

different by victims, whether by “seriousness” or by “harmfulness” (ibid, p. 336).  

 

Finally, yet another completely different categorisation of hate speech was given by Yong 

who distinguished four different categories (2011, p.386):  

 

1) “Targeted vilification”, or speech that has a main intention to intimidate, to wound or 

to insult targeted individuals or groups (ibid, p.394). Targeted hate speech may occur 

either through face-to-face encounters or without any direct contact, but its main 

characteristic is that it is “narrowly directed” at the victims (ibid). This category of hate 

speech, according to the researcher, should not be covered by the principle of free speech 

as it is hard to imagine that expressing hostility or contempt can lead to personal 

development of the speaker or correspond to any other interests that are supported by the 

principle of free speech  (ibid, p.395);  

 

2)  “Diffuse vilification” that is covered by this principle but must not be protected (ibid, 

p.386, p.396). Yong understands under this category hate speech that attempts to 

intimidate, insult or wound, but is not directed at particular individuals or groups of 

people and is addressed to a wide audience of potentially sympathetic listeners – for 

example, speeches during Nazi marches (ibid). Yong argues that this kind of speech 

constitutes somewhat “greater degree of free speech interests” than targeted vilification, 

but contextually significance of such speech is usually low because the main aim of 



 24 

diffuse vilification is to intimidate, not to disseminate any ideas (ibid, p.397). In other 

words, it is a low-value speech and its regulation is necessary for prevention of 

intimidation (ibid, p.398);  

 

3)  “Political advocacy”, or speech that propagates exclusionary or eliminationist policies 

and that, too, argues Yong, is covered by the principle of free speech, but must not be 

protected (ibid, p.386). It must be emphasized that under “exclusionary policy” Yong 

understands “exclusion from full citizenship” of particular groups of people based on race 

or religion believes, and under “eliminationist policies” – “ethnic cleansing” or “forced 

repatriation” of these groups (ibid, p.398). As such political advocacy can result in 

enactment of these principles, Yong claims that this kind of hate speech, too, must be 

unprotected by free speech doctrine (ibid);  

 

4)   Other instances of speech that presents “adverse judgements” about a particular group 

of people based on their race or religion. According to the researcher, this category of 

speech must be protected by the principle of free speech even despite its potential harm 

(ibid, p.386, p.401). Explaining his position, Yong claims that this kind of speech is 

assertions or opinions and thus it can be answered through counterarguments (ibid).  

 

In other words, Yong suggests that speech that is either uncovered (1) or unprotected (2-

3) is “regulable” and may be regulated (ibid, p.388).  

 

 



 25 

2.6. Philosophical and practical reasons to ban hate speech  

 

Before providing an overview of legal base for hate speech regulation, it is necessary to 

determine why hate speech must be regulated. Whereas this issue seems to be 

straightforward – hate speech unlawfully threatens targeted people or groups of people 

and violets their dignity – scholars argue that the consequences of hate speech are much 

more diverse and long-term, and eventually can affect the whole society.  

 

Thus, Parekh emphasized that hate speech is harmful for public debate, as it negatively 

affects communities’ moral sensibility and distorts mutual respect in society (2012, p.44). 

Hate speech creates the environment of hostility and fear and intimidates the target group 

so that these people find themselves unable to participate in public life (ibid). Of course, 

simultaneously it affects personal lives of targeted people as they feel themselves 

humiliated and discriminated against (ibid).  

 

Similar position support Gelber and McNamara, who underline that hate speech can be 

regarded as “an existential attack” on a person’s dignity and its negative effects are long-

term and enduring (2016, p.325). Researchers argue that two types of harms of hate 

speech are usually distinguished:  

  1) constitutive, or harm resulted from saying hate speech per se, and  

  2) consequential, or harm that can be a result of hate speech (ibid).  For instance, hate 

speech can lead to negative stereotypes and create an environment where there 

stereotypes are normalized (ibid).  



 26 

 

While contacting interviews with the people in Australia, targeted by hate speech, Gelber 

and McNamara observed that they acknowledge being hurt, upset and frightened (ibid, 

p.328, p.333). In a number of cases interviewers reported that fear prevented them 

expressing themselves or to act against hate speech (ibid, p.333). Even silence and 

withdrawal from social life was mentioned by interviewers as a means to avoid hate 

speech (ibid, p.334), which eventually supports the view of Parekh regarding the 

exclusion of targeted groups from social life. Similar ideas were expressed by Davids 

who argues that racist hate speech results in exclusion (2018, p.298). The scholar 

emphasizes that peaceful coexistence of people in a society is not possible in a climate of 

hate speech (ibid, p.306). Finally, D'Souza, Griffin, et al. acknowledge that one specific 

type of hate speech – namely, gendered hate speech – indirectly contribute to 

perpetuating gender-based violence (2019, p.963).   

 

Yet not everyone agrees with the idea that hate speech must be banned. American scholar 

Baker, who defines himself as a supporter of “almost absolute protection” of freedom of 

speech (2012, p. 57), provides two arguments why prohibition of hate speech is not an 

ultimate solution of the problem. Firstly, he argues that prohibition will not effectively 

reduce potential harm of such speech (ibid, p.72). Secondly, he believes that such 

measures will, on the contrary, aggravate the existent problems in society without 

properly addressing them (ibid).  

 



 27 

Baker claims that direct response and open criticism of, for instance, racist views, can be 

more effective than prohibition of hate speech (ibid). Law regulation, on the other hand, 

can cause hate speech to “go underground” so that the real extent of the issue not be 

evident (ibid, p.73). Moreover, Baker believes that prohibition may enrage people of 

groups of people who support racist views, thus encouraging them to act (ibid). Conflicts 

within a society must be approached as political, not “violent struggle” (ibid, p.74). 

Politicians must address underling origins of racism and not hate speech that results from 

it (ibid).   

 

While Baker’s arguments evidently have some basis, one could retort that there is hate 

speech that cannot be answered in public debate, simply because it is motivated by pure 

hatred of troubled individuals and thus is not reasonable or logical arguments to be 

discussed in a dialogue.  Furthermore, as was discussed above, there are various types of 

hate speech and it seems that the existence in a society of xenophobia, religious 

disagreements, misogyny and other forms of tensions does not need proof in the form of 

hate speech. However, open criticism, mentioned by Baker, as well as general 

condemnation of hate speech by the society, seem to be essential in combating this 

hurtful phenomena, especially in the internet era, where hate speech resistance cannot 

completely depend upon law enforcement.     

 

Being an American scholar, Baker represents “permissive” American approach which 

will be discussed below and which differs from the European perspective that is the 

primary focus of the current thesis.    



 28 

 2.7. International legal base for freedom of expression and hate speech’ 

regulation  

 

Dealing with freedom of expression and hate speech and its regulation, it is necessary to 

establish the general European legal base of the question. Freedom of expression as a 

standard was first prescribed in 1948, in the Article 19 of the Universal Declaration of 

Human Rights. Two years later the European Convention on Human Rights, main 

normative act of the Council of Europe, guaranteed freedom of expression in the quoted 

above Article 10 and listed possible restrictions to it in the second part the same Article.  

 

In 1966, Article 19 of the International Covenant on Civil and Political Rights reiterated 

freedom of expression and everyone’s right to hold opinions without interference. At the 

same time, Article 20 of the Covenant proclaimed that “any advocacy of national, racial 

or religious hatred that constitutes incitement to discrimination, hostility or violence shall 

be prohibited by law” (International Covenant on Civil and Political Rights, 1966, Article 

19-20).  

 

Similar approach can be found in the Article 4 of the International Convention on the 

Elimination of All Forms of Racial Discrimination (1965) which disapprove propaganda 

and organizations that advocate for superiority based on ethnicity or skin colour, or 

justify discrimination or racial hatred (International Convention on the Elimination of All 

Forms of Racial Discrimination, Article 4). However, practical interpretation of these 



 29 

Conventions differs from state to state (OSCE, Preventing and responding to hate crimes, 

2009, p.54). 

 

Additionally, incitement to genocide that is also amount to hate speech (McGonagle, 

2013, p.6) is forbidden by the Article 3 of the Convention on the Prevention and 

Punishment of the Crime of Genocide which entered into force in 1951.  

 

 

2.8. Combating hate speech: minimal and extensive regulation approaches  

 

The widespread of hate speech has become a global issue due to the Internet and modern 

technologies. As mentioned in the Introduction, this issue, can be divided into two parts: 

hate speech’s impact on the society and individuals, its contribution to the hate crimes 

and violation of human rights that is causes, and, on the other hand, the possible 

consequences of the hate speech’s regulations to the freedom of expression. According to 

the former OSCE Representative on Freedom of the Media Miklós Haraszti, the general 

trend is to extend regulation of hate speech, rather than balancing it by “more reasoned 

speech” (Haraszti, 2012, p.xiii). The same view supports Bhikhu Parekh, who claims that 

exact forms of expression that are prohibited vary from country to country, but universal 

trend is observable even for the past few decades (Parekh, 2012, p.37).  

 

With regard to hate speech’s regulation, there are two possible approaches: the first one is 

a so-called “minimal regulation approach” and the second – “an extensive regulation”. 



 30 

The idea of the minimal regulation approach is based on the assumption that actual 

instigations to actual hate crimes must be criminalized, whereas simply offensive speech 

should be handled by dialogue in the press, courts, ethics organizations (Haraszti, 2012, 

xiii). Defenders of free expression often argue that, regardless of the disseminated 

content, freedom of speech will ultimately benefit the whole society (Downs, Cowan, 

2012, p.1354). Respectively, proponents of the extensive regulation approach propose 

new speech bans into national criminal codes (Haraszti, 2012, xiii).  

 

Already in 2012 Haraszti underlined that extensive regulation approach may put at risk 

“the very existence of an international human rights standards for handling hate speech” 

(ibid). The real consequences of such an approach in practice can be visible when 

analyzing the situation with freedom of speech in the post-soviet countries, where war on 

terror was twisted into “a fight against ”extremism”, a vague term that quickly “expanded 

to encompass almost all forms of political dissent” (Richter, 2012, p.290).  

 

It is necessary to underline the word “vague” in this quotation: for instance, vague and 

too broad definition of the term “extremism” in Russian and other post-soviet countries’ 

law has given the officials a possibility to “stifle debate in the media of issues of 

increasing public interest – such as the motivations underlying insurgency” (ibid, p.305). 

In other words, post-soviet authorities use anti-extremist laws to suppress and to control 

any possible political opponents, but they are not the only ones. Similar strategy is 

employed in other parts of the world, for example, in South Africa, where the same 

vagueness of the national laws against hate speech concerns scholars who believe it can 



 31 

be used against any unpopular, critical or politically unwanted speech (Davids, 2018, 

p.298).  

  

To summarize this view, it is a lack of clear and concrete definition of extremism, which 

was and is misused by the authoritarian governments for the restricting of freedom of 

expression. Taking into account the lack of a concrete definition of hate speech in 

international law, it is possible to assume that hate speech regulation may also be misused 

by other governments or political parties in their interests. This danger is even more 

evident if we consider that, as will be demonstrated below, social media to a certain 

extent replaced traditional courts while taking their own decision as to what information 

to permit and what to prohibit.  

 

Of course, opponents of this view believe that freedom of expression is used in defense of 

hate speech. For instance, Hate Speech International3 – an initiative that is dedicated to 

combating extremism and supporting journalists that cover this topic – estimates that 

whereas hate speech leads to political violence, “a liberal paradox” is created by 

supporters of freedom of expression who insist on its protection (Hate Speech 

International, Political backdrop). This view is also supported by Parekh who argues that 

freedom of speech protects arguments, ideas and opinions, whereas hate speech weakens 

democracy by spreading irrational fears and creating a sense of insecurity among its 

                                                 
3 Hate Speech International monitors hate speech, as well as hate crimes and extremism worldwide, thus 

being a very useful tool for researchers in this area.  

 



 32 

victims (2012, p. 48). Maitra and McGowan, too, underline that harm caused by hate 

speech excludes it from any protection under free speech principle (2010, p.364).  

 

Prominent communication scholar Peter Molnar attempted to find a compromise between 

the two views. According to Molnar, the debate between proponents of minimal and 

extensive regulation approach is based on lack of information and generalization that 

result in misunderstanding (2012, p. 183). Molnar argues that supporters of freedom of 

expression do not sacrifice everything for the sake of this freedom, but rather have 

different opinions as to what is the most effective response to hate speech (ibid, p.184).  

 

Molnar himself considers art and education “in the broadest sense”, as well as an 

application of the “imminent danger test” to be the best response to hate speech in truly 

democratic countries (ibid, p.185). This test is based not on the content, but rather on the 

context of hate speech and even on the social environment: if a certain expression 

constitutes “a clear and present danger of violent action”, it must be prohibited (ibid, p 

197). If, however, no imminent threat of violence can be assumed, prejudice and hatred 

are better to counteract with cultural activities, not law enforcement, claims Molnar 

(ibid).  

 

 

 

 

 



 33 

2.9. European and American approaches towards hate speech  

 

Some scholars believe that support of minimal and extensive regulation may be explained 

by historical background, providing as example different legal approaches in European 

countries and in the USA. In particular, the difference between European and American 

approaches is evident when it concerns racist hate speech (Bleich, 2014, p.284). 

 

Discussing this difference, Greene argues that European approach towards hate speech 

regulation is based on the idea that laws must protect minorities from democracy that 

went wrong (2012, p.92). On the contrary, American approach assumes that democracy 

required that all voices must be heard and self-expression should not be limited (ibid). As 

a result, there is “restrictive” European policy towards hate speech, and “permissive” 

American (ibid, p.94). According to the Supreme Court of the USA, American 

Constitution prevents adaptation of laws that are required by both Article 4 of the 

International Convention on the Elimination of All Forms of Racial Discrimination and 

Article 20 of the International Covenant on Civil and Political Rights, discussed earlier 

(ibid, p.95). In both cases the Conventions were ratified in the USA with some 

reservations.  

 

Abrams, too, notices that there is a constitutional protection of hate speech in the USA, 

which is ultimately based on the famous First Amendment (2012, p.116). Abrams argues 

that hate speech prohibition would be a restriction based on the content, which is strictly 

against the principles of the First Amendment (ibid, p.118). The same idea is supported 



 34 

by Bleich, who argues that in the USA it is almost impossible to convict someone of 

racist hate speech, unless this speech leads to “immediate violence” or is a direct threat ( 

2016, p.284).  

 

Explaining the difference between European and American approach, Abrams suggests 

that different historical background – European experience with Holocaust and Nazi 

versus American history – can lead to different decisions regarding hate speech 

regulation (2012, p.119). The same view shares Schauer who underlines that Holocaust 

denial is regarded unlawful in Austria, Germany, France and some other European and 

non-European countries, but not in the USA where such prohibition would contradict 

freedom of speech (2012, p.130). In France, however, laws against Holocaust denial, as 

well as laws against racist hate speech, are used as a legal base against expressed hostility 

(Suk, 2012, p.148).  

 

Some historical background of discrepancy between European and American approaches 

is given by Bleich who believes the two positions started to evolve differently in 1960th 

(2016, p.284). According to the researcher, after the European Court on Human Rights 

was established in 1959 to enforce the Convention, it obtained an ultimate authority in 

Europe for drawing the line between freedom of expression and hate speech (ibid). In the 

USA such power since 1920th belonged to the Supreme Court. However, Bleich stops 

here, noticing that there is no theoretically plausible explanation of the difference 

between European and American approaches – only various hypotheses (ibid, p.285).  

 



 35 

For example, Bleich mentions different “institutional strength of the courts”, variations in 

political cultures, divergent legal base for protection of freedom of expression, generally 

diverse jurisprudence systems and even contrasting interpretations of legal texts made by 

individual judges (ibid). At the same time Bleich emphasized that also European and 

American approaches towards freedom of expression and hate speech are “analytically 

distinct”, nevertheless they do not exclude each other (ibid, p.286).  

 

Interesting idea related to the American approach towards hate speech regulation is 

expressed by Downs and Cowan who argue that free-speech defenders in the USA 

acknowledge harm of certain speech, such as Holocaust denial or Nazi’s views, but yet 

consider freedom of expression to be more important (2012, p.1354). Downs and Cowan 

observe that it is easy to defend someone with whom you agree, but not so – to defend 

your opponent’s right to speak (ibid). This being said, researchers remind that hate speech 

can be harmful and humiliating to individuals, once again underlining the complex issue 

of finding a balance between freedom of speech and hate speech regulation (ibid).  

 

 

 2.10 . Online hate speech: general overview   

 

While providing unprecedented opportunities for communication and spread of 

knowledge, Internet facilitated “multi-directional” communication between its users 

(McGonagle, 2013, p.28). Lack of barriers that existed in traditional institutionalized 



 36 

media increased opportunities to participate in public discourse, but also simplified 

dissemination of hate speech (ibid).   

 

This observation is supported by the Council of Europe that called Internet “a turbo 

accelerator” of hate speech” (Council of Europe, 2013, p.1). According to Council of 

Europe, there are two principal reasons for this phenomenon: wide access, as well as 

anonymity which excludes responsibility that should accompany freedom of expression 

(ibid).  

 

One could argue that anonymity also has its advantages when it comes to political 

debates, but indeed, the Internet gave rise to additional instruments for dissemination of 

hate speech. Nowadays online hate speech can be found not only on the websites, but is 

more common in the social media where this problem has different perspective and could 

potentially lead to grave consequences due to real names and personal information, which 

people often provide about themselves.  

 

A special aspect of online hate speech is difficulty of combating it due to the speed of 

dissemination, as well as lack of training on the technical issues which sometimes prevent 

quick investigation and identification who is responsible for dissemination of hate speech 

(OSCE, Preventing and responding to hate crimes, 2009, p.56). McGonagle debates that 

other complication for online hate speech regulation is jurisdictional issue as different 

countries have different laws regarding hate speech which, in turn, influence policies of 

internet-providers (2013, p.30). The researcher claims that websites that disseminate hate 



 37 

speech are purposely being hosted in countries with “favorable jurisdiction” for hate 

speech (ibid, p.29). 

 

During the last decade various social media encountered hate speech on their portals and 

determined to (or were made to) address this issue in their guidelines and rules for users. 

Their policies, along with those of the internet-providers, are not unified as well as their 

automatic algorithms of content adjustment are not fully transparent. Being private actors, 

their involvement in online speech regulation could potentially lead to censorship, rightly 

warns McGonagle (2013, p.30). Yet the regulation of online hate speech is de-facto left to 

these private social media.  

 

Their related policies will be analyzed in detail in the third part of the thesis.  

 

 

2.11. Alternative ways to combat hate speech   

 

Not only Molnar, but other scholars and organizations believe that apart from minimal or 

extensive regulation discussed above, there are other, secondary activities that can 

contribute to combating hate speech. Possible list of these activities is given in the 

resource guide for NGOs “Preventing and responding to hate crimes” prepared by the 

OSCE’s Office for Democratic Institutions and Human Rights in 2009. OSCE 

acknowledges that there is a strong connection between hate crimes and hate speech and 

propose following ways to counteract the latest:  



 38 

 

- raise awareness about the problem;  

- monitor the hate content and advocate for the removal of hate speech from the Internet;    

- conduct education activities, more precisely – education of the problem of cyberhate for 

teachers, students, parents,  law enforcement;  

- denounce and challenge the arguments or claims, hold politicians and public figures 

accountable before public opinion;  

- and, finally, take legal actions if hate speech «crosses the threshold into crime»  

(OSCE, Preventing and responding to hate crimes, 2009, p.53-56).   

 

Similar idea about a need of alternative responses to hate speech was expressed in 

General Recommendation No. 35 of the UN Committee on the Elimination of Racial 

Discrimination, titled “Combating racist hate speech” (2013). In particular, this 

Recommendation underlines a necessity of educational, information and cultural 

strategies for combating hate speech (2013, p.9).   

 

In turn, Council of Europe, too, implements a number of strategies to combat hate speech. 

Among others, Council of Europe proposes to deny legal protection to hate speech, 

provide access to the means of expression to minorities and advocate for an intercultural 

dialogue (McGonagle, 2013, p.8).   

 

In must be noticed that measures proposed by OSCE, UN and Council of Europe should 

be regarded as a complex activity. One could assume, however, that these and other 



 39 

possible measures should be viewed as additional and cannot fully substitute law 

intervention because it reduces hate speech, sends a powerful message and prevents 

hostility against targeted individuals or groups of people (Parekh, 2012, p.46).  

 

Finally, in the academic literature on hate speech exists an opinion, that yet another way – 

or additional way – to combat hate speech is to provide a voice to the groups targeted 

with hate speech (Gelber, McNamara, 2016, p.326). Whereas this hardly can be regarded 

as the only needed measure, it could definitely influence the debate around hate speech, 

inform society’s perception of it and, eventually, provide some support to the victims.   

 

 

 

 

 

 

 

 

 

 

 

 

 

 



 40 

 3.  Research questions and methodology 

 

As was briefly outlined in the Introduction, the main research question of this thesis can 

be formulated as “What is understood under hate speech in a court’s practice and in the 

practice of social media: do these parties have equal approaches towards hate speech and 

if not, what differences can be observed?”  

 

Additionally, several sub questions were formulated that are supposed to deepen an 

understanding of hate speech phenomena and illuminate possible discrepancy between 

courts’ approach and that of the social media:  

1) What arguments does a court use to determine what is hate speech and what is 

offensive speech that nevertheless must be protected for the sake of freedom of 

expression?  

2) What types of hate speech can be found in a court’s practice? 

3) How do major social media define hate speech?  

4) How do major social media attempt to combat hate speech? 

5) Are these methods effective and sufficient or not? 

6) What alternatives ways to combat hate speech are already exist or are being 

developed?  

7) How does social media’s practice regarding hate speech correspond to that of a court?  

8) Do social media de facto replace a judge when controlling online hate speech? 

9) What implications it could potentially have for freedom of expression?  

 



 41 

As the scope of the thesis did not permit us to analyse case law of a number of courts, it 

was chosen to focus on the practice of only one court, namely, the European Court of 

Human Rights, established in 1959. The choice of this particular court is based on its 

significance for the European law practice and for the European Convention on Human 

Rights which norms it is supposed to protect, including freedom of expression, 

guaranteed by Article 10 of the Convention. The practice of the European Court of 

Human Rights related to hate speech is analyzed and compared with guidances and 

actions of the major social media, such as Facebook, Twitter, and YouTube that were 

selected due to their popularity and widespread in society’s life.  

 

Thus, according to the portal of web-statistic Statista.com, in April 2019 Facebook had an 

active monthly audience of 2 320 millions of users, YouTube – 1 900 millions, and 

Twitter – 330 millions (see: Statista, Most popular social media worldwide). Policies of a 

number of other popular social media were excluded from the research for various 

reasons. For instance, WhatsApp was excluded as a messenger, Instagram – as a portal, 

primary orientated on visual communication, Chinese social media platforms such as me 

WeChat, QQ, QZone and Sina Weibo were not analysed due to the linguistic 

inaccessibility.  

 

The main method used in the current thesis was qualitative content analysis that is based 

on text interpretation (Mayring, 2000, p.1). It must be noted that content analysis as a 

research method has a long history in social sciences. It was defined as a systematic 

technique that involves compressing texts into content categories already by Berelson in 



 42 

1952. Holsti in 1969 described content analysis as a research technique that “objectively 

and systematically” determines particular characteristics of a text (1969, p.14). 

Krippendorff already in 1980 established that content analysis involves collecting a set of 

texts, systematic reading of them, noticing consistent features and making assumptions 

about their meaning (p.3). These systematic inferences are supposed to answer the 

research questions of the study (ibid, p.25).  

 

Stemler, too, emphasized that content analysis relies on categorization of data and coding 

(2001, p.3), with Weber earlier explaining that under a category is understood a group 

with “similar meaning or connotations” (Weber, 1990, p.37).  

 

At the same time, qualitative approach aims at understanding of a particular phenomena 

(Vaismoradi, Turunen, 2013, p. 398). Qualitative content analysis suggests that material 

should be analyzed gradually and by dividing it into analytical units (Mayring, 2000,  

p.3). The same idea was expressed by Sparker who suggested dividing content into small 

units for a description (2005).  

 

Content analysis is applied by legal scholars, too, who conduct a content analysis of legal 

texts and judicial opinions for scientific understanding of the law as such (Hall &Wright, 

2008, p.64). In fact, one of the first content analysis of the legal practice was conducted 

by political scientist Fred Kort in 1957 who recorded and categorized facts from a 

number of judgments in an attempt to predict results of similar cases (ibid, p.67). Other 

scientists soon began to use content analysis of the judicial opinions in order to 



 43 

understand judicial behavior (ibid, p.68). In the 1990 content analysis of legal practice 

became widespread among legal scholars in the USA (ibid, p.70).  

 

Hall and Wright underline that content analysis can provide different information to the 

legal researches. For example, it may be useful for counting court results (ibid, p.85) or 

evaluating legal doctrine (ibid, p.87). They further theorize that exploration of legal 

doctrine consists of understanding of judicial reasoning (ibid, p.87), conducting 

normative legal analysis (ibid, p.88) or studying the landscape of case law (ibid, p.90). 

Although this thesis cannot be regarded as a legal study, in the second part of it we 

attempt content analysis of case law with a primary purpose to understand the reasoning 

of the European Court of Human Rights regarding hate speech and with a secondary 

purpose to compose an overview of recent practice of the Court regarding this issue.   

 

In other words, content analysis conducted in the current thesis was based on two primary 

sources:   

 

1. Selected case law of the European Court of Human Rights that was found in the online 

database of the court4. Selection of the judgments for the detailed analysis was based on 

time frame (2012-2019), as it is the same period when social media began to develop 

their own policies regarding hate speech, as well as on their significance and overall input 

to the understanding of the Court’s approach towards hate speech. Significance was 

established by comparison of various secondary literature where these cases where 

                                                 
4 The database of the European Court of Human Rights can be found here:  http://hudoc.echr.coe.int/  

http://hudoc.echr.coe.int/


 44 

repeatedly mentioned by legal scholars (among others: McGonagle, 2013; Bleich, 2013; 

Sottiaux, 2011; Tulkens, 2012; Buyse, 2014). .   

 

As categorization is a central part of the qualitative content analysis (Mayring, 2000,  

p.3), an attempt to categorize the judgments was made based on the theoretical 

background discussed in paragraph 2.4.  

 

2. Guidances, codes, community standards, official statements of the three social media 

platforms (Facebook, YouTube and Twitter) that were outlined and compared in the third 

part of the thesis. Additionally, online initiatives against hate speech were discussed in 

the paragraph 5.4  

 

Observations regarding the comparison were presented in the Conclusion section.   

 

 

 

 

 

 

 

 

 

 



 45 

4. European Court of Human Rights’ approach towards hate speech  

 

4.1. Definition and types of hate speech according to the European Court of 

Human Rights  

 

Although the limits of freedom of expression were already discussed in the famous 

Handyside v. the United Kingdom case in 1976, the term “hate speech” was not included 

in the Convention and first appeared in four judgments of the Court in 1999, namely in 

Sürek v. Turkey (No. 1, par. 62); Sürek & Özdemir v. Turkey (par.63); Sürek v. Turkey 

(No. 4, par. 60) and Erdogdu & Ince v. Turkey (par. 54) (McGonagle, 2013, p.11).  

 

Interestingly enough, in none of these four judgments does the Court provide a definition 

to hate speech. Instead, once again stressing responsibilities that come together with the 

right to freedom of expression, in three of four of these judgments the Court uses the 

same wording and warns media to be especially careful when publishing views of 

organizations that proclaim violence against the State as media should not become a 

mean “of dissemination of hate speech” (Erdogdu & Ince v. Turkey, 1999, par. 54).  

 

However in the fourth of these cases – Sürek v. Turkey, 1999 – the Court went further 

and decided that whereas some information may lawfully “offend, shock or disturb”, it 

the case Sürek v. Turke the commentary in question was “hate speech and glorification of 

violence” (see: Sürek v. Turke, 1999, par.62). Thus the Court established that there is a 

new type of speech – expression that goes beyond the protection given to the offensive 



 46 

speech protected in Handyside v. the United Kingdom decision (McGonagle, 2013, p.12). 

It is also worth noticing that in all these four cases the Court connected hate speech with 

“glorification of violence”.  

 

Nevertheless, European Court of Human Rights did not provide a definition to hate 

speech – neither in 1999 not later. Without a definition, case law of the Court is the main 

source which is explored by scholars who attempt to analyse the Court’s approach 

towards hate speech (McGonagle, 2013, p.12). While doing it, some of them underlined 

that it is important not to expand the term “hate speech” indefinitely (Tulkens, 2012, 

p.295).   

 

It should be noticed that European Court of Human Rights itself distinguishes a much 

broader number of hate speech categories than was discussed in the theoretical part of 

this thesis. In 2019 the Court published a Factsheet on hate speech where a significant 

number of hate speech’s types with examples of the respective most typical cases were 

listed (ECtHR, Factsheet hate speech, 2019, p.2-8):  

 

- Ethnic hate  (Pavel Ivanov v. Russia, 2007);  

- Incitement to violence, support of terrorist activity (Roj TV A/S v. Denmark, 2018);  

- Negationism and revisionism (mentioned earlier Garaudy v. France, 2003, also M’Bala 

M’Bala v. France, 2015);  

- Racial hate (Glimmerveen and Haqenbeek v. the Netherlands, 1979);  

- Religious hate speech (Norwood v. the United Kingdom, 2004); 



 47 

- Speech that constitutes “Threat to the democratic order” (among others - Schimanek v. 

Austria, 2000); 

- Speech that constitutes “Apology of violence and incitement to hostility” (discussed 

above Sürek (no.1) v. Turkey, 1999, also Gündüz v. Turkey, 2003);  

- Hate speech in “Circulating homophobic leaflets” (mentioned above Vejdeland v. 

Others v. Sweden, 2012);  

- Speech that condones terrorism (Leroy v. France, 2008, also Stomakhin v. Russia, 

2018);  

- Speech that condones war crimes (Lehideux and Isorni v. France, 1998);  

- Speech that denigrates national identity (Dink v. Turkey, 2010);  

- Speech that amounts to extremism (Ibragim Ibragimov and Others v. Russia, 2018);  

- Displaying symbolic with “controversial historical connotation” (Fáber v. Hungary, 

2012);  

- Speech that incites ethnic hatred (Balsytė-Lideikienė v. Lithuania, 2008);  

- Speech that incites national hatred (Hösl-Daum and Others v. Poland, 2014);  

- Speech that incites racial discrimination or hatred (among others: Féret v. Belgium, 

2009, Le Pen v. France, 2010, Perinçek v. Switzerland, 2015, Šimunić v. Croatia, 2019);  

- Speech that incites religious intolerance (Erbakan v. Turkey, 2006); 

- Speech that insults states officials (Stern Taulats and Roura Capellera v. Spain, 2018);  

- Hate speech that is primary disseminated in the Internet (among others: Delfi AS v. 

Estonia, 2015, Pihl v. Sweden, 2017, Nix v. Germany, 2018, etc.).  

 



 48 

Whereas this typology seems to be overwhelmingly detailed and not completely clear (for 

instance, the difference between ethnic hate and speech that incites ethnic hatred seems to 

be very blurred, as well as the difference between religious hate speech and speech that 

incites religious intolerance), it provides a full typological overview of the cases of the 

European Court of Human Rights that concern with information that the Court itself 

qualifies as hate speech. One could argue that it underlines the idea that it is not possible 

to provide a single finite definition to hate speech which instead should be better regarded 

as an open spectrum of speeches.  

 

 

4.2. ECtHR’s algorithm of establishing hate speech 

 

European Court of Human Rights has repeatedly reiterated that there is an obligation to 

draw a line between expressing views (including such that can “offend, shock or disturb”) 

and distributing information that represents incitement to extremism or any other form of 

hate speech (ECtHR, Factsheet hate speech, 2019, p.1). The Court proposes achieving 

this by applying two principal legal norms: Articles 10 and 17 of the European 

Convention on Human Rights. Article 17 – “Prohibition of abuse of rights” is to be 

applied when commentary in question attempt to destroy the fundamental values of the 

Convention, Article 10 – in all the other cases (ECtHR, Factsheet hate speech, 2019, p.1). 

For example, the limitations of the Article 10 must be applied when the commentary in 

question threatens public safety, health or the reputation or rights of others, etc. (ibid).   

 



 49 

Nevertheless, this algorithm remains too vague to solve the collision of freedom of 

expression and hate speech in practice. Quite often it is not so easy to determine, for 

instance, whether a message represents someone’s views on national security or territorial 

integrity or contains approval of terrorist’s attacks. Is an article an example of journalist’s 

provocation or the violation of the rights of others, do we need to protect moral in this or 

that case or do we need to protect the information, which shocks and disturbs, etc.  

 

As was briefly mentioned in paragraph 2.1 of this thesis, the Court itself developed a 

classic test that is supposed to determine whether Article 10 was violated and that is 

codified in Article 10.2 of the Convention. Firstly, any interference with freedom of 

expression must be prescribed by law (European Convention on Human Rights, 1950, 

Art.10.2). Secondly, any restrictions of freedom of expression must have a “legitimate 

aim” – among others, prevent disorder and crimes or protect the reputation or rights of 

other people. Thirdly, these restrictions must be “necessary in a democratic society” 

(ibid). As emphasized by McGonagle, within these rules states-members of the Council 

of Europe may regulate freedom of expression themselves which explains why there is no 

consensus among different jurisdictions regarding what content, how and how much must 

be regulated (McGonagle, 2013, p.9).  

 

It must be emphasized that the European Court of Human Rights does not replace 

national authorities in its decisions, but rather reviews their judgments for 

correspondence with the Convention’s values (ibid). In other words, the Court’s mission 

is to review national interpretation of the Convention (ibid).  



 50 

4.3. Overview of ECtHR’s hate speech practice in 1976-2012 

 

Whereas the scope of this thesis does not allow us to conduct a thorough full-scale study 

of all the judgments related to hate speech regulation in the 60-year old history of the 

European Court of Human Rights, analysis of the recent case law requires at least a brief 

overview of selected historical and most influential judgments that formed modern 

Court’s approach towards hate speech regulation. The general line is that the Court seeks 

harmonization of freedom of expression with protection of other human rights which 

involves protection from hate propaganda (Sottiaux, 2011, p.42).  

 

The famous case Handyside v. United Kingdom, discussed above, took place in 1976 and 

is usually debated and quoted in any scholarly publication as an earliest example of the 

Court’s approach towards hate speech regulation. In fact, the Court itself quotes 

Handyside v. United Kingdom in its Factsheet and in countless hate speech judgments 

(ECtHR, Factsheet hate speech, 2019, p.1). Still in the early period of its existence the 

Court ruled in the case Glimmerveen and Hagenbeek v. the Netherlands, 1979, that 

distribution of leaflets by far right political leaders in which they appealed to “our white 

people” and claimed to remove all foreign workers, constituted a form of racial 

discrimination (Glimmerveen and Hagenbeek v. the Netherlands, 1979, par.4). Bleich 

noted that Glimmerveen and Hagenbeek asserted that their leaflets were protected by 

freedom of speech provisions of the Convention, but the Court quoted Article 17 

discussed above and dismissed their claims (2013, p.292).  

 



 51 

Another case that happened before the Court introduced the term “hate speech” in 1999, 

was Jersild v. Denmark, 1994, in which a Danish journalist Jersild was convicted for 

spreading racist statements in the interview with right-wing group that he had conducted. 

This case is significant because it established a need for a journalistic responsibility when 

dealing with extremists views: the journalist in this case was convicted mainly because he 

failed to contradict his interviewees or distance himself from their racist statements 

(McGonagle, 2013, p.13).   

 

The analysis of the scholarly literature indicates that the number of hate speech cases 

appears to increase in the 2000th. After the four judgments in 1999 that introduced the 

term “hate speech” in connection with “glorification of violence” (see par. 4.1), a series 

of important rulings took place one by one. The judgment in Garaudy v. France, 2003, 

established Holocaust denial as unlawful and incitement to hatred against Jews, although 

in one of the previous cases the Court ruled that generally Article 10 of the Convention 

protects debate about unsettled historical events and public figures about whom different 

opinions may exist (Lehideux and Isorni v. France, 1998, par. 45).  

 

The case Gündüz v. Turkey, 2003, introduced debate about religious hate speech in the 

practice of the European Court of Human Rights. In this case the Court overturned 

Turkish courts’ decision that discussions on Islam and secularism constituted incitement 

to hatred based on religion. Instead, European Court of Human Rights ruled that 

defending religious beliefs without incitement to violence does not amount to hate speech 



 52 

(Gündüz v. Turkey, 2003, par.51). In other words, the Court underlined incitement to 

violence as a necessary component of hate speech.  

 

The Court reiterated its position a year later in the case Norwood v. United Kingdom, 

2004, which concerned a member of a British far right party who displayed out of a 

window of his apartment a poster with Twin Towers if flames, accompanying by the call 

to eliminate Islam from Britain in order to protect its people (Norwood v. United 

Kingdom, 2004, par. 2). The Court decided that this fact amounted to a public attack on 

Muslims in the United Kingdom and so a conviction of Norwood by British courts did 

not constitute a breach of the Article 10 (ibid, par.4). It is worth noticing that the term 

“hate speech” did not appear in this decision even five years after its introduction in the 

Court’s practice.   

 

According to Tulkens, the essence of today’s position of the Court was expressed in the 

case Erbakan v. Turkey, 2006 (2012, p.279). In this case the Court evaluated decisions of 

Turkish courts that convicted a future Turkish Prime Minister Erbakan for inciting hatred 

in his political speech in which he discussed differences between religions and races. The 

European Court of Human Rights concluded that, as speech in question did not create 

imminent danger, the penalty imposed by the national courts would have negative 

consequences for the political debate in the country. Thus, the Court underlined negative 

consequences as important factor in establishing hate speech. Simultaneously, the Court 

reiterated that in democratic society it may be necessary to prevent dissemination of 



 53 

expressions that incite or justify hatred on the grounds of intolerance (Erbakan v. Turkey, 

par. 56).  

 

Another significant case related to ethnic hate speech is Pavel Ivanov v. Russia, 2007, in 

which the Court convicted Russian journalist Ivanov in spreading hatred via his 

newspaper. In this case the speech in question was series of anti-Semitic publications in 

which Ivanov accused the entire ethnic group of Jews in all Russian problems, and called 

for their exclusion from social life (Pavel Ivanov v. Russia, 2007, The Facts). The Court 

had no doubts that this sort of speech constituted a violation of the underlying values of 

the Convention (ibid, The Law, par.1). However, in this judgment the Court did not use 

the term “hate speech” but referred to Ivanov’s publications as “incitement of ethnic, 

racial and religious hatred” (ibid, the Facts). This underlines the fact that the use of the 

term “hate speech” is not unanimous, and in 2007 the judges still preferred to use more 

general terminology.  

 

Incitement to racial hatred was further explored in the judgements  Soulas and Others v. 

France, 2008 and  Féret v. Belgium, 2009. In both these cases the speech – a published 

book and leaflets respectively – was directed against immigrants and incited hatred and 

violence against them. In the first case, the author of the book in question went so far as 

to propagate an “ethnic war” (Soulas and Others v. France, The Facts). In the second, a 

chairman of a political party Féret called to stop “Islamification of Belgium” (Féret v. 

Belgium, 2009, par.9). In both cases the European Court of Human Rights supported the 



 54 

national courts and reiterated that conviction of the applicants was in the interest of order 

in the society.  

 

Finally, two more case that must be mentioned in this very brief overview of the 

European Court of Human Rights’ practice in the period 1976-2012 and related to hate 

speech, are the cases Le Pen v. France, 2010 and Dink v. Turkey, 2010. The first of them 

is usually quoted together with the Féret v. Belgium, 2009, because of the similar nature. 

Thus, the applicant in this case was a prominent French politician Le Pen whom French 

courts found guilty in incitement of hatred and discrimination against Muslims in France 

for his far-right statements in an interview. The European Court of Human Rights upheld 

the decisions of the national courts, confirming that political comments that presented a 

whole religious community as a national threat are likely to result in “rejection and 

hostility” against these people (Le Pen v. France, 2010, The Law, par.1).  

 

On the contrary, the case Dink v. Turkey, 2010 represents a situation where the state – 

Turkey – condemned a journalist Firat Dink for his articles on the identity of Turkish 

people of Armenian origin. A year later, the journalist was killed. The relatives brought 

the case to the European Court which concluded that there was no “pressing social need” 

to find the journalist guilty by national courts because the articles did not incite violence 

or revolt and were not disrespectful or insulting (Dink v. Turkey, 2010, par.136). The 

journalist was merely expressing his opinions on the issue of public interest (ibid, 

par.135). It must be noted that additionally the Court highlighted the fact that authorities 

failed to protect the journalist (ibid, par.91).  



 55 

To sum it up, the argumentation of the Court in the discussed judgments is presented in 

the following Table 1:   

 

 

Table 1. Argumentation of the ECtHR in selected hate speech-related cases 1976-2012  

 

Judgement  Argumentation in support of the 

utterance in question being hate 

speech  

Argumentation in support of 

the utterance in question being 

protected speech (not hate 

speech) 

1. Handyside v. 

United Kingdom, 

1976 

 Ideas that can “offend, shock 

or disturb” must be free in the 

interests of “pluralism, 

tolerance and 

broadmindedness” (par.49) 

2. Glimmerveen 

and Hagenbeek 

v. the 

Netherlands, 

1979 

Political leaflets with appeal to 

remove all foreign workers, 

constituted a form of racial 

discrimination (par.4) 

 

3. Jersild v. 

Denmark, 1994 

Journalist convicted because he 

failed to contradict his 

interviewees or distance himself 

from their racist statements 

 

4. Lehideux and 

Isorni v. France, 

1988 

 Generally, the Convention 

protects debate about unsettled 

historical events and public 

figures about whom different 

opinions may exist (par.45) 

5-8. Sürek & 

Özdemir v. 

Turkey, 1999, 

Sürek v. Turkey, 

1999 (No. 4), 

Erdogdu & Ince 

v. Turkey, 1999 

Sürek v. Turkey, 

1999 

First use of the term “hate 

speech” by ECtHR, no 

definition is given, but 

responsibilities are highlighted: 

media should not become a 

mean “of dissemination of hate 

speech” (Erdogdu & Ince v. 

Turkey, 1999, par. 54). 

 

The Court established a 

connection between hate speech 

and “glorification of violence” 

(Sürek v. Turke, 1999, par.62) 

 

 



 56 

9. Garaudy v. 

France, 2003 

Holocaust denial is unlawful 

and constitutes incitement to 

hatred against Jews (par.45) 

 

9. Gündüz v. 

Turkey, 2003 

 Defending religious believes 

without incitement to violence 

does not amount to hate 

speech (par.51) 

10. Norwood v. 

United Kingdom, 

2004 

Call to eliminate Islam from 

Britain amounted to a public 

attack on Muslims in the United 

Kingdom (par.4) 

 

11. Erbakan v. 

Turkey, 2006 

 Because speech in question did 

not create imminent danger, 

the penalty imposed by the 

national courts would have 

negative consequences for the 

political debate (par.56) 

12. Pavel Ivanov 

v. Russia, 2007 

Accusation of the entire ethnic 

group and call for Jews to be 

excluded from social life 

constituted a violation of the 

underlying values of the 

Convention and incitement of 

ethnic hatred (The Law, par.1) 

 

13. Soulas and 

Others v. France, 

2008 

Propaganda of an “ethnic war” 

is incitement to hatred  

 

14. Féret v. 

Belgium, 2009 

Distributing leaflets with appeal 

to stop “Islamification of 

Belgium” is incitement to hatred 

(par.9) 

 

15. Le Pen v. 

France, 2010.   

 

Political comments that present 

a whole religious community as 

a national threat are likely to 

result in “rejection and 

hostility” against these people 

and therefore, conviction of the 

speaker is relevant and 

sufficient (The Law, par.1) 

 

16. Dink v. 

Turkey, 2010 

 There was no “pressing social 

need” to find the journalist 

guilty by national courts 

because the articles did not 

incite violence or revolt, were 

not disrespectful or insulting 

and the journalist was 



 57 

expressing his opinions on the 

issue of public interest 

(par.135-136) 

 

 

In the next paragraph more detailed analysis of the arguments provided in the judgments 

passed by the European Court of Human Rights in 2012-2019 – already after the 

widespread of social media – will be provided.  

 

 

4.4. ECtHR’s case law concerning hate speech in 2012-2019 

 

Despite the typology found in the Factsheet hate speech of the European Court of Human 

Rights and discussed above, we propose to categories the analysis by content of the cases, 

as it was done by Tulkens who distinguished only four categories of hate speech: 

glorification of violence, religious intolerance, racial hate speech and sexual orientation 

hate speech (2012).  

 

In this paragraph following fifteen judgments of the European Court of Human Rights 

from the period 2012-2019 will be analyzed in details:  

 

- Political/racial//ethnic hate speech (Aksu v. Turkey, 2012;  Öner and Türk v. Turkey, 

2015; Balázs v. Hungary 2015; Perinçek v. Switzerland, 2015;  Király and Dömötör 

v. Hungary, 2017);  



 58 

- Religious hate speech (Raelien Suisse v Switzerland, 2012; Belkacem v. Belgium, 

2017); 

- Sexual orientation hate speech (Vejdeland and Others v. Sweden, 2012; Identoba and 

Others v. Georgia, 2015; M.C. and A.C. v. Romania, 2016);   

- Online hate speech (Delfi AS v. Estonia, 2015; Magyar Tartalomszolgáltatók 

Egyesülete and Index.hu Zrt v. Hungary, 2016; Pihl v. Sweden, 2017; Smajić v. Bosnia 

and Herzegovina, 2018; Savva Terentyev v. Russia, 2018 ). 

  

The results of this analysis will be presented in Table 2 in paragraph 4.5.  

 

 

4.4.1. Political, racial, ethnic hate speech  

 

Aksu v. Turkey, 2012 

 

As was mentioned in the chapter 4.1 of the present paper, the European Court of Human 

Rights distinguishes, among others, political, racial, national and anti-constitutional hate 

speech. However, as will be evident from the detailed analysis of the judgments, this 

separation is not always clear and in some cases, is very blurred. That is why, for the 

purpose of the present thesis, it was decided to combine the next five judgments in one 

category, namely, political, racial and ethnic hate speech.  

 



 59 

The first of the five judgments to be discussed in this category is the case of Aksu 

v. Turkey, 2012 where the alleged hate speech was contained in the book “The Gypsies 

of Turkey” written by the Associate Professor Ali Rafet Özkan and published by the 

Turkish Ministry of Culture. The book in question was a research about people of Roma 

origin in which the author describes the ethnic and cultural features of the respective 

nation. Among other things, the author claimed that people of Roma origin “were 

engaged in illegal activities” and lived as “thieves, pickpockets, swindlers, robbers, 

usurers, beggars, drug dealers, prostitutes and brothel keepers” (Aksu v. Turkey, 2012, 

par.14). These particular remarks attracted the attention of the applicant of Roma origin 

and he filled the complaint in 2002, claiming that the passages were insulting and 

humiliating. The national courts in long and complicated proceedings found no proof that 

such remarks constitute an attack on the applicant or the other people of Roma origin.  

 

In a separate but related case, the same applicant claimed that several entries in the new 

published Turkish dictionary contained discriminatory and insulting information about 

the Roma people. For instance, the applicant especially mentioned such entries as “Gypsy 

tent’: a dirty and poor place” and ‘Becoming a Gypsy’: displaying miserly behavior” 

(ibid, par.28). Several national courts dismissed the complaint on the ground that the 

entries reflect historical and sociological reality and were not meant to insult a particular 

ethnic group.  

 

The case came to the European Court of Human Rights that investigated the complaint 

and found a need for balance between the applicant’s rights and public interest in 



 60 

freedom of expression. Restating, that freedom of expression is one of the most essential 

foundations of a democracy, the Court reminded about the information that can “offend, 

shock or disturb” and that was protected in the case Handyside v. the United Kingdom, 

1976. The Court then proceeded with the reminder that freedom of expression may be  

subject to the exceptions described in the second part of the Article 10 of the European 

Convention on Human Rights, but the necessity of such restrictions must be established 

“convincingly” (ibid, par.64). 

 

Once again, the Court underlined that the analysis of the passages from the book in 

question must be conducted in the context of the whole book, including the research 

method used by the author who contacted the members of the Roma community, as well 

as the police and relevant authorities (ibid, par.20). Furthermore, the author claimed that 

he lived with the Roma community and observed their lifestyle from the scientific point 

of view (ibid, par.11).   

 

Additionally, the European Court stated that the applicant had an opportunity to review 

his case before two different levels of national jurisdiction and the Ministry of Culture 

windrowed the remaining copies of the book. Taking it all into account, the Court 

concluded that the necessary balance between the rights of the Roma community and 

freedom of expression was found, even considering the vulnerable position of the Roma 

people in the country. According to the court, the government should do more to battle 

the negative stereotypes about the Roma community, but in the present case the Turkish 

authorities did not misuse their power and did not violate anybody’s rights. The 



 61 

expressions used in the book are part of a spoken daily Turkish, as much as the 

dictionary’s entries from the second part of the case. The entries must have been labeled 

as “insulting”, but an absence of this note was not enough to constitute a violation.  

 

It is necessary to emphasize that once again the court considered the context around 

particular words and phrases, and mentioned both the case of Handyside v. the United 

Kingdom, 1976 and the European Convention on Human Rights while establishing a 

balance between offensive speech and freedom of expression.  

 

 

Öner and Türk v. Turkey, 2015 

 

Another case that was considered to be an example of alleged political and racial hate 

speech also originated in Turkey. The applicants in the Öner and Türk v. Turkey, 2015, 

attended public celebrations and made speeches about Kurdish people. Among others, the 

applicants talked about the civilians killed by the security forces in Cizre, Nusaybin and 

Şırnak, the poisoning of the Kurdish leader and claimed that the authorities did not 

supported democratization and did not solve the Kurdish problem (Öner and Türk 

v. Turkey, 2015, par.6).  

 

Shortly after the event, the public prosecutor charged the applicants with terrorist 

propaganda; the national court found them guilty and sentenced them to one year and 

eight month of imprisonment. Though the sentence was eventually suspended, the 



 62 

applicants filed a complaint to the European Court of Human Rights, claiming that their 

right for freedom of expression guaranteed to them by Article 10 of the Convention was 

violated. The applicants emphasized that in their speech they spoke about the necessity to 

find peaceful methods to solve the Kurdish problem, did not advocate any kind of 

violence and did not called upon people to commit any kind of illegal actions (ibid, 

par.19).  

 

After reviewing the case, the European Court concluded that there was an interference 

with the applicants’ rights to freedom of expression and it was not necessary in a 

democratic society. From the Court’s perspective, the speech in question constituted a 

critical assessment of the authorities and did not contained terrorist propaganda. 

Moreover, the Court specifically mentioned that the applicant’s speech did not 

“encourage violence, armed resistance or an uprising” (ibid, par.24) and did not incited 

violence by “by instilling a deep-seated and irrational hatred against identifiable persons” 

– and thus was not hate speech (ibid).  

 

Interestingly enough, this judgment is one of the very few instances where the European 

Court of Human Rights specifically mentioned the term “hate speech”, adding to it a 

significant characteristic. Thus, from this judgment it is possible to assume that the Court 

understands under hate speech a speech that is capable of inciting violence by appealing 

to the irrational hatred against a person or a group of people.  

 



 63 

Additionally, the case Öner and Türk v. Turkey, 2015 is a good example of blurring of 

the categories of political and ethnic hate speech or alleged hate speech as the initial 

speech of the applicants was about the Kurdish problem and their discrimination in the 

society, as well as about the authorities’ policy towards this issue.  

 

Balázs v. Hungary 2015 

 

The case Balázs v. Hungary 2015, can also be classified as a political or ethnic hate 

speech case. In it, the applicant of Roma origin and his girlfriend were leaving a club 

when three men approached and started to insult them with “Dirty gypsy, do you need a 

cigarette? Here is money!” (par.10). The situation developed in a fight that was 

concluded by the police arrival. Subsequently, the national courts failed to establish a link 

between the insult with national hatred and the fight, and the case came into the European 

Court.  

 

The Court concluded that the state has an obligation to establish any possible racist 

motive and ethnic hatred behind the attack. Even though proving racial motivation could 

be difficult, the national courts should have taken into account that people of Roma origin 

are especially vulnerable in the Hungarian society and require an additional protection 

(ibid, par.53).  

 

Moreover, the Court underlined that any racist verbal abuse is highly relevant in cases 

involving ethnic minorities (ibid, par.61). When only the victim’s characteristic is 



 64 

involved, the attack can be classified as hate crimes even if the motives were mixed (ibid, 

par.70). Therefore, concludes the court, the state clearly violated the rights of the 

applicant by not conducting a thorough investigation (ibid, par.75).  

 

While relevant for the present paper, it is necessary to mention that in the case Balázs 

v. Hungary 2015 the term “hate speech” was not used as such in the court’s assessment 

part of the judgment, but was substituted by the phrase “racist verbal abuse” which in the 

present case can be regarded as a synonym. Regarding the typology, the case Balázs 

v. Hungary 2015 provides an example of ethnic hate speech.   

 

 

Perinçek v. Switzerland, 2015 

 

Finally, the last selected case found for the category of political, racial and ethnic hate 

speech is the case Perinçek v. Switzerland, 2015. In this case the applicant was a doctor 

of law and a chairman of the Turkish Workers’ Party who in 2005 took part in a various 

press conferences where he publicly denied the genocide of the Armenian people by the 

Ottoman Empire in 1915. In one of his speeches, he claimed that “the allegations of the 

‘Armenian genocide’ are an international lie” and that the Kurdish and the Armenian 

problem “did not even exist” (Perinçek v. Switzerland, 2015, par.13).  

 

After the conferences the Switzerland-Armenia Association filed a complaint against the 

applicant and the Lausanne District Police Court found the applicant guilty of racial 



 65 

discrimination as the Armenian genocide was a proven fact (ibid, par.186). After the 

appeal in the national courts, the case came to the European Court of Human Rights 

where the applicant claimed that the courts breached his freedom of expression.  

 

In 2013 a Chamber of the Court found a violation of Article 10 of the Convention, but at 

the Government’s request the case came to the Grand Chamber of the Court which 

analyzed, among others, whether the speech in question can be regarded as hate speech 

by comparing it to the similar cases such as, among others, Gündüz v. Turkey, 2003  and 

Erbakan v. Turkey, 2006.  

 

After this analysis, the Court concluded that the national courts failed to establish a social 

need or necessity of the applicant’s conviction in a democratic society, as the authorities 

overstepped their power and suppressed a debate with huge public interest (ibid, par. 

241). The mere denial of Armenian genocide is insufficient to be qualified as an 

incitement of hatred towards a particular nation and, accordingly, the applicant had the 

right for open discussion of the issue even if it is sensitive and may be offensive, as it is a 

fundamental aspect of freedom of expression (ibid, The Court’s assesment). 

 

Once again it is necessary to emphasize that the Court did not only analyzed the 

particular words or speech, but regarded it in a broader context of the historical and 

political situation. Moreover, the Court focused on whether the speech in question had a 

public interest and whether a conviction of the speaker was necessary in a democratic 

society – two more crucial questions that can be used in establishing whether particular 



 66 

content constitutes hate speech or not. From this perspective, the difference between the 

case Király and Dömötör v. Hungary, 2017 analyzed below and the cases Perinçek 

v. Switzerland, 2015 and Öner and Türk v. Turkey, 2015 is especially evident.  

 

 

Király and Dömötör v. Hungary, 2017 

 

Finally, the fifth case to be analyzed in this category is Király and Dömötör v. Hungary 

that was passed in January 2017.  In this case the applicants – two Hungarians of Roma 

origin – complained that the state failed to protect them during the anti-Roma 

demonstration that has taken place in 2012. This demonstration was organized by the far-

right groups who described it on their websites as “against Roma criminality” (Király and 

Dömötör v. Hungary, 2017, par.7).  

 

During the events the leaders of the far-rights groups called Roma “not “normal”, and 

claimed that Roma criminality was “omnipresent” in Hungary and that this ethnic group 

brought “only destruction, devastation and fear” (ibid, par.10). On this ground, the 

leaders called the participants to “sweep out the “rubbish” from the country” and stated 

that the Hungarians had rights to use all possible means to achieve this purpose (ibid).  

 

The applicants who were in the neighborhood during the demonstration heard these and 

other speeches and later submitted a complaint to the national courts. According to the 

applicants, the police and eventually the state failed to protect them from the racist 



 67 

threats. The Hungarian courts, however, concluded that the police acted professionally 

and had no legal basis to stop the demonstration (ibid, par.15).  

 

The European Court of Human Rights did not question the actions of the Hungarian 

police, but concentrated mainly on the intimidating speeches that the applicants were 

exposed to due to their ethnicity. While the Hungarian courts did not investigate this part 

on the ground that the statements made during the march were not included in the initial 

complaint, the European Court of Human Rights reminded about the discrimination of 

Roma minority in the country and repeated examples of hate speech and even hate-

motivated killings (ibid, par.40).  

 

According to the European Court of Human Rights, harassment motivated by racism with 

no physical violence was still a crime and a violation of the applicants’ rights.  The Court 

especially underlined a tense political situation in the country (ibid, par.73) and 

acknowledged that in all similar cases the outcome depended on the various factors as the 

Court’s approach to the cases with an alleged hate speech was “highly context-specific” 

(ibid. par.74).  

 

In other words, the Court did not analyze particular phrases or words that were said 

during the far-right demonstration, but rather the whole situation of hatred and threat that 

was created. The context appears to be more significant for the Court than the precise 

utterances. In the case Király and Dömötör v. Hungary, 2017 the European Court 

concluded that national courts failed to investigate the context of the speeches and, by 



 68 

establishing that the statements were hateful and abusive, but did not exited any violence, 

eventually failed to protect the rights of the applicants.  

 

 

4.4.2. Religious hate speech 

 

Raelien Suisse v Switzerland, 2012 

 

Two of the selected recent judgments of the European Court of Human Rights could be 

classified as religious hate speech: Raelien Suisse v Switzerland, 2012 and Belkacem v. 

Belgium, 2017. In the first of these cases the applicant was an association of the Raelian 

Movement that was registered in 1977 with an aim to make first contacts with 

extraterrestrials and establishing good contacts with them (Raelien Suisse v Switzerland, 

2012, par.10). In 2001 the organization requested the authorization from the police to 

conduct a poster campaign during which a poster “The Message from Extraterrestrials” 

was supposed to be demonstrated (ibid, par.14).  It was expected that together with the 

contact details of the organization the phrase “Science at last replaces religion” would 

have been written on the poster (ibid).  

 

The authorities, however, refused to grant permission to the event, citing the previous 

national courts’ decisions that the movement was engaged in the activities against the 

public order (ibid, par.15). The national court once again stated that the Raelian 

Movement could not have a religious freedom because it was regarded as “a dangerous 



 69 

sect” as it has advocated, among other things, for the human cloning and “geniocracy” 

(ibid, par, 16).   

 

After national courts in Switzerland upholded this decision, the case went to the 

European Court that reiterated the importance of establishing the nature of the speech 

(ibid, par.61). It was clear that the poster wanted to draw attention of the public to the 

movement with a religious connotation and did not contribute to the political debate in 

Switzerland, thus being closer to the commercial advertisement than to the political 

speech, which gives authorities more freedom (ibid, par.62). After establishing that fact, 

the Court underlined that different states may have reasons to impose restrictions in such 

issues (ibid, par.63). In the present case the Court found a prohibition of the poster 

demonstration to be a reasonable restriction in order to protect health, morals and the 

rights of others in the society (ibid, par.72).  

 

Moreover, the Court found the position of the authorities to be in balance, as they did not 

banned the whole association or its website and even without the poster campaign it had 

other means to disseminate its ideas and express its beliefs (par.73). Accordingly, the 

decision of the police to deny the permission was relevant and proportional and answered 

the need of the society and, therefore, there was no violation of Article 10 (par.76).  

 

It is important to emphasize that the official judgment of the Court was complemented by 

the dissenting opinion of one of the judges who wrote, among others, that the speech in 

question – in particular the phrase “Science at last replaces religion” did not constitute 



 70 

religious hate speech and did not denigrate any religion (ibid, Scientific atheism).  The 

judge then mentioned the Article 10 of the Convention and his belief that it prohibits the 

state to play a role of a watchman and to prescribe what is right or wrong (ibid). 

Accordingly, the judge concluded that ban of the poster did not correspond to the social 

need in his opinion (ibid).  

 

While analyzing this case, it is necessary to notice that the Court once again examined the 

speech in a broader context, namely, the nature of the organization and its beliefs, as well 

as the purpose of the speech in question. That is similar to the Court’s approach in the 

cases with alleged political and national hate speech where it also took into account 

whether the speech was made with a purpose to incite hatred and intolerance, or in order 

to trigger public debate.  

 

Belkacem v. Belgium, 2017 

 

Another selected case – Belkacem v. Belgium, 2017 – represents a more typical example 

of religious hate speech than the previous case. In this case the applicant was convicted 

by Belgian courts for incitement to hatred in his YouTube videos where he called on his 

audience, among other things, to fight with non-Muslims and “teach them a lesson” 

(Belkacem v. Belgium, 2017, par.8). In the European Court the applicant claimed to 

merely express his opinion and exercise his freedom of expression.  

 



 71 

The Court, however, unanimously and “without any doubts” concluded that the applicant 

disseminated hateful content and sought to incite violence against all non-Muslims which 

is clearly against the values of the Convention (ibid, par.33). Additionally, the Court 

highlighted that defending Sharia while advocating for violence can amount to hate 

speech and that states have an obligation to contradict “religious fundamentalism” (ibid, 

par.34). 

 

 

4.4.3. Hate speech based on sexual orientation 

 

Vejdeland and Others v. Sweden, 2012 

 

As was briefly mentioned in the theoretical part of this thesis, the case Vejdeland and 

Others v. Sweden, 2012 was the case where the European Court of Human Rights 

recognised homophobic hate speech for the first time (McGonagle, 2013, p.12). The 

applicants in this case distributed leaflets in a school in which they claimed that 

homosexuality was a “deviant sexual proclivity” responsible for HIV and AIDS and 

destructive for a society (Vejdeland and Others v. Sweden, 2012, par.8). In their defence 

the applicants claimed that they merely wanted to start a debate (ibid, par.10).  

 

The European Court, however, concluded that even of these statements did not propagate 

hateful acts as such, they were still serious allegations that reinforced prejudices (ibid, 

par.54). The Court reiterated that discrimination based on sexual orientation should be 



 72 

considered as serious as discrimination on the ground of race, sex, colour or origin (ibid, 

par.42). Therefore, national courts did not violate Article 10 by convincing the applicant.   

 

 

Identoba and Others v. Georgia, 2015 

 

Another case of the European Court that concerns sexual orientation hate speech is 

Identoba and Others v. Georgia, 2015. In this case the complaint was filed by 14 people 

and a non-governmental organization set up to promote and protect the rights of LGBT 

people. According to the applicants, in May 2012 during a peaceful demonstration for the 

International day against Homophobia they were insulted and threatened by the counter-

demonstration of members of a religious group (Identoba and Others v. Georgia, 2015, 

par.10). In particular, the applicants were accused of being “seek”, “immoral” and 

“perverts” and on this ground the participants of the religious demonstration stated that 

they “should be burnt to death” and “crushed” (ibid, par.13 and par.15). The threats were 

followed by the attacks and some of the applicants suffered physical trauma (ibid, par.18)  

 

National authorities failed to investigate the case properly and the case came to the 

European Court which concluded that the applicants clearly became the target of hate 

speech and aggressive actions (ibid, par.70). The whole situation contained an intense 

anxiety, with a homophobic bias as an additional factor (ibid).  

 



 73 

The domestic authorities, however, failed to investigate the homophobic motives of the 

event and did not determine homophobic hate speech (par.77). In its conclusion, the 

Court once again stated that the applicants were left in anxiety and fear by violence which 

“consisted mostly of hate speech and serious threats” (ibid, par.70) and thus there was a 

clear violation of the applicants’ rights.  

  

M.C. and A.C. v. Romania, 2016 

 

In a similar case - M.C. and A.C. v. Romania, 2016 - hate speech again was addressed 

towards the participants of the LGBT-rally, but this time in Bucharest. After the event, 

the applicants were physically attacked by a group of six young men and a woman who 

were shouting “You poofs go to the Netherlands!” (M.C. and A.C. v. Romania, 2016, 

par.9).  

 

Similar to the previous case, Romanian authorities failed to protect the rights of the 

applicants.  The European court of human rights found that the authorities did not 

investigate homophobic motives behind the attack especially giving that the speech in 

question was “clearly homophobic” hate speech (ibid, par.124).   

 

In other words, in both Identoba and Others v. Georgia, 2015 and M.C. and A.C. 

v. Romania, 2016 the European Court had little doubt about the nature of the speech and 

concluded it to be sexual orientation hate speech on the ground of the context, on one 

side, and taking into account the whole situation and the following attacks, on the other.  



 74 

 

4.4.4. Online hate speech cases in the practice of the ECtHR 

 

Delfi AS v. Estonia, 2015 

 

Five of the selected cases could be classified as online hate speech.  Chronologically first 

of them – Delfi AS v. Estonia, 2015. The applicant of this case was a company that 

owned the largest news portal in Estonia which publishes up to 330 news per day both in 

Estonian and Russian (Delfi AS v. Estonia, 2015, par.11). In 2006 one of the articles 

published by the news portal attracted the attention of the readers and consequently 

received 185 comments (ibid, par.17). Approximately twenty of them contained personal 

threats and offensive language against a person mentioned in the article (ibid).  

 

For instance, among the comments were “bloody shitheads”, “go and drown yourself”, “a 

good man lives a long time, a shitty man a day or two”, etc. (ibid, par.18). The lawyers of 

the concerned person filled a lawsuit and at the same day the news portal deleted the 

offensive comments (ibid, par.19). The national court, however, decided that as the news 

portal reserved the right to delete inappropriate comments, it has taken insufficient 

measures to protect the rights of others (ibid, par.26). Moreover, the national court of 

appeal reiterated that the news portal was not a technical intermediary, but a “provider of 

content services”and therefore had respective obligations (ibid, par.29).   

 



 75 

Disagreed with the decisions of the national juridical system, the Delfi Company brought 

the case to the European Court of Human Rights that eventually shared the position of the 

national courts. Thus, the European Court emphasized that, whereas it acknowledges how 

beneficial Internet can be for freedom of expression, it must be taken into account that all 

kinds of unlawful speech, including hate speech, can be easily disseminated online (ibid, 

par.110).  Therefore, it is important to find a balance between freedom of expression, on 

one the hand, and prohibiting unlawful speech that constitutes a violation of personality 

rights and other underlying values of the Convention (ibid).  

 

According to the Court, Internet news portals have “duties and responsibilities” when 

their users disseminate unlawful speech through their sites (ibid, par.113). Taking into 

account that the applicant owned one of the biggest news portals in Estonia, the nature of 

some comments on the portal was known before the present case (ibid, par.117) As was 

established by the Supreme Court of Estonia and reiterated by the European court, the 

comments in question constitute hate speech and speech that “directly advocated 

violence” (ibid). That is why the automatic filter-system used by the portal to filter the 

instances of hate speech was insufficient and failed to prevent the violation of the rights 

of others (ibid, par.156).   

 

Based on these considerations and taking into account the nature of the comments, the 

Court upheld the decision of the national courts and concluded that there was no violation 

of Article 10 of the Convention in convicting the news portal.  

 



 76 

Magyar Tartalomszolgáltatók Egyesülete and Index.hu Zrt v. Hungary, 2016 

 

The second case concerned with online hate speech is Magyar Tartalomszolgáltatók 

Egyesülete and Index.hu Zrt v. Hungary, 2016. The case involves a self-regulatory body 

of Hungarian Internet content providers and a major news portal of Hungary who were 

the applicants of the case.  

 

In 2010, when both applicants allowed users’ comments of their portals without 

premoderation, the applicants published an article that illustrated their opinion about two 

real estate websites (Magyar Tartalomszolgáltatók Egyesülete and Index.hu Zrt v. 

Hungary, 2016, par.11).  The article attracted the attention of the audience and among the 

comments were phrases like “They have talked about these two rubbish real estate 

websites a thousand times already” and “Is this not that Benkő-Sándor-sort-of sly, 

rubbish, mug company again?” (ibid, par.12).  

 

The company-owner of the websites filled a lawsuit, accusing the applicants that their 

opinion and the users’ comments were false and had a negative influence on their 

reputation (ibid, par.15). The applicants removed users’ comments, but the national 

courts still found them guilty in infringing the websites’ company-owner’ reputation 

(ibid, par.17).  

 

When the case came to the European Court of Human Rights, it compared it to the case 

Delfi AS v. Estonia, 2015. However, the court concluded the Hungarian case was 



 77 

different: after the analysis of the speech, the court found that the users’ comments were 

“offensive and vulgar”, but not unlawful and did not amount to hate speech (ibid, par.64).  

 

Once again, the court underlined the necessity to find a balance between the rights of all 

concerned parties. If the users’ comments constitute hate speech, the state may impose 

liability upon news portals. But the present case did not involve such context and 

therefore the court concluded that there had been a violation of the rights of the applicants 

(ibid, par.91).  

 

 

Pihl v. Sweden, 2017  

 

In a similar case, the applicant was a subject of an online defamation, namely, of an 

accusation of being a member if a Nazi party (Pihl v. Sweden, 2017, par.3). The applicant 

replied online that this information was wrong and the administration of the blog in 

question deleted the previous entry (ibid, par.6). Still, the applicant claimed that national 

courts refused to hold the administration of the web-site liable and did not protect his 

reputation (ibid, par.15). 

 

The European Court, however, declared that in this case a balance between freedom of 

expression and applicant’s rights was found by national courts as the defamatory post was 

quickly removed and an apology issued (ibid, par.29 and 30). Additionally, the Court 



 78 

highlighted that although offensive, the post did not incite violence and did not amount to 

hate speech against the applicant (ibid, par.37).   

 

 

Smajić v. Bosnia and Herzegovina, 2018  

 

The applicant of this case was convicted for posting online statements in which he 

described military actions that might be undertaken by Bosnians against Serb villages in 

the event of a new war and expressed insults against Serbs (Smajić v. Bosnia and 

Herzegovina, 2018, par.5). The applicant claimed to discuss a matter of public interest 

(ibid, par.29).  

 

The European Court unanimously rejected these allegations, stating that conviction of the 

applicant was prescribed by law for the protection of the rights of others (ibid, par.32). 

Furthermore, the Court underlined that, while hypothetical, the applicant’s statement 

related to a very sensitive subject in Bosnian society and so national courts had sufficient 

reasons to convict him (ibid, par.39).  

 

 

Savva Terentyev v. Russia, 2018  

 

The case Savva Terentyev v. Russia, 2018 can probably be regarded as a classic example 

of “Handyside v. United Kingdom rule” which states that certain expressions may 



 79 

“offend, shock or disturb”, but must, nevertheless, remain free in the interests of 

“pluralism, tolerance and broadmindedness” (Handyside v. United Kingdom, 1976, 

par.49). In this case an applicant posted remarks about Russian police, calling them 

“faithful dogs” and making other insulting remarks about them  (Savva Terentyev v. 

Russia, 2018, par.10).   

 

While acknowledging that the applicant’s comments were expressed in an offensive 

form, the Court unanimously concluded that they did not incite hatred (ibid, par.84). The 

comments in question were “an emotional reaction” of the applicant to the abusive 

actions of the police and did not provoke any violence (ibid). Therefore, the Court stated 

the criminal conviction of the applicant for a sentence of one year’s imprisonment was 

disproportionate and not justified by “pressing social need” (ibid, par.86).  

 

 

4.5. Summary of the ECtHR’s practice concerning hate speech 

 

To summarize the analysis of the case law of the European Court, it is necessary to 

emphasize certain criteria that the Court employs in order to determine whether the 

speech in question is hate speech or not and what judgment to pass. Those criteria are:  

 

1) Nature of the speech: whether it is capable of inciting violence or hatred or not;  

2) Purpose of the speech: whether it is aiming at inciting hatred or not;  



 80 

3) Significance of the speech: whether it has public interest and contributes to the public 

debate or not;  

3) Context of the speech: whether there are tensions that accompany the speech in 

question or not;   

4) Necessity of convincing the speech and the speaker in the democratic society;   

5) Whether a balance between freedom of expression and rights of others was taken into 

account by the national courts, or not.  

Further arguments of the European Court from the analyzed cases are summarized in 

Table 2 below.  

 

Table 2. Argumentation of the ECtHR in selected hate speech-related cases 2012-2019 

 

Judgement Argumentation in support of the 

utterance in question being hate 

speech  

Argumentation in support of 

the utterance in question 

being protected speech (not 

hate speech) 

1.  Aksu v. Turkey, 

2012 

 

 A need for balance between 

the applicant’s rights and 

public interest in freedom of 

expression is underlined. 

 

Necessity of restrictions of 

freedom of information must 

be established 

“convincingly” (par.64). 

 

The Court reiterates the 

importance of the context 

and concludes that balance 

between the rights of the 

Roma community and 

freedom of expression was 

found.  

 

2.  Öner and Türk  The Court mentions the term 



 81 

v. Turkey, 2015 “hate speech”, adding that it 

is capable of inciting 

violence by appealing to the 

irrational hatred against a 

person or a group of people 

(par.24). 

 

The Court concludes that the 

speech in question was not 

hate speech, but critical 

assessment of the authorities 

that did not contained 

terrorist propaganda and did 

not encourage violence 

(par.24) 

 

3.  Balázs v. Hungary 

2015 

State has an obligation to 

establish any possible racist 

motive and ethnic hatred behind 

the attack (par.53). 

 

Any racist verbal abuse is highly 

relevant in cases involving ethnic 

minorities (par.61). 

 

 

 

4.  Perinçek 
v. Switzerland, 2015 

 Public statements that 

concern a matter of public 

interest must have higher 

protection under Article 10 

of the Convention  par. 

(197).  

 

The conviction of the 

speaker was not necessary in 

a democratic society 

(par.158). 

 

The statement in question 

did not call for hatred or 

intolerance and the context 

lacked tensions, therefore 

the speech of the applicant 

must have been protected by 

the Article 10. 

 



 82 

5.  Király and 

Dömötör v. Hungary, 

2017 

The Court emphasized the 

importance of the context and 

states that cases with an alleged 

hate speech are “highly context-

specific” (par.74). 

 

6.  Raelien Suisse v 

Switzerland, 2012 

 

 

It is important to determine the 

type of speech and its purposes. 

Speech with commercial 

purposes – such as commercial 

advertisement – should have less 

protection by the Article 10 than 

political debate (par.62). 

 

It is important to search for 

relevant and proportional 

restrictions to speech that raises 

concerns (par.75).In this case, the 

decision of the national courts to 

ban the poster was relevant 

(par.76).  

 

 

7. Belkacem v. 

Belgium, 2017 

 

Advocating for religious 

fundamentalism while calling for 

violence can amount to hate 

speech (par.34). States have an 

obligation to prohibit it (ibid). 

 

General attack on people based 

on their religious beliefs or lack 

of such is incompatible with the 

values of the Convention and 

cannot be protected by the 

Article 10 (par.33).  

 

 

 

8. Vejdeland and 

Others v. Sweden, 

2012 

 

Even if certain statements  do not 

propagate hateful acts as such but 

are serious allegations that 

reinforce prejudices, they might 

be prohibited (par.54 and 55). 

 

Discrimination based on sexual 

orientation should be considered 

as serious as discrimination on 

the ground of race, sex, colour or 

origin (par.42) 

 



 83 

9.  Identoba and 

Others v. Georgia, 

2015 

 

Homophobic motives must be 

investigated in hate speech and 

subsequent threats.   

 

The context of the situation must 

be taken into account. 

 

10. M.C. and A.C. 

v. Romania, 2016 

 

 

 

Authorities must investigate 

homophobic motives.  

 

The context and, if applicable, 

the following attacks must be 

taken into account.  

 

 

 

11. Delfi AS 

v. Estonia, 2015 

The Court acknowledges that 

Internet can be beneficial for 

freedom of expression, while at 

the same time enabling 

dissemination of unlawful 

speech, including hate speech.  

 

Therefore, a balance between 

freedom of expression and 

prohibiting unlawful speech must 

be found.  

 

Internet news portals have 

“duties and responsibilities” 

when their users disseminate 

unlawful speech through their 

sites (ibid, par.113) 

 

In this case, automatic filter-

system used by the portal to filter 

the instances of hate speech was 

insufficient and failed to prevent 

the violation of the rights of 

others (ibid, par.156).   

 

 

12.  Magyar 

Tartalomszolgáltatók 

Egyesülete and 

Index.hu Zrt v. 

Hungary, 2016 

 

 

  

Analysis of the speech itself 

must be conducted. 

“Offensive and vulgar” 

comments do not necessarily 

amount to hate speech 

(par.64).  



 84 

 

 

 

 

 

 

 

13.  Pihl v. Sweden, 

2017  

 

 Although offensive, the 

comment in question did not 

incite to violence and did 

not amount to hate speech 

(par.37). 

 

National court found a 

balance between freedom of 

expression and protection of 

the applicant’s rights as the 

comment was quickly 

removed and an apology 

issued (par.29 and 30).  

14.  Smajić v. Bosnia 

and Herzegovina, 

2018  

 

Conviction of the applicant was 

prescribed by law for the 

protection of the rights of others 

(par.32). 

 

Even hypothetical statements 

might be prohibited, taking into 

account the context – in this case, 

sensitive subject of ethnic 

relations between Bosnians and 

Serbs (par.39). 

 

15.  Savva Terentyev 

v. Russia, 2018  

 

 Insulting and offensive 

remarks might still be 

protected by the Article 10, 

provided that they do not 

incite hatred or violence 

(par.69).  

 

Criminal conviction is 

disproportionate for 

offensive speech without 

incitement to hatred 

(par.86).   



 85 

 5.   Social media and online hate speech 

 

As was illustrated by the judgments of the European Court of Human Rights, disputes 

related to the alleged hate speech online have already reached the courts. It is reasonable 

to assume, however, that the vast majority of hate speech incidents that have happened or 

are happening in the Internet never reaches courts or any other similar institution. So 

what are the ways to combat hate speech online, in particular, in major social media? Are 

they effective and how do they correspond to the case law of the European Court of 

Human Rights? Who decides what hate speech on the Internet is? These and other 

questions are discussed in this chapter through the analysis of the guidances and practise 

of the major social media – Facebook, Google (YouTube) and Twitter. 

 

As was noticed in one of the recent studies on fake news – another topic of current 

interest, also closely connected to the freedom of expression – social media in western 

countries enjoy not being held responsible for the content that their users publish on the 

platforms (Niklewicz, K., 2017, p.29). According to the author of this study, this is a 

result of a mutual desire of the US and the EU to protect industry without full 

understanding of the potential of these new media (ibid). In other words, this legal 

vacuum creates a situation when social media themselves are able to decide what content 

they want to prohibit or restrict and how they want to do it.  

 



 86 

When it comes to the online hate speech regulation, it was decided to analyse community 

standards, guidances, users’ rules and even official statements as a primary source of 

information regarding the positions of the respective social media.  

 

 

5.1. Hate speech definitions in social media’s guidances   

 

Hate speech in Facebook  

 

The primary source for Facebook’s hate speech policy is Facebook Community Standards 

that describe the policy and rules of this social media. Among others, in the Facebook 

Community Standards sections like “Violence and Criminal behaviour” and 

“Objectionable content” can be found. Interestingly enough, hate speech is assigned to 

the latter, whereas the former discusses issues such as violence, promoting crime, 

dangerous organizations etc. (see: Facebook Community Standards). Hate speech, on the 

other hand, is placed among “adult nudity/sexual activity” and “violence and graphic 

content” (ibid). This might be regarded as confusing, giving that hate speech is isolated 

by Facebook in a separate category from criminal activity and violence, even though it 

can be closely related issues.  

 

In the Facebook Community Standards it is stated that the social media understands under 

hate speech a content that directly attacks people based on specific characteristics such as 

race, nationality, religion, sex, gender, sexual orientation etc. (Facebook Community 



 87 

Standards, section Hate speech). It is evident that this definition closely corresponds to 

the definitions of hate speech discussed in the chapter 2.2 of the present thesis. Moreover, 

Facebook’s list is even broader and includes additional categories.  

 

In a separate section of the social media’s help centre - Tools for addressing abuse – 

Facebook encourages its users to use special tools that were implemented to avoid 

abusive content. For example, Facebook advices to unfriend or block a person who sent 

an offensive message, use privacy settings and report the abuse and abuser (Facebook, 

Tools for addressing abuse). Report link is provided near the content on Facebook, which 

makes this procedure relatively easy to use.  

 

Finally, in an additional statement, Facebook’s VP of Global Public Policy Marne Levine 

claims that Facebook prohibits content that is harmful – such as anything that can result 

in real violence in the real world or anything that emotionally distresses an individual – 

but allows controversial or offensive content (Levine, 2013). Whereas this seems to be in 

accordance with the Article 10 and with the case law of the European Court discussed 

above, it is not clear how exactly Facebook’s content moderators separate offensive 

content from the content that distresses its users. In reality, this boundary is so blurred 

that even the European Court quite often takes its decisions not unanimously, but with a 

number of dissenting opinions. How the social media can promise to distinguish one from 

another, remains an open question. 

 



 88 

In fact, the responses to this statement clearly indicate that in practice the rules and policy 

of the social media do not function that smoothly as promised. Thus, the statement by 

Marne Levine was published in 2013 and subsequently has received more than 2,400 

comments. Many users criticize Facebook for its policy and actions regarding hate speech 

and demonstrate alleged mistakes and passive approach towards the reports of abusive 

content (see commentaries in Levine, 2013). Whereas it is not possible to prove all the 

users’ claims and published links, content analysis of the comments clearly indicates that 

Facebook’s approach does not fully solve the problem of hate speech on this platform.  

 

This is also evident from the two studies conducted by German Ministry of Justice in 

2016 and 2017. In 2016, the study of Facebook’s practice regarding hate speech 

regulation has shown that only 46% of flagged criminal content was deleted by the social 

media, out of which 42% – during the promised 24 hours (Löschung rechtswidriger 

Hassbeiträge bei Facebook, 2016). On the other hand, when the rest was later reported via 

direct email at the Facebook support team, 84% of criminal content was deleted, 48% - 

during the first 24 hours (ibid).  

 

In 2017 the results of the same study were even more alarming: only 39% of flagged 

criminal content was deleted by Facebook, 33% – during the first 24 hours (Löschung 

rechtswidriger Hassbeiträge bei Facebook, 2017). After the remaining hate speech was 

reported via email, 88 percent of criminal content was deleted, 76 percent – within 24 

hours (ibid). 

 



 89 

That is why organizations such as the International Auschwitz Committee expressed their 

frustration and dissatisfaction with the practice of Facebook regarding hate speech. The 

Committee in particular stated that Facebook displays signs of arrogance towards this 

problem and thus contributes “to poisoning the social climate” especially in Germany, but 

also in the world in general (Deutsche Welle, 2016).  

  

Hate speech in YouTube (Google)  

 

One of the most popular platforms in the Internet – YouTube – also frequently encounters 

hate speech published by its users. YouTube, which belongs to the Google Company, has 

its own policy regarding this issue.  

 

Thus, in the Policy center of the site a specific section “Hate speech policy” can be found. 

In this section YouTube defines hate speech as content that promotes violence or incite 

hatred against groups or individuals based on such characteristics as race, ethnicity, 

religion, disability, age, gender, sexual orientation or veteran status (YouTube, Policies, 

safety, and reporting). These categories can be compared to the ones described by 

Facebook: whereas there are some minor differences, such as mentioning of the veteran 

status by YouTube, in general the definitions may be considered similar.  

 

YouTube, too, proposes its users several possibilities to combat hate speech on the 

platform: block the user, flag the video, file an abuse report in case numerous violation by 

the same user are found or report a violation of the local laws. Both flagging and  

https://support.google.com/youtube/topic/2676378?hl=en&ref_topic=6151248
https://support.google.com/youtube/topic/2676378?hl=en&ref_topic=6151248


 90 

reporting tool are easily acceptable and convenient to use (YouTube, Report 

inappropriate content). 

. 

However, another study conducted by the German Ministry of Justice estimated that in 

2016 YouTube deleted only 10% of the flagged content (Löschung rechtswidriger 

Hassbeiträge bei Facebook, YouTube und Twitter, 2016). Direct report by a user with so-

called privileged account (possible only on YouTube and Twitter, but not on Facebook) 

has led to the removing of another 35% of the criminal content (ibid). Finally, after direct 

report of abuse content via email, YouTube deleted 53% of such content, leaving only 

2% online (ibid). 

 

The results were better in a similar study in 2017. YouTube has significantly improved its 

response towards flagged content and deleted 90% percent of it (Löschung rechtswidriger 

Hassbeiträge bei Facebook, YouTube und Twitter, 2017). After a direct email to the 

support center, the remaining 10% of criminal content were also removed from the 

platform. 

 

Hate speech in Twitter  

 

In its turn, Twitter has its own rules that describe content boundaries and limitations 

connected to the abusive behaviour. Twitter does not use the term “hate speech”, but 

mentions hateful conduct alongside with violent threats, harassment, multiple account 

abuse, self-harm impersonation, etc. (Twitter rules)  



 91 

 

Twitter’s rules define hateful conduct as promotion of violence against people based on 

race, ethnical or national origin, sexual orientation, gender, religion, age, disability or 

disease – with a slight difference, the same categories that were mentioned in the 

Facebook Community Standards (Twitter Rules). Thus, it is possible to conclude that 

although there is no universally accepted definition of hate speech in the normative acts, 

all three platforms that are analysed in the present thesis define hate speech or hateful 

conduct via similar definitions.  

 

Nevertheless, the study commissioned by the German Ministry of Justice, quoted above, 

estimates that in 2016 Twitter deleted only 1% of flagged criminal content (Löschung 

rechtswidriger Hassbeiträge bei Facebook, YouTube und Twitter, 2016). Another 75% 

were removed after direct report by user with a privilege account. Finally, 6% were 

deleted after contact via email to the support center, leaving 18% of reported criminal 

content online (ibid). 

 

According to a similar study in 2017, Twitter did not improve its practice regarding 

flagged criminal content: only 1% was removed (Löschung rechtswidriger Hassbeiträge 

bei Facebook, YouTube und Twitter, 2017). However, direct report and contact via email 

to the support center proved to be more effective: eventually, 100% of the criminal 

content were deleted by Twitter (ibid).  

 



 92 

To sum it up, definitions of hate speech in three major social media are presented in 

Table 3:  

 

Table 3. Hate speech definitions by Facebook, YouTube and Twitter  

 

Facebook Content that directly attacks people based on specific characteristics such as 

race, nationality, religion, sex, gender, sexual orientation, disability, etc.   

Attempts at “humor or satire” that might be perceived as “threat or attack” 

are allowed. (Facebook Community Standards, section Hate speech). 

YouTube 

(Google) 

Content that promotes violence or incites hatred against groups or 

individuals based on such characteristics as race, ethnicity, religion, 

disability, age, gender, sexual orientation or veteran status (YouTube, 

Policies, safety, and reporting). 

 

Twitter  No mentioning of the term “hate speech”, but hateful conduct that is defined 

as “promotion of violence” against people based on race, ethnical or 

national origin, sexual orientation, gender, religion, age, disability or 

disease (Twitter rules).  

 

 

 

 

 

https://support.google.com/youtube/topic/2676378?hl=en&ref_topic=6151248


 93 

5.2. Regulation of hate speech on social media  

 

Social media employ three main strategies in regulating hate speech on the platforms: use 

of automatic regulation5, reports submitted by users and content moderation that is done 

by its content-reviewers (Laub, 2019). At the end of 2018 Facebook alone employed 15 

000 content-reviewers who moderate content in more than 50 languages (Facebook 

Newsroom, 2018).  

 

Already in 2015 three major technological companies – Facebook, Google and Twitter  - 

stated that they were going to delete hate speech from their platforms within 24 hours 

(Reuters, 2015). This decision was taken amid the so-called refugee crisis in 2015, as an 

additional measure to combat rising online racism.  

 

Just a year later, in 2016 the European Commission issued a Code of Conduct on 

countering illegal online hate speech (European Commission, 2016). It is a voluntary 

code that demands to remove the majority of hate speech within 24 hours, while defining 

hate speech as a conduct that is “publicly inciting to violence or hatred” towards groups 

or individual members of a group that is distinguished by special characteristics such as 

race, religious beliefs, nationality or ethnicity” (ibid). 

 

Facebook, Twitter, YouTube, and Microsoft all signed the Code of Conduct in May 2016 

(European Commission, 2017). In 2017, European Commission published a subsequent 

                                                 
5 Artificial intelligence will be discussed below in paragraph 5.3. 



 94 

press-release in which it underlined an important progress in combating hate speech 

online. However, challenges remain significant: according to the European Commission’s 

data, in 2016-2017 IT companies  

- removed hateful content only in 59% reported cases;  

- reviewed 51% of the reported hateful content within 24 hours (ibid).  

 

Finally, in the fourth review of the implementation of the Code of Conduct, published in 

2019, European Commission concluded that hateful content flagged by users  

- was deleted by IT companies in 72% of cases,  

- was reviewed within 24 hours on average in 89% of cases (European Commission, 

2019, p.1).  

 

Altogether, during a period of 6 weeks in November-December 2018 the IT companies 

that signed the Code of Conduct6 received 4392 notifications of the hateful content from 

the countries-members of the EU (ibid).  

 

According to the Commission’s review, deletion of the content by IT companies depends 

on its nature: content that advocates for violence against specific groups of people was 

removed in 85,5% of the cases, whereas the defamatory speech and images were deleted 

only in 58,5% of the cases (ibid, p.3). The most common types of hate speech were hate 

speech based on nationality (including speech against migrants), hate speech based on 

                                                 
6 Apart from the above mentioned Facebook, Twitter, YouTube, and Microsoft, Instagram (279) and 

Google+ have joined the Code of  Conduct in 2018 (European Commission, 2019, p.2).  



 95 

sexual orientation and hate speech based on religious believes, specifically – anti-Muslim 

speech (ibid, p.5).  

 

Independently from the European Commission and taking into account that percentage of 

removal of the flagging content was disproportionately weak in 2016-2017,  in June 2017 

German Parliament has voted to adopt a Social Networks Enforcement Law that is 

supposed to make social media more accountable for the content posted by their users. 

This law requires social media to pay up to 50 million euros in case they fail to remove 

hate speech from their platforms. Similar propositions can be heard in France and Great 

Britain (Toor, 2017).   

 

Whereas this measure may be seen as a resolute and strong attempt to combat hate speech 

in the major social media, some organizations expressed their concerns regarding the 

effect of this law on the freedom of speech in Germany. For instance, David Kaye, UN 

Special Rapporteur to the High Commissioner for Human Rights, noticed that the 

German law gives private firms too much responsibility to police freedom of expression 

(McGoogan, 2017). According to Kaye, legislation, capable of limiting free speech, must 

be applied by an independent body (ibid).  

 

Some other experts expressed their concerns regarding this measure. Mirko Hohmann, a 

Project Manager at the Global Public Policy Institute (GPPi) claimed that such a law will 

make social media companies delete content “excessively”, after they facing fines of up 

to 50 million euros (McGoogan, 2017). This may lead to errors and removal of lawful 

http://www.bmjv.de/SharedDocs/Gesetzgebungsverfahren/Dokumente/RegE_NetzDG.pdf?__blob=publicationFile&v=2
http://www.gppi.net/issue-areas/data-technology-politics/


 96 

content as a precaution (ibid).  Whereas the influence of the German law on hate speech 

and freedom of speech is still to be examined, there are other ways to combat hateful 

content online. 

 

 

5.3. Alternative ways of combating online hate speech: attempts of 

automatic regulation 

 

In recent years a new approach towards online hate speech regulation has been developed 

– namely, automatic regulation of such content. Artificial intelligence means that hate 

speech is being regarded from the point of view of computer science. In 2018-2019 this 

approach is still a relatively new field and available tools for automatic hate speech 

regulation are scarce, among other things – due to lack of systematic data (Fortuna, 

Nunes, 2018, p.3). But the number of articles on automatic hate speech regulation from 

computer science and engineering is rising in recent years, which indicates that 

technologies are being developed (ibid, p.22).   

 

Algorithms for detecting hate speech are based on text mining and machine learning 

technologies that attempt to classify hate speech (ibid, p.16). Interestingly enough, 

researchers from the computer science emphasize that one of the problems in conducting 

this classification is low level of agreement regarding hate speech classification “by 

humans” (ibid, p.25).  

 



 97 

While acknowledging that automatic regulation of hate speech remains an open issue, 

some researchers report positive results of binary classification “hate speech-not hate 

speech” (Del Vigna et al, 2017, p.94). Taking into account the huge amount of user-

generated content, computer scientists underline that human moderators are not capable 

of monitoring major social media, and thus automatic regulation has a great potential for 

the future (ibid).  

 

As was mentioned above, Facebook, YouTube and Twitter all employ automatic content 

regulation that to a certain extent concerns hate speech, but the availability of data 

regarding this regulation, its accuracy, objectivity and further consequences of such 

approach remain an issue for another study. In the current thesis it is important to notice 

that currently social media also rely on human moderators who apply the rules specified 

in the guidances inconsistently (Tobin et al, 2017). Facebook seems to acknowledge this 

problem, with Vice President of the company Justin Osofsky saying to journalists “we 

must do better” (ibid). At the same time, in 2019 Mark Zuckerberg called for a “global 

regulation” that could establish content and data standards (Laub, 2019).  

 

 

5.4. Other initiatives to counteract online hate speech  

 

Finally, one more approach related to online hate speech regulation must be mentioned –

initiatives, independent from governments or social media. Most prominent examples of 

such initiatives is “Hate Speech International”. 

https://www.propublica.org/people/ariana-tobin


 98 

 

“Hate Speech International” is an independent network of journalists who conduct 

cooperative and international research on extremism and hate speech online and offline.  

The initiative was founded by a Norwegian journalists Kjetil Stormark and Øyvind 

Strømmen and received grants from Norwegian  Freedom of Expression Foundation (Fritt 

Ord) and Ministry of Foreign Affairs.  

 

“Hate Speech International”, however, operates independently and aims at displaying 

radicalization and extremist networks (Hate Speech International, About). For this 

purpose “Hate Speech International” regularly conducts related studies and publishes 

them on their website.  

 

 “Hate Speech International” is not the only initiative to combat hate speech that can be 

found online. For instance, similar aims – to monitor, to report, to educate about hate 

speech and policies and tools against it –  had a project “No hate speech youth campaign” 

that was conducted by the Council of Europe in 2012-2017, as well as presently active 

Italian project eMore and pan European project MANDOLA, funded by Rights, Equality 

and Citizenship (REC) Programme of the European Commission.  of the European 

Union.   

 

 

 

 

 



 99 

6. Conclusion  

 

In the present thesis the collision between freedom of expression and hate speech has 

been explored from two different perspectives: the one of the European Court of Human 

Rights, established in 1959, and the one of the three major social media platforms, 

developed during the last decade. In particular, case law of the Court related to hate 

speech has been analysed, as well as community standards and users’ rules of Facebook, 

YouTube and Twitter.  

 

Several conclusions have been made during analysis of the case law of the European 

Court of Human Rights. The most important of them were:  

 

1) European Court of Human Rights does not provide a single definite definition to the 

term “hate speech”, preferring to leave it open in order to approach each case on an 

individual basis. This attitude very much corresponds to the general lack of a single 

definition of the term “hate speech”, as was observed during the review of scientific 

literature. Existent definitions, discussed in the paragraph 2.2., permitted to construct 

following broad and flexible hate speech definition for the purposes of the present thesis: 

we propose to understand under hate speech, utterances that (1) encourage hatred (2) 

towards individuals or groups of people (3) based on various particular characteristics (4) 

which are stigmatized and (5) are employed to legitimize hostility. Furthermore, hate 

speech (6) may be understood as a continuum that (7) does not necessarily result in 

violence, but (8) silences, discriminates and threatens targeted individuals or groups;  



 100 

 

2)  The term “hate speech” was not included in the European Convention on Human 

Rights. European Court of Human Rights first used this term in four judgments, passed in 

1999, namely in Sürek v. Turkey (No. 1, par. 62); Sürek & Özdemir v. Turkey (par.63); 

Sürek v. Turkey (No. 4, par. 60) and Erdogdu & Ince v. Turkey (par. 54). Without 

providing an exact definition of the term, in the fourth of these cases the Court stressed 

that whereas some information may lawfully “offend, shock or disturb”, it the case Sürek 

v. Turke the utterance in question was “hate speech and glorification of violence” ( Sürek 

v. Turke, 1999, par.62). In other words, the Court emphasized the fact that there is a new 

type of speech – expression that goes beyond the protection given to the offensive speech 

by the judgment in Handyside v. the United Kingdom, 1976, and specifically underlines 

violence as a necessary component of it. Yet, even after the introduction of the term “hate 

speech” in its practice, the Court quite often omits it and uses instead various 

synonymous phrases;  

 

3) In 2019 the Court published an updated version of its Factsheet hate speech where it 

provides a typology of cases related to hate speech. This typology seems to be 

unnecessary detailed and not completely straightforward, yet it provides a full overview 

of the case law that concerns with information that the Court itself qualifies as hate 

speech. One could assume that such approach stresses an idea that it is not possible to 

provide a single finite definition to hate speech which instead should be regarded as an 

open spectrum of speech, even if some experts emphasize a need not to expand the term 

“hate speech” indefinitely (Tulkens, 2012, p.295);   



 101 

 

4) Without a definition, case law of the Court remains the main source for scholars to 

analyse the Court’s approach towards hate speech. The analysis of the case law, indeed, 

illustrates how the Court determines whether the speech in question is hate speech or not. 

Thus, the Court seems to employ several criteria on a case to case basis:  

 a) Nature of the speech; 

b) Purpose of the speech (incitement of hatred, violence, and fear as opposite to starting 

public debate; 

c) Significance of the speech (presence or absence of public interest); 

d) Context of the speech; 

e) Necessity of convincing the speech and the speaker in a democratic society; 

f) Presence of a balance between freedom of expression and rights of others in decisions 

made by the national courts. 

 

5) Additionally, following argumentation of the Court was found in the analysed 

judgments from the period 2012-2019:   

a) There should be a balance between the protection of individual rights and public 

interest in freedom of expression (Aksu v. Turkey, 2012); 

b) Necessity of restrictions of freedom of information must be established “convincingly” 

(ibid, par.64);  

c) Hate speech is connected by the Court to inciting violence by “appealing to the 

irrational hatred against a person or a group of people” (Öner and Türk v. Turkey, 2015, 

par.24);  



 102 

d) Critical assessment of the authorities without incitement to violence does not constitute 

hate speech (ibid);  

e) Racist verbal abuse is highly relevant in cases involving ethnic minorities and states 

have an obligation to establish possible racist motive behinds attacks (Balázs v. Hungary 

2015, par.53);  

f) Public statements that concern a matter of public interest must have higher protection 

under the Article 10 (Perinçek v. Switzerland, 2015, par. 197);  

g) On the contrary, speech with commercial purposes – such as commercial 

advertisement – should have less protection by the Article 10 than political debate 

(Raelien Suisse v Switzerland, 2012, par.62);  

h) Religious fundamentalism coupled with incitement to violence amounts to hate speech, 

as much as attack on people based on their religious beliefs (Belkacem v. Belgium, 2017, 

par.34);  

i) Even if certain statements do not propagate hatred but are allegations that reinforce 

prejudices, they might be prohibited (Vejdeland and Others v. Sweden, 2012, 54);  

k) Discrimination based on sexual orientation should be considered as serious as 

discrimination on the ground of race, sex, colour or origin (Vejdeland and Others v. 

Sweden, 2012, par.42). Homophobic motives must be investigated (Identoba and Others 

v. Georgia, 2015).  

l) Internet news portals have “duties and responsibilities” when their users disseminate 

unlawful speech through their sites (Delfi AS v. Estonia, 2015, par.113);  

m) Even hypothetical statements might be prohibited, taking into account the context 

(Smajić v. Bosnia and Herzegovina, 2018, par.39);  



 103 

n) Criminal conviction is disproportionate for offensive speech without incitement to 

hatred (Savva Terentyev v. Russia, 2018, par.86).  

 

At the same time, subsequent analysis of the social media’s approach towards hate speech 

provided the following observations:  

 

6) Taking into account an amount of user-generated content, it is possible to assume that 

the majority of hate speech incidents now happens online and never reaches courts. Three 

major social media – Facebook, YouTube (Google), Twitter – all have developed 

community standards and users’ rules where they define hate speech and prohibit it on its 

platforms. Whereas there are some minor differences between the definitions of hate 

speech by these companies, it is possible to acknowledge that they are more or less 

identical. The problem lies in the implementation of these rules;  

 

7) For instance, Facebook claims to distinguish “offensive content” from the content that 

“distresses its users”. This is a daunting task even for the European Court that quite often 

takes its decisions not unanimously, but with a number of dissenting opinions. How 

exactly and how accurately social media approach this issue, remains an open question. 

More transparency from social media is needed in order to investigate this issue;  

 

8) Several studies conducted by German Ministry of Justice and by the European 

Commission in recent years indicated that social media failed to remove 100% of content 

flagged as hate speech, although this statistic improves year after year;  



 104 

 

9) In western countries, social media are not being held responsible for the content 

published by their users. That means that social media themselves decide what contact 

they want to prohibit or restrict. Yet, there is a tendency in the opposite direction. In 2016 

the European Commission issued a Code of Conduct on countering illegal online hate 

speech that was signed by Facebook, YouTube (Google), Twitter, among others. 

Simultaneously, in June 2017 German Parliament has voted to adopt a Social Networks 

Enforcement Law that declares social media responsible for illegal hateful content on 

their platforms;  

 

10) This law has been criticised by experts who believe that it gives too much 

responsibility to police freedom of expression to the private companies. Among others, 

this concern was expressed by UN Special Rapporteur to the High Commissioner for 

Human Rights David Kaye who urged instead to organise an independent body for online 

content regulation. Interestingly enough, Facebook CEO Mark Zuckerberg recently called 

for a “global regulation” of content and data, too (Laub, 2019);  

  

11) Eventually, at the present moment online hate speech regulation lies with social 

media that employ three strategies: automatic regulation, users’ reports and content 

moderation done by human reviewers. In 2018 Facebook alone employed 15 000 content-

reviewers. These people, de-facto, serve as new judges in a new court. Their qualification 

for this task and consistency with established practice, such as, for instance, the practice 

of the European Court of Human Rights, are highly questionable;    

http://www.bmjv.de/SharedDocs/Gesetzgebungsverfahren/Dokumente/RegE_NetzDG.pdf?__blob=publicationFile&v=2
http://www.bmjv.de/SharedDocs/Gesetzgebungsverfahren/Dokumente/RegE_NetzDG.pdf?__blob=publicationFile&v=2


 105 

 

12) Meanwhile, a new approach towards online hate speech regulation is being developed 

by computer scientists who believe that human moderators cannot monitor major social 

media due to the amount of user-generated content. Automatic content regulation is based 

on text mining and machine learning technologies that attempt to classify hate speech. 

Whereas automatic regulation might result in decreasing of online hate speech, it is 

reasonable to assume that application of this technology will create new threats and new 

restrictions for freedom of expression. In reality, it means that society is going to 

substitute already not sufficiently qualified human moderators with self-educated 

automatic technologies that will decide for us what content is allowed for humans and 

what it judges necessary to prohibit;   

 

13) Overall, this recent popularity of extensive regulation approach seems to indicate that 

human civilisation failed to take its hatred under control. Instead of promoting alternative 

measures – such as improving education, rising awareness about the problem, increasing 

tolerance and intercultural dialogue, providing voice to the victims of hate speech – 

society came to a conclusion that prohibition and automatic recognition will conclusively 

solve the problem, as if removal of online hate speech will at the same time remove 

hatred from the offline communities.  

 

Automatic hate speech recognition is, indeed, needed for a vast amount of user-generated 

content; but without supervision by international qualified body, it cannot guarantee 

freedom of expression, as much as without broad educational and cultural measures, 



 106 

restrictions imposed by social media alone cannot possibly protect human rights in the 

long-term perspective.  

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 



 107 

7. References 

Articles and books 

 

Abrams. (2012). On American Hate Speech Law / In: The content and context of hate 

speech. Rethinking Regulation and Responses |ed. by| Herz, M., Molnar, P. Cambridge 

Univ. Press, p.116-129.  

 

Baker, C. E. (2012) Hate speech / In: The content and context of hate speech. Rethinking 

Regulation and Responses |ed. by| Herz, M., Molnar, P. Cambridge Univ. Press, p.57-80.   

 

Berelson, B. (1952). Content Analysis in Communication Research. Glencoe, Ill: The 

Free Press. 

 

Bleich , E. (2014). Freedom of Expression versus Racist Hate Speech: Explaining 

Differences Between High Court Regulations in the USA and Europe. Journal of Ethnic 

and Migration Studies, 40:2, p. 283-300.  

 

Bonotti, M. (2017). Religion, hate speech and non-domination. Special Issue: Religion 

and Public Life, Vol. 17(2), p. 259–274.  

 

Buyse, A. (2014). Dangerous expressions: The ECtHR, violence and free speech. ICLQ 

vol 63, pp 491–503.  

 



 108 

Davids, N. (2018). On the (in)tolerance of hate speech: does it have legitimacy in a 

democracy?, Ethics and Education, 13:3, p. 296-308.  

 

Del Vigna, F., Cimino, A., Dell’Orletta, F., Petrocchi, M., Tesconi, M. (2017). Hate me, 

hate me not: Hate speech detection on Facebook. Proceedings of the 1st Italian 

Conference on Cybersecurity, p. 86–95.  

 

Downs, D. M., Cowan, G. (2012). Predicting the Importance of Freedom of Speech and 

the Perceived Harm of Hate Speech. Journal of Applied Social Psychology, 42, 6, p.. 

1353–1375. 

 

D'Souza, T., Griffin, L., Shackleton, N., Walt, D. (2018). Harming Women with Words: 

The Failure of Australian Law to Prohibit Gendered Hate Speech, 41 U.N.S.W.Law 

Journal. 939, p.939-976.     

 

Fortuna, P. Nunes, S. (2018). A Survey on Automatic Detection of Hate Speech in Text. 

ACM Comput. Surv. 51, 4, Article 85. 

 

Gelber, K,  McNamara, L. (2016). Evidencing the harms of hate speech/ Social Identities, 

22:3, p. 324-341.  

 



 109 

Greene, J. (2012) Hate speech and the Demos / In: The content and context of hate 

speech. Rethinking Regulation and Responses |ed. by| Herz, M., Molnar, P., Cambridge 

Univ. Press, p.92-116.  

 

Hall, M. A., Wright R. F. (2008). Systematic Content Analysis of Judicial Opinions. 

California Law Review, Vol. 96, No. 1, p. 63-122.  

 

Haraszti, M. (2012). Foreword: hate speech and the Coming Death of the International 

Standard before it was born / In: The content and context of hate speech. Rethinking 

Regulation and Responses |ed. by| Herz, M., Molnar, P., Cambridge Univ. Press ; p. xiii-

xix.  

 

Holoubek, M., Kassai, K., Traimer, M. (2014). Grundzüge des Rechts der Massenmedien, 

Verlag Österreich.  

 

Holsti, O.R. (1969). Content Analysis for the Social Sciences and Humanities. Reading, 

MA: Addison-Wesley.  

 

Korn, G. (2014). Einfuhrung in das Kommunikationsrecht. Facultas Verlags- und 

Buchhandels AG, p. 49-50. 

 

Krippendorff, K. (1980). Content Analysis: An Introduction to Its Methodology. Newbury 

Park, CA: Sage. (2d ed., 2004).   



 110 

 

Maitra, I,  McGowan, M. K. (2010). On Racist Hate Speech and the Scope of a Free 

Speech Principle. The Canadian Journal of Law and Jurisprudence, 23, p 343-372.  

 

Mayring, P. (2000) Qualitative Content Analysis. Forum Qualitative Sozialforschung, 

Volume 1, No. 2, Art. 20.  

 

McGonagle, T. (2013). The Council of Europe against online hate speech: Conundrums 

and challenges. MCM; No. 2013(005). Belgrade: Republic of Serbia, Ministry of Culture 

and Information. 

 

Molnar, P. (2012). Responding to “Hate Speech” with Art, Education, and the Imminent 

Danger Test”// In: The content and context of hate speech. Rethinking Regulation and 

Responses |ed. by| Herz, M., Molnar, P. Cambridge Univ. Press, p.183-198.  

 

Niklewicz, K. (2017). Weeding out Fake News: An Approach to Social Media 

Regulation. Brussels, Wilfried Martens Centre for European Studies. Retrieved from:  

https://martenscentre.eu/publications/weeding-out-fake-news-approach-social-media-

regulation  

 

Parekh, B. (2012). Is there a case for banning hate speech? / In: The content and context 

of hate speech. Rethinking Regulation and Responses |ed. by| Herz, M., Molnar, P., 

Cambridge Univ. Press ; p.37-57.  

https://martenscentre.eu/publications/weeding-out-fake-news-approach-social-media-regulation
https://martenscentre.eu/publications/weeding-out-fake-news-approach-social-media-regulation


 111 

 

Richardson-Self, L. (2017). Woman-Hating: On Misogyny, Sexism, and Hate Speech. 

Hypatia vol. 33, no. 2, p. 256-272.  

 

Richter A. (2012). One step beyond hate speech / In: The content and context of hate 

speech. Rethinking Regulation and Responses |ed. by| Herz, M., Molnar, P., Cambridge 

Univ. Press, p.290-306.  

 

Schauer, F. (2012). Social Epistemology, Holocause Denial, and the Post-Millian 

Calculus / In: The content and context of hate speech. Rethinking Regulation and 

Responses |ed. by| Herz, M., Molnar, P., Cambridge Univ. Press ; p.129-144.   

 

Sottiaux, S. (2011). ‘Bad Tendencies’ in the ECtHR’s ‘Hate Speech’ Jurisprudence. 

European Constitutional Law Review, 7, p. 40–63.  

 

Sparker A. (2005). Narrative analysis: exploring the whats and hows of personal stories. 

Holloway I (ed.).Qualitative Research in Health Care (1st edn). Berkshire: Open 

University Press, p. 191–208. 

 

Stemler, S. (2001). An overview of content analysis. Practical Assessment, Research & 

Evaluation, 7(17). 

 



 112 

Suk, J. C. (2012). Denying Experience: Holocaust Denial and the Free-Speech Theory of 

the State/ In: The content and context of hate speech. Rethinking Regulation and 

Responses |ed. by| Herz, M., Molnar, P. Cambridge Univ. Press, p.144-164.   

 

Tulkens, F. (2012). When to say is to do: Freedom of expression and hate speech in the 

case-law of the European Court of Human Rights. In: Freedom of Expression: Essays in 

honour of Nicolas Bratza /ed.by/ Casadevall, J., Myjer, E., O’Boyle M., Austin, A. 

Oisterwijk, The Netherlands, Wolf Legal Publishers, p. 279-295.  

 

Yong, C. (2011). Does Freedom of Speech Include Hate Speech? Springer 

Science+Business Media B.V., Res Publica, 17, p. 385–403.  

 

Vaismoradi, M., Turunen, H. (2013). Content analysis and thematic analysis: 

Implications forconducting a qualitative descriptive study. Nursing and Health Sciences, 

,15, p. 398–405.  

 

Weber, R. P. (1990). Basic Content Analysis, 2nd ed. Newbury Park, CA. 

 

Reports, declarations and other normative acts  

 

Council of Europe. (1997). Recommendation No. R 97(20) of the Committee of Ministers 

of the Council of Europe to Member States on “hate speech”. Retrieved from: 



 113 

http://www.coe.int/t/dghl/standardsetting/hrpolicy/other_committees/dh-

lgbt_docs/CM_Rec(97)20_en.pdf 

 

Council of Europe. (2013). Report «The hate factor in political speech: Where do 

responsibilities lie?». Retrieved from:  

https://rm.coe.int/16800c170e  

 

European Convention on Human Rights. (1950). Retrieved from:   

https://www.echr.coe.int/Documents/Convention_ENG.pdf%23page=9  

 

European Commission. (2016). Code of Conduct on countering illegal online hate 

speech. Retrieved from: 

http://europa.eu/rapid/press-release_IP-16-1937_en.htm 

 

European Commission. (2017). Countering online hate speech. Commission initiative 

with social media platforms and civil society shows progress. Retrieved from:   

http://europa.eu/rapid/press-release_IP-17-1471_en.htm  

 

European Commission. (2019). Factsheet - 4th monitoring round of the Code of Conduct. 

Retrieved from:  https://ec.europa.eu/info/files/factsheet-4th-monitoring-round-code-

conduct_en 

 

http://www.coe.int/t/dghl/standardsetting/hrpolicy/other_committees/dh-lgbt_docs/CM_Rec(97)20_en.pdf
http://www.coe.int/t/dghl/standardsetting/hrpolicy/other_committees/dh-lgbt_docs/CM_Rec(97)20_en.pdf
https://rm.coe.int/16800c170e
https://www.echr.coe.int/Documents/Convention_ENG.pdf%23page=9
http://europa.eu/rapid/press-release_IP-16-1937_en.htm
http://europa.eu/rapid/press-release_IP-17-1471_en.htm
https://ec.europa.eu/info/files/factsheet-4th-monitoring-round-code-conduct_en
https://ec.europa.eu/info/files/factsheet-4th-monitoring-round-code-conduct_en


 114 

European Court of Human Rights (2019). Factsheet hate speech. Retrieved from:  

http://www.echr.coe.int/Documents/FS_hate_speech_ENG.pdf 

 

International Convention on the Elimination of All Forms of Racial Discrimination. 

(1965). Retrieved from:  

http://www.ohchr.org/EN/ProfessionalInterest/Pages/CERD.aspx  

 

International Covenant on Civil and Political Rights. (1966). Retrieved from: 

https://www.ohchr.org/en/professionalinterest/pages/ccpr.aspx 

 

OSCE. (2005). Combating hate crimes in the OSCE region: an overview of Statistics, 

Legislation, and National Initiatives.  Organization for Security and Co-operation in 

Europe. The Representative on Freedom of the Media.  

 

OSCE (2009). Hate Crime Laws. A practical Guide.  Organization for Security and Co-

operation in Europe, The Representative on Freedom of the Media.  

 

OSCE (2009). Preventing and responding to hate crimes. A resource guide for NGOs in 

the OSCE region. Organization for Security and Co-operation in Europe, The 

Representative on Freedom of the Media.  

       

Universal Declaration of Human Rights. (1948). Retrieved from:  

 http://www.un.org/en/documents/udhr/  

http://www.echr.coe.int/Documents/FS_hate_speech_ENG.pdf
http://www.ohchr.org/EN/ProfessionalInterest/Pages/CERD.aspx
https://www.ohchr.org/en/professionalinterest/pages/ccpr.aspx
http://www.un.org/en/documents/udhr/


 115 

 

UN Committee on the Elimination of Racial Discrimination (CERD). (2013). General 

recommendation No. 35 : Combating racist hate speech,  CERD/C/GC/35. Retrieved 

from: https://www.refworld.org/docid/53f457db4.html [accessed 13 May 2019] 

 

 

Case law of the European Court of Human Rights 

(Access via hudoc.echr.coe.int):  

Aksu v. Turkey, 2012 

Balázs v. Hungary 2015 

Balsytė-Lideikienė v. Lithuania, 2008  

Belkacem v. Belgium, 2017 

Delfi AS v. Estonia, 2015 

Dink v. Turkey, 2010 

Erbakan v. Turkey, 2006 

Erdogdu & Ince v. Turkey, 1999  

Fáber v. Hungary, 2012 

Féret v. Belgium, 2009  

Garaudy v. France, 2003  

Glimmerveen and Haqenbeek v. the Netherlands, 1979 

Gündüz v. Turkey, 2003  

Handyside v. United Kingdom, 1976 

Hösl-Daum and Others v. Poland, 2014 



 116 

Ibragim Ibragimov and Others v. Russia, 2018  

Identoba and Others v. Georgia, 2015 

Ivanov v. Russia, 2007  

Jersild v. Denmark, 1994 

Király and Dömötör v. Hungary, 2017 

Lehideux and Isorni v. France, 1998 

Le Pen v. France, 2010   

Leroy v. France, 2008 

Magyar Tartalomszolgáltatók Egyesülete and Index.hu Zrt v. Hungary, 2016 

M’Bala M’Bala v. France, 2015 

M.C. and A.C. v. Romania, 2016 

Nix v. Germany, 2018  

Norwood v. United Kingdom, 2004 

Öner and Türk v. Turkey, 2015 

Pavel Ivanov v. Russia, 2007  

Perinçek v. Switzerland, 2015 

Pihl v. Sweden, 2017  

Raelien Suisse v Switzerland, 2012 

Roj TV A/S v. Denmark, 2018 

Schimanek v. Austria, 2000  

Šimunić v. Croatia, 2019  

Smajić v. Bosnia and Herzegovina, 2018  

Soulas and Others v. France, 2008 



 117 

Stern Taulats and Roura Capellera v. Spain, 2018 

Stomakhin v. Russia, 2018  

Sürek v. Turkey (No. 1), 1999 

Sürek & Özdemir v. Turkey, 1999 

Sürek v. Turkey (No. 4), 1999  

Vejdeland & others v. Sweden, 2012  

 

 

Web-Resources  

 

European Union Agency for fundamental rights, Hate Speech and Hate Crimes Against 

LGBT Persons. Access via:   

http://fra.europa.eu/sites/default/files/fra_uploads/1226-Factsheet-homophobia-hate-

speech-crime_EN.pdf  

 

Facebook Community Standards. Access via:   

https://www.facebook.com/communitystandards 

 

Facebook Newsroom. (2018). Facts About Content Review on Facebook. Access via:   

https://newsroom.fb.com/news/2018/12/content-review-facts/ 

 

Facebook. Tools for Addressing Abuse. Access via:   

https://www.facebook.com/help/tools 

http://fra.europa.eu/sites/default/files/fra_uploads/1226-Factsheet-homophobia-hate-speech-crime_EN.pdf
http://fra.europa.eu/sites/default/files/fra_uploads/1226-Factsheet-homophobia-hate-speech-crime_EN.pdf
https://www.facebook.com/communitystandards
https://newsroom.fb.com/news/2018/12/content-review-facts/
https://www.facebook.com/help/tools


 118 

 

Hate Speech International. Access via:   

https://www.hate-speech.org/about/political-backdrop/  

 

Hate speech Monitor. Access via:   

https://www.hate-speech.org/hatespeech-monitor/ 

 

Löschung rechtswidriger Hassbeiträge bei Facebook, 2016. Access via:   

http://www.fair-im-

netz.de/WebS/NHS/SharedDocs/Downloads/DE/09262016_Testergebnisse_Facebook.pd

f?__blob=publicationFile&v=2  

 

Löschung rechtswidriger Hassbeiträge bei Facebook, 2017. Access via:   

http://www.fair-im-

netz.de/WebS/NHS/SharedDocs/Downloads/DE/03142017_Recherchebericht_Facebook.

pdf?__blob=publicationFile&v=3 

 

Löschung rechtswidriger Hassbeiträge bei Facebook, YouTube und Twitter, 2016. 

Access via:   

https://www.bmjv.de/SharedDocs/Downloads/DE/Artikel/09262016_Testergebnisse_jug

endschutz_net_Hasskriminalitaet.pdf?__blob=publicationFile&v=1 

 

Löschung rechtswidriger Hassbeiträge bei Facebook, YouTube und Twitter, 2017. 

Access via:   

https://www.hate-speech.org/about/political-backdrop/
https://www.hate-speech.org/hatespeech-monitor/
http://www.fair-im-netz.de/WebS/NHS/SharedDocs/Downloads/DE/09262016_Testergebnisse_Facebook.pdf?__blob=publicationFile&v=2
http://www.fair-im-netz.de/WebS/NHS/SharedDocs/Downloads/DE/09262016_Testergebnisse_Facebook.pdf?__blob=publicationFile&v=2
http://www.fair-im-netz.de/WebS/NHS/SharedDocs/Downloads/DE/09262016_Testergebnisse_Facebook.pdf?__blob=publicationFile&v=2
http://www.fair-im-netz.de/WebS/NHS/SharedDocs/Downloads/DE/03142017_Recherchebericht_Facebook.pdf?__blob=publicationFile&v=3
http://www.fair-im-netz.de/WebS/NHS/SharedDocs/Downloads/DE/03142017_Recherchebericht_Facebook.pdf?__blob=publicationFile&v=3
http://www.fair-im-netz.de/WebS/NHS/SharedDocs/Downloads/DE/03142017_Recherchebericht_Facebook.pdf?__blob=publicationFile&v=3
https://www.bmjv.de/SharedDocs/Downloads/DE/Artikel/09262016_Testergebnisse_jugendschutz_net_Hasskriminalitaet.pdf?__blob=publicationFile&v=1
https://www.bmjv.de/SharedDocs/Downloads/DE/Artikel/09262016_Testergebnisse_jugendschutz_net_Hasskriminalitaet.pdf?__blob=publicationFile&v=1


 119 

http://www.bmjv.de/SharedDocs/Downloads/DE/Artikel/03142017_Monitoring_jugends

chutz.net.pdf?__blob=publicationFile&v=1 

 

 MANDOLA project. Access via:   

http://mandola-project.eu/  

 

No hate speech youth campaign. Access via:   

https://www.coe.int/en/web/no-hate-campaign 

 

Project eMore. Access via:   

https://www.emoreproject.eu/  

 

Statista. Most popular social networks worldwide as of April 2019. Access via:  

https://www.statista.com/statistics/272014/global-social-networks-ranked-by-number-of-

users/ 

 

Twitter Rules. Access via:   

https://support.twitter.com/articles/18311 

 

YouTube, Policies, safety, and reporting. Access via:   

https://support.google.com/youtube/answer/2801939?hl=en  

 

YouTube, Report inappropriate content. Access via:   

https://support.google.com/youtube/answer/2802027 

http://www.bmjv.de/SharedDocs/Downloads/DE/Artikel/03142017_Monitoring_jugendschutz.net.pdf?__blob=publicationFile&v=1
http://www.bmjv.de/SharedDocs/Downloads/DE/Artikel/03142017_Monitoring_jugendschutz.net.pdf?__blob=publicationFile&v=1
http://mandola-project.eu/
https://www.coe.int/en/web/no-hate-campaign
https://www.emoreproject.eu/
https://www.statista.com/statistics/272014/global-social-networks-ranked-by-number-of-users/
https://www.statista.com/statistics/272014/global-social-networks-ranked-by-number-of-users/
https://support.twitter.com/articles/18311
https://support.google.com/youtube/topic/2676378?hl=en&ref_topic=6151248
https://support.google.com/youtube/answer/2801939?hl=en
https://support.google.com/youtube/answer/2802027


 120 

Press articles  

Deutsche Welle. (2016).  International Auschwitz Committee accuses Facebook of 

arrogant approach to hate speech. Retrieved from 

http://www.dw.com/en/international-auschwitz-committee-accuses-facebook-of-arrogant-

approach-to-hate-speech/a-36787344 

 

Laub, Z. (2019). Hate Speech on Social Media: Global Comparisons.Council on foregn 

relations. Retrieved from: https://www.cfr.org/backgrounder/hate-speech-social-media-

global-comparisons 

 

Levine, M. (2013). Controversial, Harmful and Hateful Speech on Facebook. Retrieved 

from: https://www.facebook.com/notes/facebook-safety/controversial-harmful-and-

hateful-speech-on-facebook/574430655911054  

 

McGoogan, C. (2017). Germany to fine Facebook and YouTube €50m if they fail to 

delete hate speech. Retrieved from:  

http://www.telegraph.co.uk/technology/2017/06/30/germany-fine-facebook-youtube-

50m-fail-delete-hate-speech/ 

 

Reuters. (2015). Facebook, Google, Twitter agree to delete hate speech in 24 hours. 

Retrieved from: http://www.reuters.com/article/us-germany-internet-

idUSKBN0TY27R20151215 

 

http://www.dw.com/en/international-auschwitz-committee-accuses-facebook-of-arrogant-approach-to-hate-speech/a-36787344
http://www.dw.com/en/international-auschwitz-committee-accuses-facebook-of-arrogant-approach-to-hate-speech/a-36787344
https://www.cfr.org/bio/zachary-laub
https://www.cfr.org/backgrounder/hate-speech-social-media-global-comparisons
https://www.cfr.org/backgrounder/hate-speech-social-media-global-comparisons
https://www.facebook.com/notes/facebook-safety/controversial-harmful-and-hateful-speech-on-facebook/574430655911054
https://www.facebook.com/notes/facebook-safety/controversial-harmful-and-hateful-speech-on-facebook/574430655911054
http://www.telegraph.co.uk/technology/2017/06/30/germany-fine-facebook-youtube-50m-fail-delete-hate-speech/
http://www.telegraph.co.uk/technology/2017/06/30/germany-fine-facebook-youtube-50m-fail-delete-hate-speech/
http://www.reuters.com/article/us-germany-internet-idUSKBN0TY27R20151215
http://www.reuters.com/article/us-germany-internet-idUSKBN0TY27R20151215


 121 

Tobin, A,, Varner, M, Angwin, J. (2017). Facebook’s Uneven Enforcement of Hate 

Speech Rules Allows Vile Posts to Stay Up. Pro Publica. Retrieved from: 

https://www.propublica.org/article/facebook-enforcement-hate-speech-rules-mistakes 

 

Toor, A. (2017). Germany wants to fine Facebook over hate speech, raising fears of 

censorship. Retrieved from:  

https://www.theverge.com/2017/6/23/15852048/germany-hate-speech-facebook-twitter-

fine-censorship 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

https://www.propublica.org/people/ariana-tobin
https://www.propublica.org/people/madeleine-varner
https://www.propublica.org/people/julia-angwin
https://www.propublica.org/article/facebook-enforcement-hate-speech-rules-mistakes
https://www.theverge.com/2017/6/23/15852048/germany-hate-speech-facebook-twitter-fine-censorship
https://www.theverge.com/2017/6/23/15852048/germany-hate-speech-facebook-twitter-fine-censorship


 122 

8. Abstract  

One of the most essential problems of modern media law is the collision between 

freedom of expression and hate speech. Rapid development of the Internet and social 

media have greatly contributed to the realization of freedom of expression worldwide, but 

at the same time increased dissemination of hate speech. In the new media environment, 

establishing control over hate speech while simultaneously protecting freedom of 

expression became a global issue.  

 

This thesis attempts at looking at this problem from two different perspectives: from the 

position of the European Court of Human Rights which has a long history of passing 

judgments in cases related to hate speech, and from the point of view of the major social 

media which in the last decade have developed their own rules and measures regarding 

this issue.  

 

In order to establish how the European Court of Human Rights decides whether the 

speech in question is hate speech or not, more than 30 judgments of the Court are 

analyzed in detail. The most significant arguments from the judgments are summarised in 

two tables. 

 

The Court’s approach is then compared with the one of the social media that attempt to 

combat hate speech with the help of human moderators and automatic filtering systems. 

A number of observations resulted from this comparison are provided in the conclusion.  

 



 123 

Keywords: freedom of expression, hate speech, European Court of Human Rights, case 

law, social media 

 

 

 

9. Zusammenfassung  

 

Eines der größten Probleme des modernen Medienrechts ist die Kollision von 

Meinungsfreiheit und Hassrede. Die rasche Entwicklung des Internets und der sozialen 

Medien hat weltweit zur Verwirklichung der Meinungsfreiheit beigetragen, gleichzeitig 

hat sie aber die Verbreitung von Hassreden verstärkt. In der neuen Medienlandschaft 

wurde die Kontrolle über Hassreden bei gleichzeitigem Schutz der Meinungsfreiheit zu 

einem globalen Thema.  

 

Diese Masterarbeit versucht dieses Problem aus zwei verschiedenen Perspektiven zu 

betrachten: aus der Position des Europäischen Gerichtshofs für Menschenrechte, der seit 

langem die Fälle im Zusammenhang mit Hassrede beurteilt, und aus der Sicht der 

wichtigsten sozialen Medien, die im letzten Jahrzehnt ihre eigenen Regeln und 

Maßnahmen zu diesem Thema entwickelt haben. 

 

Um festzustellen, wie der Europäische Gerichtshof für Menschenrechte entscheidet, ob es 

sich um eine Hassrede handelt oder nicht, werden mehr als 30 Urteile des Gerichtshofs 

im Detail analysiert. Die wichtigsten Argumente aus den Urteilen sind in zwei Tabellen 

zusammengefasst. 



 124 

 

Der Ansatz des Gerichtshofs wird dann mit dem der sozialen Medien, die Hassreden mit 

Hilfe von Moderatoren und automatischen Filtersystemen zu bekämpfen versuchen, 

verglichen. Die Beobachtungen, die sich aus diesem Vergleich ergeben, werden in der 

Schlussfolgerung präsentiert. 

 

Suchbegriffe: Meinungsfreiheit, Hassrede, Europäischer Gerichtshof für Menschenrechte, 

Rechtsprechung, soziale Medien