Clotilde Pégorier,
Lecturer in Law, University of Essex
The issue of
hate speech regulation has again moved in recent years to the forefront of
legal and political debate in Europe. To note that questions in this area are
complex, and often generate diverging opinions as to the appropriate balance
between legislation and the protection of rights, is no novelty. What is
striking, however, is the marked difference in the tendencies of those “natural
born twins” (Gabriel
Toggenburg), the EU and the Council of Europe, in their respective
approaches to hate speech. How might this be explained? And what, crucially, might
be the wider legislative implications at European level?
The EU and the Fight Against Online Hate Speech
First, let us
consider the EU’s efforts in this context, which might here be exemplified in
relation to the battle against online hate speech. In response to the problem
and threat of terrorism and radicalisation, and prompted in particular by the
attack in Brussels on 22 March 2016, the EU decided to intensify its work on fighting
hate speech – a campaign upon which they had embarked some eight years earlier with
the adoption of the Council
Framework Decision 2008/913/JHA on combating certain forms and expressions
of racism and xenophobia by means of criminal law. As part of its security agenda
for the period 2015-2020, the Commission presented, on 14 June 2016, a communication
outlining action in seven specific areas where cooperation at EU level could
effectively support Member States in preventing and countering radicalisation. Alert
to the ever greater role played by the internet in the dissemination of views
and ideologies, the European Commission took the step of consulting IT
companies with the intention of creating legislation designed to inhibit the
online spread of illegal content inciting violence.
In pursuing such
an initiative, the Commission was, in fact, expanding upon longer-standing
awareness of the importance of preventing the spread of hate speech via media
forms. As Advocate General Yves Bot concluded in his Opinion
from 5 May 2011 with respect to the cases
C-244/10 and C-245/10:
Member states are to ensure that television broadcasts do not
contain any incitement to hatred on grounds of race, sex, religion or
nationality, must be interpreted as also prohibiting broadcasts which, in
attempting to justify a group classify as a ‘terrorist’ organisation by the
European Union, may create reactions of animosity or rejection between
communities of different ethnic or cultural origin (para 93).
In May 2016, the
European Commission had, moreover, already signed an online Code
of Conduct on countering illegal hate speech online with four of the
biggest internet companies – namely, Twitter, Facebook, YouTube and Microsoft. The
code is not legally binding yet would appear to indicate a willingness on the
part of the named IT companies to support the EU’s drive to prevent online hate
– a willingness that owes in some measure, no doubt, to the protections
supplied by Articles 12 to 14 of the e-Commerce
Directive of 8 June 2000, commonly known as the ‘safe harbour’ provisions. According
to Article 12, the provider of a service cannot be held liable for any
information it transmits – including hate speech – as long as it: (a) does not
initiate the transmission; (b) does not select the receiver of the
transmission; and (c) does not select or modify the information contained in
the transmission. Article 14 limits the liability of providers of “information
society services” still further when such services consist only of the “storage
of information” provided by a recipient of the services. This provision only
applies where the provider does not control or have knowledge of the illegal
activity or information; or having gained knowledge or awareness of such
illegal activity expeditiously removes or disables the links to the activity.
However we
speculate on the primary motives of the IT companies, of prime significance is
that they are assisting the EU in its fight against online hate speech. The
Code encourages social media companies to take quick action as soon as a valid
notification of online hate speech has been received, e.g. by removing or
disabling access. It also underlines that, in order to combat the spread of
illegal hate speech, “it is essential to ensure that relevant national laws
transposing the Council Framework Decision 2008/913/JHA are fully enforced by
Member States in the online as well as the in the offline environment.” With the
adoption of the Council
Framework Decision, the EU considered that Member states were permitted to enact
criminal sanctions against anyone:
publicly condoning, denying or grossly trivialising crimes of
genocide, crimes against humanity and war crimes as defined in Articles 6, 7
and 8 of the Statute of the International Criminal Court, directed against a
group of persons or a member of such a group defined by reference to race,
colour, religion, descent or national or ethnic origin when the conduct is
carried out in a manner likely to incite to violence or hatred against such a
group or a member of such a group.
Reading such
provisions takes us to the heart of one of the key dilemmas at the core of
current debates on hate speech – namely, the definition and understanding of
the concept itself. A brief excursus on this point seems warranted here. For
the question of definition remains somewhat thorny – hate speech is a term that
is, at once, both over- and underdetermined. As Anne Weber puts it in her Manual
on Hate Speech for the Council of Europe in 2009:
No universally accepted definition of the term “hate speech” exists,
despite its frequent usage. Though most States have adopted legislation banning
expressions amounting to “hate speech”, definitions differ slightly when
determining what is being banned.
This is
undeniably true. Yet there are international and national sources that provide
useful guidance. The Council of Europe’s Committee of Ministers’ Recommendation
97(20) on “Hate Speech” defined it as follows:
[T]he term “hate speech” shall be understood as covering 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.
Also relevant
here are the provisions of Art. 20, para. 2 of the International Covenant on
Civil and Political Rights (ICPPR) of 1966, which stipulate that “any advocacy
of national, racial or religious hatred that constitutes incitement to
discrimination, hostility or violence shall be prohibited by law.” An
authoritative interpretation of Art. 20, para. 2 is supplied by General Comment
No. 34 by the Human Rights Committee, which reads:
What distinguishes the acts addressed in article 20 from other acts
that may also be subject to limitations, is that for the acts addressed in
article 20, the Covenant indicates the specific response required from the
State: their prohibition by law. It is only to this extent that article 20 may
be considered as lex specialis with regard to article 19 [which establishes
other limitations on freedom of expression]. The acts referred to in article
20, paragraph 2, must cumulatively (a) advocate, (b) be for purposes of
national, racial or religious hatred, and, (c) constitute incitement to
discrimination, hostility or violence. By “advocacy” is meant public forms of
expression that are intended to elicit action or response. By “hatred” is meant
intense emotions of opprobrium, enmity and detestation towards a target group.
“Incitement” refers to the need for the advocacy to be likely to trigger
imminent acts of discrimination, hostility or violence. It would be sufficient
that the incitement relate to any one of the three outcomes: discrimination,
hostility or violence (para. 51).
This
interpretation provides perhaps the fullest, and most useful, elucidation of
hate speech – one that does most to capture its particular power to harm. Read
in conjunction with modern understandings of the potential of online media to
contribute to the dissemination of political views, and to generate and spread
‘hatred’, it casts particularly sharp light, moreover, on how, by enlisting the
support of IT companies, the EU is taking a progressive – and legitimate –
stand in trying to confront modern hate speech in one of its most threatening
forms.
The Council of Europe: The Protection of Freedom of Expression
Over the Fight against (Online) Hate Speech?
The situation is
somewhat different, however, in the case of the other main European organ, the
Council of Europe, which appears to be taking a much more cautious approach. A
latest manifestation of this has been in the Perinçek
case, for instance, where the European Court of Human Rights (ECtHR) decided,
on 15 October 2015, that Switzerland’s criminalisation of Doğu Perinçek for
genocide denial constituted a violation of Article 10 of the European Convention
on Human Rights (ECHR). The Court finding here was that the restriction on
freedom of expression imposed by the Swiss authorities was not proportionate.
This is but the
latest sign of a divergence in the attitudes and response from two European
organs to the issue of hate speech, reflecting a breach within Europe with regards
to the status of hate speech in relation to freedom of expression, the latter
itself a fundamental notion of both the ECHR and the Charter of
Fundamental Rights of the European Union (Article 11).
The prevention
and prohibition of online hate speech has been on the agenda of the Council of
Europe since at least 2001, when the Convention
on Cybercrime was adopted. In 2003, an Additional
Protocol concerning the criminalisation of acts of a racist and xenophobic
nature committed through computer systems was adopted. According to this
Additional Protocol:
1. Each Party shall adopt such legislative measures as may be
necessary to establish the following conduct as criminal offences under its
domestic law, when committed intentionally and without right: distributing or
otherwise making available, through a computer system to the public, material
which denies, grossly minimises, approves or justifies acts constituting
genocide or crimes against humanity, as defined by international law and
recognised as such by final and binding decisions of the International Military
Tribunal, established by the London Agreement of 8 August 1945, or of any
other international court established by relevant international instruments and
whose jurisdiction is recognised by that Party. […]
In June 2016, however,
at the same time that the EU Code of Conduct was adopted, the Council of Europe
Secretary General, concerned about internet censorship, decided that rules for
blocking and removing illegal content must be transparent and proportionate.
This opinion
came after his report
on the state of democracy, human rights and the rule of law, based on a
study conducted by the Swiss Institute of Comparative Law and identifying a
number of shortcomings in some states, became public.
In the report,
the Secretary General clearly stated that:
In the majority of member states, the legal framework on blocking,
filtering and removal of Internet content meets the requirements of being
prescribed by law, pursuing legitimate aims and being necessary in a democratic
society, in accordance with Article 10 of the Convention. Exceptions remain
however, notably with regard to laws regulating hate speech and
counter-terrorism (p. 33).
In view of this,
one can understand why the Grand Chamber of ECtHR decided in the Perinçek case that the Swiss
criminal provision was disproportionate and did not fulfil the criteria of
being necessary in a democratic society. Yet Art. 261bis of the Swiss
penal code provides that ‘any person who publicly denigrates or
discriminates against another or a group of persons on the grounds of their
race, ethnic origin or religion in a manner that violates human dignity,
whether verbally, in writing or pictorially, by using gestures, through acts of
aggression or by other means, or any person who on any of these grounds denies,
trivialises or seeks justification for genocide or other crimes against
humanity, […] is liable to a custodial sentence not exceeding three years or to
a monetary penalty’. It is difficult to see how a criminal law could be much
more transparent or clearer here. The decision to uphold Perinçek’s claim to a
violation of Art. 10 ECHR certainly delivered a blow to the fight against hate
speech at the EU level – as was duly noted by judges Spielmann (president of
the Grand Chamber), Casadevall, Berro, De Gaetano, Sicilianos, Silvis and
Kūris, in their joint
dissenting opinion:
With regard to the finding that there was no obligation on
Switzerland to criminalise the applicant’s statements (see paragraphs 258-68),
we confess to having serious doubts as to the relevance of the reasoning. Can
it not be maintained, on the contrary, that a (regional) custom is gradually
emerging through the practice of States, the European Union (Framework Decision
2008/913/JHA) or ECRI (Policy Recommendation no. 7)? We would also note that
beyond Europe, the United Nations Committee on the Elimination of Racial
Discrimination has repeatedly recommended criminalising negationist discourse.
Can all these developments be disregarded at a stroke by examining the case in
terms of an alleged conflict of obligations? (para 10)
It thus seems that
the Council of Europe is taking a retrogressive step in the fight against hate
speech – both offline and online – as the laws in place regulating hate speech do
not appear to be in line with the ECHR. The approach of the Council also lies
in opposition to that being taken by the EU, rendering the position of EU
members difficult: should they criminalise online hate speech, or should they rather
grant greater weight to Art. 10 ECHR? Indeed, what if Switzerland was an EU
member? By criminalising genocide denial as a form of hate speech liable to
incite violence, as it initially did in the Perinçek case, Switzerland complied
with the Framework Council decision. In so doing, however, it contravened Art.
10 of the ECHR and was thus found guilty of a violation by the ECtHR.
Conclusion
The question of
how to square the protection of freedom of expression with the imposition of
criminal sanctions for hate speech is, doubtless, one which is difficult. Yet
wherever one draws the line between acceptable and unacceptable limits on
freedom of expression, it seems apparent that, at the European level, the EU
and the Council of Europe should be working together much more coherently in
attempting to confront the issue of online (and offline) hate speech.
To this end, the
Council of Europe should liaise more closely with the EU – not least as the
Secretary General, in his 2016
report, commented that:
-In addition to
calling on member states to implement in full the recommendations in this
report, I urge them to make clear their commitment to the European Convention
on Human Rights and the Strasbourg Court. Our Convention system can never be
taken for granted: it depends on the active and constructive engagement of all
governments. By embedding these fundamental freedoms into the legal, political
and social fabric of their nations, Europe’s leaders can build democracies
which are more open and inclusive and, as a result, more secure (p.5)
In order to
facilitate a more consistent approach across Europe, it seems clear that the European
Court of Human Rights itself has to be prepared to allow for greater
restrictions to be placed on freedom of expression, precisely as noted by the
judges in their dissenting opinion in the Perinçek
case. As long as the Strasbourg Court continues to permit freedom of expression
to be used as a catch-all defence, it will remain extremely difficult to combat
online hate speech and to develop a common European standard. Two measures thus
seem necessary. Firstly, a common understanding of what hate speech is and
entails should be striven for – the interpretation supplied by General Comment
No. 34 by the Human Rights Committee provides useful initial orientation,
not least in the manner that it explicates key notions of ‘incitement’ and
‘hatred’, and in the way that it outlines the possible effects of hate speech
beyond physical violence. Secondly, there needs to be common agreement on the
way in which such forms threaten democratic values – how they violate ‘the
respect of the rights or reputations of others’ and may imperil ‘national
security’, ‘public order’, or ‘public health or morals’, and thus constitute a
legitimate restriction on freedom of expression provisions.
Barnard &
Peers: chapter 9
JHA4: chapter II:6
Photo credit: European
Centre for Press and Media Freedom
Comment from Adrian Hunt, Birmingham Law School, University of Birmingham (part 1)
ReplyDeleteThanks for this thought provoking piece, Clotilde. I would like to make a comment on the premise of the article that the decision in Perinçek reflects a difference between the EU position as reflected in the Framework Decision [FD], and the position adopted under the Convention.
It is not at all clear to me - and I cannot see where this article identifies - a clear division between the EU and the ECHR, as regards the relevant rules and principles in a situation such as this.
The decision in Perinçek was, as this piece explains, based upon proportionality grounds. However the subsequent critique of it set out in the blog focuses not on proportionality, but rather on the "prescribed by law" requirement, which did not form the basis for the reasoning of the European Court of Human Rights in the case at all. The critique by the Council of Europe Secretary General referred to in blog post raised concerns about the method of internet censorship, and was primarily premised on the prescribed by law requirement; not the proportionality type issue considered in Perinçek. This is because some EU countries when dealing with speech on the internet have preferred to operate informal processes where public authorities contact ISPs etc expressing concern about particular types of content, “suggesting” the contents removal, which the ISPs may then often do. The interference with speech in such a way operates unofficially and arguably/obviously therefore presents problems in terms of the prescribed by law requirement.
part 2:
DeleteBut that was not the matter at issue in Perinçek, which accepted that the Swiss law did not offend in the prescribed by law sense. Furthermore, Perinçek accepted that each category offence set out in the FD, is capable of being compatible with the Convention.
Thus, Perinçek accepts that speech (such as genocide denial) which is carried out in a manner likely to incite to violence or hatred, may be criminalized without such criminalization being incompatible with the Convention. (see the provisions in the Framework Decision in Article 1 (a) to (d) are ok.). However in the instance case on the facts (for reasons explained below) the court concluded that the speech was not likely to incite violence or hatred.
part 3:
DeletePerinçek also accepts that measures which criminalise such speech ‘carried out in a manner likely to disturb public order’ (FD Article 2) is also permitted by Article 10, (but in that context this is to be given the narrow meaning denoted by “the prevention of disorder” in the English text, as distinct from the possible wider notion argued for by the Swiss Goverment on the basis of the French language text). Since the Swiss Governments specific arguments on this ground were not able to point to the fact that the speech in the particular case was likely to provoke disorder” in the narrow sense, that part of their argument failed.
Perinçek, also accepts that criminalization of speech of this kind which is threatening, abusive or insulting (FD Article 2)(and does not incite hatred or provoke public disorder) may be compatible with the convention, since
'the negative stereotyping of an ethnic group was capable, when reaching a certain level, of having an impact on the group’s sense of identity and on its members’ feelings of self-worth and self-confidence. It could thus affect their “private life” within the meaning of Article 8 § 1 of the Convention' [Perinçek para 200 - relying on Aksu v. Turkey].
And the court concluded on the facts that
‘statements bore on a matter of public interest and did not amount to a call for hatred or intolerance, that the context in which they were made was not marked by heightened tensions or special historical overtones in Switzerland, that the statements cannot be regarded as affecting the dignity of the members of the Armenian community to the point of requiring a criminal law response in Switzerland, ……., that the Swiss courts appear to have censured the applicant for voicing an opinion that diverged from the established ones in Switzerland, and that the interference took the serious form of a criminal conviction – the Court concludes that it was not necessary, in a democratic society, to subject the applicant to a criminal penalty in order to protect the rights of the Armenian community at stake in the present case.’
You and I may disagree with this assessment but it does not in and of itself signify a difference between the position in EU Law and under the Convention. It is rather a specific decision of a court on its facts. The approach of the dissenting judges does not as such offer a different basis in principle to the majority, but rather indicates that they concluded differently as regards the proportionality of the measures in the light of their reading of the facts.
Thus, the assertion which the article makes that the ‘Strasbourg Court continues to permit freedom of expression to be used as a catch-all defence’ arguably seriously over simplifies the process of reasoning adopted in resolving this case, and the many other cases referred in Perinçek, some of which went one way, and some of which went the other.
The call which the article makes therefore for some sort of dialogue to arrive at some sort of agreed position seems strange because it is difficult to discern from the article what it that that the EU and ECHR differ on, which might be the focus of a dialogue, such as to produce a different rule/set of principles such as might definitively lead to different outcomes than the one arrived at in this case? Which part of the FD points to a situation where the conclusion in this case should inevitably lead to a finding that the conviction of Perinçek should be compatible with the Convention? Indeed the FD (Article 7), itself provides that the Framework decision shall not have the effect of modifying the obligation to respect fundamental rights and fundamental legal principles, including freedom of expression, referring both to Art 6 TEU, and the concept more generally (this would include article 10 ECHR - see 14th preambular para).