‘The question is not whether we will be extremists but what kind of extremists we will be. Will we be extremists for hate, or will we be extremists for love? Will we be extremists for the preservation of injustice, or will we be extremists for the cause of justice?’

  —Martin Luther King, Letters from a Birmingham Jail, 1963.

‘The poisonous messages of extremists must not be allowed to drown out the voices of the moderate majority’.

  —UK Government, Prevent Strategy, 2011.


Since September 2001 the international community has witnessed a proliferation of legislation ostensibly designed and intended to protect citizens from individuals and groups in the process of and prepared to use acts of violence against society. As part of this allegedly protective movement numerous States have bolstered, enacted or proposed to enact laws focused on outlawing so-called extremist speech, drawing, whether explicitly or implicitly, a direct line from people publicly expressing ‘extreme’ views to their committing acts of terrorist violence. Such laws, however, have concerning human rights implications, and whilst banning or punishing extremism might be expected of authoritarian governments, the scope of UN-led, USA and EU counter-extremism initiatives indicates a concerning trend of liberal States embracing opportunities to impose severe restrictions on ‘extreme’ speech.1

Labelling views and opinions as extremist is far from new2 but 9/11 can be empirically identified as the major catalyst for the contemporary resurgence of the idea of extremism as a threat. It was through the aftermath of 9/11 that extremism became embedded in domestic and international discourse and practice as the predominant ‘root’ of terrorism. Broadly speaking, terrorists pursue political or societal change through violent means. Ideological motive distinguishes terrorism from other criminal violence and willingness to kill or harm civilians to achieve their ends marks terrorists out from other political activists. ‘Traditional’ counter-terrorism activities, however, focused only on securing and protecting against extant violent plots; strategies sought neither to understand the ‘root causes’ of terrorism nor do anything to address the underlying conditions creating terrorists.3 11 September 2001 changed this. The events of that day thereafter linked the commission of terrorist atrocities with belief in extreme doctrine and drove forward a new global approach to counter-terrorism that fixated on combating harmful expressions of ‘extreme’ ideologies.4 This new approach to counter-terrorism gradually, subtly but inexorably, bound terrorism ever-tighter with the idea of ‘extremism’. Without ever precisely defining ‘extremism’ the publicised imperative of States became to defeat this ‘evil’, almost thereby reducing battling actual terrorist violence to a secondary responsibility.5 This approach was spurred on by the UN, which in 2015 firmly established ‘eliminating extremist speech’ as an urgent and critical goal on the worldwide political agenda.6

In the wake of the globalised ‘war on terror’ counter-extremism has become a widespread, high-agenda policy aim, particularly for liberal States.7 Despite this politicised focus on extremism, however, domestic legislatures have largely skirted round implementing explicit laws prohibiting or penalising extremist speech: freedom of expression sanctions on extremism have more commonly been imposed by defining extremism in policy documents then suppressing identified extremist ideologies through targeted application of laws against hate speech and criminal offences such as ‘inciting terrorism’.8 Notably however, the UK, spearheading the liberal State approach and building on its 2015 ‘Counter-Extremism Strategy’ which defined extremism as ‘vocal or active opposition to fundamental British values including democracy, the rule of law, individual liberty, mutual respect and tolerance of different faiths and beliefs, as well as calling for the deaths of members of the armed forces’,9 is currently in the process of enacting primary legislation explicitly referencing extremism. The UK ‘Counter-Extremism and Safeguarding Bill’ announced in the Queen’s Speech of 18 May 2016 is expected to codify the above definition of extremism and impose a range of mainly civil measures directed towards eliminating allegedly dangerous extremist speech from public dialogue.10

Whatever form States’ extremism legislation takes, laws that restrict the right to freedom of expression citing an imperative of ‘countering extremism’ are of questionable compatibility with international human rights law.11 Freedom of expression is universally considered to be one of the most fundamental and important civil and political rights, deserving of particularly strong protection.12 It is well-established that the right, enshrined in the Universal Declaration on Human Rights, the International Covenant on Civil and Political Rights (‘ICCPR’) and a range of regional conventions and declarations,13 encompasses the right to hold, express and receive a very broad range of views and ideas, including those that are unpalatable, controversial and shocking.14 It is further clear that though not absolute the right to freedom of expression may only be restricted in very narrow circumstances.15 In all cases restrictions must be enacted in accordance with law, serve a legitimate purpose and be necessary to achieve that identified purpose.16 In view of these high standards, imposing civil or criminal penalties on people expressing extremist views and ideas is potentially very problematic. In particular, it is doubtful whether any sufficiently precise definition of extremism can be constructed to satisfy legality; though governments are empowered and expected to protect their people from harm it is debatable whether curtailing extremist speech serves any legitimate objective; and, the need to eliminate a portion of free speech to protect at all is very much open for discussion.

The UN Special Rapporteur on the Promotion and Protection of Human Rights and Fundamental Freedoms while Countering Terrorism (‘UN Terrorism Rapporteur’) re-stressed in early 2016 the need for legislation pertaining to extremism to comply with international human rights law.17 However, whether or not it can theoretically do so presently remains an open question. This paper addresses that gap, investigating the considerations and concerns that domestic legal reliance on the concept of extremism raises in connection with international human rights law. It examines extremism in connection with the right to freedom of expression and asks from a theoretical perspective whether domestic legislation restricting extremist speech does or can comply with the tripartite test for restrictions set out in Article 19(3) ICCPR. Systematic analysis exploring the facets of extremism explicitly within the framework of international human rights law has not to date been undertaken; this paper’s detailed exploration of the legal facets of extremism consequently informs and enriches both prior and future related discourse on counter-terrorism, counter-extremism and freedom of expression generally.

The paper’s structure reflects the Article 19(3) criteria: the first section investigates the compliance of extremism legislation with the principle of legality, the second scrutinises the legitimate purposes extremism legislation might serve and the third queries whether or how extremism legislation can satisfy the strict test of necessity. As a whole, the paper adopts a black-letter interpretive approach, testing the fit of the content and rationale of domestic laws referencing extremism with the established precepts of international law. In view of the nuanced legal, political and socio-legal dynamics involved in extremism, the paper draws upon an array of multi-disciplinary sources to interpret international law and inform its analysis. Further, as the relationship between national and international law can be more easily analysed in abstract when referring to a specific legal order, the paper grounds its theoretical discussion by using UK definitions and documents to dynamically exemplify the real-world application of its arguments and conclusions. The UK is selected for illustrative purposes because it has a detailed counter-extremism policy which is already acting as a benchmark for other liberal States18 and its international political standing together with the intense deliberations of its legislative process may particularly give its Counter-Extremism Bill (if and once enacted) a wider legitimising and emboldening impact in both liberal and proto-liberal States.19 Though primarily referencing a European jurisdiction, the paper similarly chooses to frame its analysis within Article 19 ICCPR rather than Article 10 of the European Convention on Human Rights (‘ECHR’) in order to highlight the worldwide application of the paper’s discussion and conclusions.20

It is acknowledged that the State approaches, academic literature and the UN’s discourse on security, terrorism and extremism on which it relies are generally infused by terrorism and extremism as ‘Islamist’. However, as it unpacks the layers and facets of extremism this paper takes care not to limit its analysis to any particular extremist ideology. Referring variously to ‘extremist speech’ and ‘extremist expressions’, the paper at all times adopts the definition of ‘expression’ formulated by the UN Human Rights Committee in its General Comment No. 34.21 It is assumed that all forms of extremist expression encompass a political element, even if modally amounting to creative endeavour.

The Challenge of Legally Restricting Extremist Expressions

Ascertaining whether domestic legislation which prohibits under civil law provisions or criminally penalises public expressions of extremism does or could comply with the requirements of Article 19(3) ICCPR first requires asking whether such laws are or could be enacted ‘in accordance with law’. Crucially, though the three limbs of Article 19(3) collectively concern the rule of law, discussion of the object of and need for fundamental rights restrictions a priori requires the terms of restrictive legislation to be internally lawful.

Legality Under The ICCPR

It is well-established in international law that laws must be clear, accessible and formulated with sufficient precision to enable individuals to regulate their conduct accordingly.22 It contravenes the principle of legality to prohibit or sanction conduct without a clear definition of what exactly the law is targeting.23 The level of precision required of laws depends on the type of law in question, the field it is designed to cover and the number and status of those to whom it is addressed.24 Laws must not be arbitrary or unreasonable and should contain adequate safeguards against the illegal or abusive imposition of restrictions on fundamental rights through their implementation.25 Legality is further measured by how effective laws are in protecting the right to non-discrimination: laws that inherently discriminate breach Article 26 ICCPR.26 In respect of freedom of expression, unclear legal restrictions have a particularly pernicious effect, with individuals tending towards excessive self-censorship in the absence of clarity over where exactly the boundary to illegality lies.27 The fundamental importance of free speech for democracy therefore demands that laws infringing upon this right be drafted particularly narrowly and precisely28 and that restrictions be enshrined in positive legislation.29

The ICCPR’s structure closely aligns the general protection of freedom of expression under Article 19 with the protective directive of Article 20(2), making this provision equally important in any discussion on freedom of expression. Article 20(2) puts States under an active duty to prohibit ‘advocacy of national, racial or religious hatred that constitutes incitement to discrimination, hostility or violence’. Jurisprudence of the European Court of Human Rights (‘ECtHR’) has clarified that there is a dynamic component to ‘incitement’. It is not enough to express accord with a hostile, discriminatory or violent worldview, even passionately; to amount to ‘incitement’ statements must positively intend to encourage or solicit other people to engage in hostility, discrimination or acts of violence.30 Despite Article 20(2) definitively requiring law-making, domestic measures implementing its strictures must equally comply with the tripartite test for permissible restrictions set out in Article 19(3).31

Further, whether seeking to restrict speech under Article 19(3) or Article 20(2) a close nexus between the relevant expression and a risk of it causing harm is vital for legality. The UN Rabat Plan of Action32 recommends that in assessing whether speech should be criminalised by States as ‘incitement’ regard should be had to the general context, the speaker, his/her intent, the content of the message, its form, the extent of the speech at issue and the likelihood of harm occurring, including its imminence. These factors are also generally useful as a point of reference when defining ‘harmful’ speech. To determine whether speech advocates hate the Rabat Plan encourages States to refer to ARTICLE 19’s Camden Principles, which define hatred as ‘intense and irrational emotions of opprobrium, enmity and detestation’.33

Drawing this guidance together in the context of extremism, it may be briefly concluded that legally restricting extremist free expression requires constructing a narrow definition of extremism that identifies precisely what it entails and links it causally to hate and/or violence yet does not target, either deliberately or inadvertently, any particular group or individual. That established, this paper now questions whether and how formulation of such a definition could be possible.

Legislating Against Extremism

As identified above, legislation restricting extremist speech needs to refer to a definition of extremism which targets with precision an identified harm. Terminology is important not only to satisfy legality; language chosen here also impacts on the second and third limbs of Article 19(3). The objective that restrictions on extremist speech may be said to serve and the need to restrict extremists’ freedom of expression will both be affected by the way in which extremism is defined. Put another way, it may be extremely difficult or impossible to satisfy any of the limbs of Article 19(3) if ‘extremism’ is obliquely framed in porous and multitudinous terms.

The UK Joint Human Rights Committee emphasised in their scrutiny of the Counter-Extremism Bill that extremism is a tricky concept to pin down.34 However, politically, domestic and international approaches to extremism are predicated on it being a necessary precursor to terrorism. Consequently, to assess what, if any, terminology can accurately capture the threat of extremism without also being illegitimately broad, vague or discriminatory, this paper first seeks to understand how contemporary conceptions of extremism have come to be permeated by an enduring link between extremists and terrorists.

Extremism and Terrorism

A gradual but inexorable shift in State perceptions of terrorism since September 2001 transformed the ‘terrorist’ into a morally charged being. Although the term ‘terrorism’ had held a pejorative connotation long before 9/11,35 it was until that devastating day mainly engaged simply to describe a tactic used by political militants vying for regime change.36 It took the advent of ‘war on terror’ parlance to fully embed ‘terrorist’ as a marker of identity and imbue the title with truly foul connotations.37 To define an act as one of terror is now to condemn it as utterly immoral.38 Persons labelled as terrorist are largely characterised as evil39 and the label has permanence: once categorised, terrorists are deemed so regardless of and in spite of any future action, inaction or voiced opinion.40

The post-9/11 transformation of terrorism from (broadly) an illegal act committed by an actor with no particular moral status to an immoral act committed by an immoral actor first enabled and later encouraged counter-terrorism strategies to extend beyond foiling terrorist plots and delve into the ‘root causes’ of terrorism.41 Additionally, once conceptually terrorism became about ‘bad people’ it was a relatively straightforward shift for counter-terrorism discourse to come to regard terrorists as the maturation of ‘extreme’ people; the violent end result of people intensely imbibing ‘extreme’ doctrine. A more nuanced approach to terrorism taking account of the array of political and socio-economic factors feeding into the phenomenon has eventually gained a certain amount of ground in liberal States42 but initially the importance and impact of wider circumstances as markers of the likelihood of terrorism was generally diminished amid the charged political atmosphere post-9/11. The ‘problem’ of extremism was awarded a place of predominance in security discourse which it has always since retained.43

Post-9/11 embedding of extremism underpinning terrorism also linked extremism with ‘radicalisation’. Focusing on the dangers of extremist ideologies supposes that individuals develop extremist political views then seek to put them into practice. It emphasises individual choice in taking violent, terrorist action and implies that individuals ‘radicalised’ through the process of receiving and adopting extremist views voluntarily join terrorist movements, finding in them ideological congruence, rather than being recruited or coerced.44 The aim of post-9/11 ‘counter-extremism’ was thus conceived as being to prevent ‘radicalisation’, cutting the threat of terrorism off at its source. ‘Counter-extremism’ measures intruding on civil freedoms were simply cast as eliminating sooner the risk of violence. Nonetheless, despite the rise and continuance of a perceptible international fixation on extremism and ‘radicalisation’ as the crux of terrorism the validity of any direct linear progression from expression to action has been and continues to be stridently critiqued. Terrorism studies freely admit the difficulties in explaining terrorism with reference to any singular concept,45 the UN Terrorism Rapporteur in early 2016 again decried the existence of a ‘fixed trajectory’ to radical violence46 and an (albeit isolated) recent study has, slightly controversially, indicated that experience of ordinary criminal violence is a more significant indicator of future terrorist action than holding or promoting any extremist ideology.47 Honing in on extremism in terrorist theory thus largely or effectively ignores the ‘complex picture of social, cultural, economic, political and psychological factors which all converge in a negative dynamic to produce those who commit terrorist acts’.48

Consequently, though extremism may be politically viewed through the prism of terrorism, as a matter of law Article 19(3)’s requirement of legality is not automatically thereby met. Even accepting that empirical evidence points to extremism as being a partial contributor to terrorism, expressions of extremism generally include no direct calls to terrorist action; these would in any event fall within the strictures of Article 20(2) ICCPR and be prima facie outlawed. In fact, when recalling the particular importance of narrowness where restrictions upon freedom of expression are concerned, it is arguable that only expressions of extremism meeting the threshold of Article 20(2) do or could comprise the causal nexus to harm necessary for legality: in the absence of linear progression from speech to action, without incitement even the most vitriolic statement of extremism is arguably too far removed from any eventual violence for legality to be satisfied.

Nonetheless, liberal approaches to counter-extremism implicitly reject any notion of ‘non-violent extremism’. Despite blind assertions to the contrary, liberal States’ counter-extremism is underpinned by an assertion that even if ‘extremists’ do not always become terrorists, extremism ‘can create an atmosphere conducive to terrorism and can popularise views which terrorists then exploit’.49 Exemplifying this, the UK Counter-Extremism Strategy identifies extremists as ‘carefully portray[ing] violence as inevitable in achieving the desired end state required by their ideology’ and/or using rhetoric characterised by violent language.50 This indirect causal connection between extremism and terrorism could potentially provide a definitional nexus with harm sufficient to satisfy Article 19(3)’s legality criteria. Schmid, having researched extensively into terrorism and extremism definitions, considers it naïve and dangerous to assume that one can hold extremist beliefs but not be inclined to use extremist (i.e. violent) methods to realise them when the opportunity presents itself: he argues that even ostensibly not-violent organisations can act as ‘stepping stones’ to violence as both are ultimately party to a common agenda.51 Extrapolating from the theory of fear speech posited by Buyse,52 which notes that conflicts escalate when parties use both speech and action in tactics of increasing harshness, it is feasible to reason that extremist expressions referring to violence, especially where underpinned by values running counter to the tenets of liberal democracy, could have a ‘conflict escalating’ effect, increasing the likelihood of terrorist violence occurring. Indirect causality must be approached with caution though: the more stretched the link between extremism and terrorism becomes, the greater strength there is to arguments that only where incitement exists and speech becomes categorised under Article 20(2) might legality be satisfied.

Furthermore, even if extremism can theoretically be legally related to terrorism, the connection depends upon a fixed conception that the key marker of extremist ideologies justifying restriction of fundamental rights is anti-liberal sentiment (consider the UK’s contrasting of extremism with ‘British values’ – a phrase open to question and interpretation but reasonably understood generally to encompass the underpinning values of liberal society); however, as a stand-alone term extremism is a nebulous concept embracing a host of non-mainstream views. Notably, use of the term ‘ecological extremists’ has increased exponentially with the rise of conservation activism to describe the views held by a certain cohort of environmental and animal rights protesters. These political ‘extremists’ actively embrace the label for their own reasons but though they may be ‘extremists’ in the sense that they are arguably equally ‘absolutists, locked in on their particular viewpoint, largely incapable of understanding, if not intolerant, of other perspectives’ and ‘having an ideology that is ‘the truth’ to be defended at all costs’,53 in pursuit of their goals ‘ecological extremists’ neither aim to destroy the democratic order nor advocate hate or support violence – except perhaps in terms of property destruction.54 Particularly where seeking to restrict expressions broader than incitement, therefore, the term ‘extremism’ needs careful further explication to ensure laws do not inadvertently net ‘extremists’ such as these whose ideas and motives have no connection to the scourge of global terrorism.

Defining and Codifying Extremism

Even within the terrorism paradigm, the question remains: can any satisfactorily narrow and targeted legal definition of extremism be formulated to meet Article 19(3)’s legality criteria? Extremism has no agreed definition in international law. In fact, despite the term ‘violent extremism’ forming the crux of the UN General Assembly’s 2015 ‘Plan of Action to Prevent Violent Extremism’ this document rather worryingly opens by asserting that even this arguably more limited formulation eludes definition.55 The addition of ‘violent’ to extremism provides no de facto clarity. Delegating to States the responsibility for constructing individual definitions of extremism, the Plan of Action gives no guidance on the substantive content to extremism, merely reminding States that they must enact law that complies with international human rights standards.56 Reiterating this warning, representatives from the UN, the Organization for Security and Co-operation in Europe (‘OSCE’), the Organisation of American States (‘OAS’) and the African Commission on Human and Peoples’ Rights (‘ACHPR’) jointly stated in May 2016 that unless given an appropriately exact definition the concept of extremism should not be used as the basis for restricting freedom of expression.57 This supplemented an earlier joint statement stipulating that domestic laws should not employ such vague terms as ‘glorifying’, ‘promoting’ or ‘supporting’ extremism.58

The UN and regional bodies may legitimately be trying to respect domestic sovereignty in refusing to codify extremism in international law and/or proffering no detail on its substantive content, but for States seeking to ensure that extremism legislation is enacted and implemented in compliance with international law the lack of international clarity on what is meant when referring to extremism is both unhelpful and problematic. ‘Success’ in fighting the global threat of terrorism through employing extremism restrictions is fatally undermined by domestic definitions of extremism that are ill-harmonised; formulated and applied disjointedly according to what States discretely perceive as threatening. A joint NGO submission to the UN in February 2016 outlined severe concerns that State-generated definitions of extremism presently capture angry but non-threatening expressions of unpopular ideas and government criticism along with truly dangerous expressions of bigoted and violent ideologies.59 An international-level definition of extremism would thus be both appropriate and beneficial: material specification of universally applicable markers of extremism in international law could vitally help to prevent abusive local language and align domestic laws that are, at least ostensibly, premised on targeting a singular, globalised harm.

Whether or not defining at an international level, however, translating extremism into adequate and appropriate legal terminology still presents a Herculean task.60 The definitional issues highlighted by NGOs in their joint submission to the UN61 are not limited to authoritarian States; as noted above, even the reference to ‘British values’ in the UK definition of extremism is potentially problematic, becoming imbued with different content depending on who is articulating the phrase, where and when, as well as arguably lending itself to allow an interpretation of extremism entirely devoid of any ultimate connection to violence. Some NGOs have opined that the very inability to articulate clearly and precisely what unlawful extremism encompasses evidences that legislation curtailing it has ‘no place’ in liberal democracy.62 Certainly, with each stage of scrutiny yielding further questions and with litigation challenges to the Counter-Extremism Strategy now also being mounted63 it seems less and less likely that the UK Counter-Extremism Bill will ever pass into law.

One scholar does stand out as having proffered a list of extremism traits that could potentially found a rule-of-law compliant definition. Etymologically, to be ‘extreme’ is to be furthest from the centre.64 This implies a normative assessment; a benchmark against which views, opinions and ideologies are judged.65 Extending his view of extremism as intrinsically violent, Schmid suggests that the key need in narrowing extremism is to distinguish it from other political speech and behaviour that does not pose a fundamental threat to liberal democracy. Thus he identifies twenty specific markers of ‘extremists’.66 Whilst caution must be exercised in relying on these markers since context can shift the benchmark, Schmid’s list is very thorough and, positively, hones in on extremist speech that betrays a violent mindset aligned with those who commit terrorist acts. A definition of extremism adopting Schmid’s markers in their entirety arguably has the potential to be very precise indeed, targeting a narrow range of harm.

Even use of Schmid’s list, though, does not address the matter of discrimination. The NGO joint submission to the UN referenced above additionally identified that extremism definitions frequently either inadvertently encompass or are being deliberately used to suppress unwanted democratic debate, dissent and minority opinions.67 This echoes Post’s contention that laws claiming to be objectively justified as preventing ‘hate speech’ are in reality being deployed to enforce ‘norms of propriety’ in matters such as race, nationality and ethnicity.68 Whilst targeted application is highly problematic it does not of itself negate legality; however, laws containing inherent risks of discrimination are fatally undermined, and evidence suggests that the risks of discrimination inherent in extremism definitions cannot be cured through ‘inclusivity’. The Global Counter-Terrorism Forum’s Ankara Memorandum 2014 recommends that to prevent discrimination States should avoid identifying extremism with any particular religion, culture, ethnic group, nationality or race.69 In line with this, and in a manner typical of the liberal agenda, the UK asserts that it targets ‘the full spectrum of extremism: violent and non-violent, Islamist and neo-Nazi – hate and fear in all their forms’.70 However, though prima facie non-discriminatory, this inclusionary statement is equally disturbingly wide-ranging. Kundani expresses fear that, coupled with sweeping terms positively defining extremism, it leaves troubling scope for manipulation of the term extremist and could well result in the ‘demonisation’ of anyone holding radically fringe views, labelling them as having a ‘fanatical mindset’ rather than just a ‘possibly misguided opinion’.71 A further testament to the potential that ‘inclusive approaches’ hold for widespread repression is the ‘unlikely coalition’ of individuals and groups in the UK holding disparate and even opposing views that have joined forces to campaign implementation of the planned extremism legislation.72

Therefore, and worryingly enough, given the extent of extremism policy and legislation existing and pending worldwide, this paper cannot conclude with any confidence that a definition of extremism compliant with international law’s principles of legality can be constructed. In order to proceed, therefore, the remainder of this paper lightly and hesitantly assumes that with care and attention to detail it might theoretically be possible for domestic legislation restricting free expression of extremist views to satisfy Article 19(3) ICCPR’s criteria of legality.

Legitimate Reasons to Restrict Expressions of Extremism

Assuming extremism legislation to be legal, this section now considers whether extremism legislation does or could serve any legitimate purpose. Article 19(3) ICCPR provides for two umbrella reasons to restrict free expression: protecting national security, public order, public health or morals; and, respecting the rights or reputations of others. As identified above, liberal States enacting extremism legislation claim that it links to terrorism: the UK explicitly identifies extremism as posing an intense and palpable threat to life and limb, democratic values and the fabric of society.73 Two parts of the umbrella are therefore potentially applicable: protecting national security and respecting the rights of others. Taking each purpose in turn, this section unpacks the relevant international law and evaluates whether extremism legislation does or could conceptually fit.

Protecting National Security

Determining whether extremism legislation protects national security comprises three stages: First, it must be ascertained how ‘national security’ is defined within the ICCPR and what is required under the ICCPR for national security to be legitimately invoked. Second, these identified legal standards must be applied to delimit if and how expressions of extremism in fact threaten national security. Third, providing extremism does in law and fact pose a national security threat, it must be established whether measures restricting it are in practice genuinely directed at countering that threat.

‘National Security’ under the ICCPR

The first step towards evaluating whether the alleged threat posed by extremists’ exercise of their freedom of expression is a matter of national security is to define a national security threat. There is, however, at present no definitive international jurisprudence on what constitutes ‘national security’: respect for sovereignty grants States a wide discretion to self-define national security threats and pursue individually tailored security policies.74 This lack of international definition is somewhat alarming: the UN Special Rapporteur on the promotion and protection of the right to freedom of opinion and expression expressed severe concerns in his 2013 report that an ‘amorphous’ concept of national security is ‘vulnerable to manipulation’ by States wishing both to invade upon fundamental rights and cloak their actions in secrecy.75 Though written in the context of surveillance, these concerns have equal relevance to extremism: problems of rule of law compliance with vague notions of extremism are compounded when ‘national security’ is equally impossible to pin down. Helpfully, however, the UN Human Rights Committee (‘the Committee’) has taken a critical approach towards States invoking national security, robustly and repeatedly requiring States to identify the precise nature of any asserted security threat and expressing its disapproval of un-evidenced claims in striking tones.76 This positively indicates that though States may not be confined to any single conception of ‘national security’ under the ICCPR they still need to produce strong evidence to back up any recourse to this as a legitimate aim.77 For states to link extremism with national security, more is needed than assertions that extremism-equals-terrorism.

Nonetheless, at present ‘national security’ tends towards a catch-all phrase, used to cover both internal civil disturbances and external, global threats to sovereignty. Troublingly, such breadth facilitates deliberate or inadvertent escalation of local threats to a higher plane, inherently justifying greater severity of measures in response and directly impacting on the appropriate probative force of questions over necessity under Article 19(3)’s third limb. Soft law guidelines however offer precise and narrow definitions of ‘national security’ and a strengthened application of these would be welcomed. The UN Siracusa Principles provide that national security may only justify rights-restrictive measures where these are taken to protect ‘the existence of the nation or its territorial integrity or political independence.’78 They also provide that national security cannot be used as a pretext to impose vague or arbitrary limitations on fundamental rights.79 This means that illegitimately expansive definitions of extremism cannot be remedied purely by oblique references thereafter to national security. The Johannesburg Principles80 further specify that national security may only be used to limit freedom of expression where there is a ‘direct and immediate connection’ between expressions intending to incite violence and the likelihood of or occurrence of such violence.81 Nowak’s CCPR Commentary also identifies ‘national security’ as a ‘political or military threat to the entire nation’, including ‘publication of a direct call to violent overthrow of the government in an atmosphere of political unrest’.82 Under these definitions, extremism only indirectly linked to violence could not be subsumed into the murky waters of national security.

It is rational and coherent to limit ‘protecting national security’ to ‘preserving the safety of State territory and protecting the integrity of its vital institutions from violent action’; adoption of a definition comprising these criteria by both States and international bodies could vitally promote transparency and accountability in State practice. Nonetheless, under the ICCPR as it stands States may apparently invoke this objective against any perceived threat provided they can justify their reasoning. This being the case, the next question to ask is; do extremists pose a national security threat?

Extremists as the Enemy Within

This paper accepts that extremism may be linked to terrorism. However, without more, this ultimately posits terrorism as the threat to national security with extremism a mere warning signal. To ascertain whether extremism could be identified as a discrete national security risk, this section reflects on an idea independently but concurrently raised by three counter-terrorism scholars; namely, that society has in recent years returned to a paradigm whereby States use their power to seek out and oust the ‘enemy within’.

Clavell, Lomell and Galli each powerfully argue that post-9/11 counter-terrorism approaches have revived a dormant pattern of States perceiving the enemy to be not only abroad but also ever-present at home.83 Lomell and Clavell further suggest that this ‘enemy within’ model re-arose in society post-9/11 because that event eroded the distinction between ‘national’ and ‘international’ security, triggering a global move towards philosophical and operational security approaches that construct citizens as enemies.84 This notion has historical credence: in ages past liberal States perceived and rejected, for example, communists as enemies existing within society; ‘extremists’ are merely the contemporary iteration of this phenomenon. Further, political positioning of extremism as a thread in the interlinked global threat of terrorism significantly blurs security remits: domestic laws address matters within sovereign borders but local counter-extremism measures are enacted in full awareness that internal expressions of extremism might causally result in harm abroad, thus taking into account both domestic and international security considerations.

Clavell uniquely constructs a conceptual narrative of how ‘extremists’ became ‘enemies within’ liberal States. She notes that in post-1980s Britain seemingly progressive and liberal models of crime prevention provided the logic for more intrusive and less tolerant uses of State power. Pervasive ideas about ‘prevention’ led policing to negatively become about monitoring and extinguishing ‘deviant’ behaviour; marking out ‘others’ in society. Previously unwelcome fringe conduct was increasingly deemed completely intolerable. Through this process, Clavell suggests, ‘extremists’ as ‘deviants’ were inexorably converted from holders of odious viewpoints into downright dangerous beings. Clavell further contends that post-9/11 there has been an amplified blurring of the line between local fears and foreign threats, leading to an obfuscation of the remits of criminal law and defence policy in ‘the drive to fortify cities as well as borders against powerful, unpredictable global actors bent on destruction’.85 She reasons that this domestic/international boundary confusion redefined ‘deviants’ as posing not only a local but an international risk.86 Extrapolating from this theory, extremists became enemies; no longer ‘mere’ domestic criminals but a more all-encompassing national security problem.

Clavell’s assessment of State action as illiberal creep may be disputed: liberalism arguably contains a positive obligation to protect society from its destruction, which would under Clavell’s theory be the result were ‘enemies within’ not identified and dealt with. However, taken together with Lomell and Galli, Clavell’s historical narrative does lucidly explain how extremism may legitimately be found to threaten national security. Even referencing the narrow conception of national security set out within the Siracusa and Johannesburg Principles there can be no dispute that enemies by their very nature target the institutions of government and the existence of civil community. Consequently, once secured in position as ‘enemies’ the security rationale underpinning any action taken against extremists would be firmly in place.

This is but one theory, though, and the potential practical ramifications of acting against extremists under a conception of their being ‘enemies within’ are concerning. Enemies are by their very nature outside society, not entitled to its protection and a risk to its existence. Regarding extremists as enemies therefore immediately heightens the appropriate level of response to their presence, which has a significant impact on the intensity of review conducted under Article 19(3)’s third criteria of necessity.

National vs Individual Security

‘Enemy’ or not, there still remains the question of counter-extremism’s intention: is it legitimately targeting a defined national security threat or illegitimately aimed at assuaging a general ongoing sense of insecurity? Domestic and international laws and policies appear increasingly focused on the goal of improving ‘individual security’ (subjectively defined feelings or perceptions of safety) rather than pursuing objectively defined national security goals. Murphy, for example, described UK counter-terrorism policy as ‘exhibiting characteristics of both risk-based government and moral panic’.87 Buhelt more broadly employed the beguiling phrase ‘liquid fear’ to describe the imprecise and unclear dread with no clear address or cause that he posits has become a decisive factor in national and international security policy-making since the singular organisation identified in the initial aftermath of 9/11 has been revealed to be a more diffuse and scattered network.88 Clavell further theorises that this ‘liquid fear’ coupled with enlarged definitions of security which identify myriad risks and threats both locally and globally and have populaces demanding ever more ‘security’ to combat the ‘insecurity’ has distorted the meaning of national security, subsuming it within or subverting it to individual security.89 The ‘enemy within’ construct could thus be both masking and fueling State security aims that are not about securing the nation at all but rather merely tap into a general feeling of peril that has been rising as a political power since September 2001. If this is so, even the broadest interpretation of ‘national security’ under Article 19(3) ICCPR might not be satisfied.

The criminological concept of ‘moral panic’ particularly assists in explaining how legal distortion of ‘national’ into ‘individual’ security can occur. A term first developed and popularised by Cohen probing the societal, political and legislative response to public disturbances by young people in a seaside town in 1960s Britain, ‘moral panic’ describes a situation whereby public fear and State intervention grossly exceed an objective threat posed to society.90 Cohen theorised that ‘moral panic’ does not happen spontaneously but is the result of a complex interplay of actors exaggerating or distorting negative behaviours (real or perceived) of certain individuals or groups. Media framing heightens the deviancy or criminality of identified ‘folk devils’, presenting them as the embodiment of evil and making them the antagonists in the ensuing public panic and State response. Political rhetoric positions the State as holding the moral high-ground against this evil, fueling public agitation, and as rule of law enforcers are expected as part of their general remit to detect, apprehend and punish evil, nebulous panic solidifies into disproportionately punitive legislation. Cohen’s theory was originally limited to a relatively minor clash of a youth subculture with authority but was subsequently expanded and applied to other manifestations of ‘deviant’ behaviour,91 and, as with extremism, his ideas had new life breathed into them post-9/11. Examining the USA response to that security breach and tragedy, Rothe and Muzzatti described the US as a culture ‘embedded in fear’ and concluded that State presentations of the threat of terrorism had generated a ‘moral panic’ according to Cohen’s precepts.92

It is arguable that contemporary presentations of ‘extremists’ equally subscribe to Cohen’s model. ‘Moral panic’ is surely complemented and exacerbated by the paradigm of the ‘enemy within’ which transforms deviancy into an active threat, increasing tension between individuals and groups. As the State wages war on its own citizens it shifts from defending responses to concrete threats to protecting more diffuse matters such as ‘social cohesiveness and identity’.93 This in turn fuels negative public perceptions of extremism, leads to calls for ‘more security’ and results in legislation borne out of ‘liquid fear’. Cohen’s theory admittedly only has partial applicability to extremism: a ‘moral panic’ in socio-criminal theory only exists to the extent that the public press for State action against the ‘folk devils’ yet the extent of public backlash in Britain against the impact of the Government’s extremism policies94 indicates that any panic about ‘extremists’ is mainly pushed from a State level, not wholly against public sentiment but certainly without its complete blessing. Still, when States invoke national security as the ‘legitimate’ aim in extremism legislation Cohen’s theory does go some way towards supporting the possibility that this is a fallacy; States are in truth pursuing visions of individual security.

Ultimately, given the legal scope to claim almost any end as national security, the link from extremism to terrorism and the conceptual placing of extremists as ‘enemies within’, it may be concluded that extremism legislation could potentially serve national security ends within the terms of Article 19(3) ICCPR. However, the word ‘potentially’ is key. National security carries deeply embedded connotations of sovereign prerogative. A certain level of transparency is vital in order to prevent the rule of law sliding into rule by executive dictate. Accepting without further enquiry that legislation restricting free speech is enacted because extremism is a danger to national security may subconsciously negatively impact the intensity of review engaged in when assessing the necessity of any such legislation. To guard against this, States must justify not just the extent of their actions but also the underpinning rationale.

Respecting the Rights of Others

The second purpose that extremism legislation could potentially serve under Article 19(3) ICCPR is respecting the rights of others. This is premised on extremists’ freedom of expression conflicting with rights held by others in society. Balancing competing rights against one another is a core component of human rights practice; a plethora of domestic and international jurisprudence exists giving preference to one individual over another and identifying various factors given weight in the exercise. Despite its legal prevalence, however, the ‘balancing exercise’ is fraught with difficulty and controversy. Two central questions have relevance for extremism legislation: which ‘rights’ are being respected and who are the ‘others’ concerned?

‘Rights’ under the ICCPR

The ICCPR contains no clear definition of a ‘right’.95 The UN Siracusa Principles specify that ‘when a conflict exists between a right protected under the ICCPR and one which is not, recognition and consideration should be given to the fact that the Covenant seeks to protect the most fundamental rights and freedoms’.96 This statement accepts that ‘rights’ may exist outside the bounds of those set down in international treaties. UN General Comment No. 34 could possibly be read otherwise. It specifies that ‘rights’ ‘includes human rights as recognized in the Covenant and more generally in international human rights law’.97 On one view, this limits rights to those those explicitly delimited by existing legal documents. However, jurisprudence of the European Court of Human Rights (‘ECtHR’) forms part of ‘international human rights law’ and comprehensively reviewing ECtHR case-law on balancing conflicting rights Bomhoff98 notes that the ECtHR takes a wide understanding of ‘rights’ extending far beyond those codified in the ECHR. Highlighting particularly the decision in Otto-Preminger-Institut v. Austria,99 which protected a ‘right to religious feelings’, Bomhff concludes there is currently next to nothing that would likely be excluded by the ECtHR as a legitimate ground for the limitation of fundamental rights.100 In view of the lack of boundary on the category of ‘rights’ set by the ECtHR and the close linguistic corollary between the ECHR and the ICCPR101 it may be concluded that ‘rights’ may be equally broadly defined under the ICCPR.

Such a sweeping approach has potentially devastating implications. Tsakyrakis102 forcefully argues that international law’s ‘definitional generosity’103 risks losing sight of the moral message conveyed by demarcating only certain matters as ‘fundamental rights’.104 Tsakyrakis contends that human rights are not merely ‘quantities of freedom’ but protect people as ‘moral agents’.105 Consequently, he argues, jettisoning any notion of quantifiable ‘rights’ and allowing any interest or consideration to compete as if a ‘right’ in a balancing process fatally compromises that which was originally denoted by lawmakers as being so important.106 Bomhoff too is concerned by this: though he fails to identify any limit to the term ‘rights’ he insists that one must nonetheless implicitly exist as too broad an application of ‘rights’ risks undermining the integrity of balancing as an enterprise.107 This is a reasonable anxiety. Definitional flexibility can be positive, enabling courts to take account of evolving understandings of rights and provide better protection to vulnerable individuals who don’t exactly fit within established categories. Flexibility of approach towards balancing is also arguably useful in enabling economic, social and cultural rights to be openly and fairly weighed against civil and political rights despite these being separated within the international legal order. But where ‘rights’ begin to be effectively created entirely outside established treaty frameworks the notion of a delimited set of ‘rights’ guaranteed to all individuals because of our common humanity – upon which all of international human rights law rests – worryingly begins to crumble.

Nonetheless, as the law stands, in the context of extremism it is possible that something as vague as a ‘right to feel secure’ could be used legitimately by States to restrict free speech. In fact, the marginally more concrete ‘right to individual security’ parallels the emerging developmental idea of a right to ‘human security’, which emphasises individual welfare in the right to security provision by States.108 These ‘rights’ in place therefore, the next stage is to define the ‘others’ holding them who are in need of protection.

Identifying ‘Others’

Not only is there no limit to the ‘rights’ which may be at stake, no set criteria exists to identify the ‘others’ affected. Crucially, States must not through defining ‘others’ simply or inadvertently pit one extremist ideology against another or illegitimately protect ‘moderate’ individuals from feeling threatened by those espousing different, unwelcome ideologies. Certainly, it would seem illegitimate to regard ‘others’ as those liable to have their voice ‘drowned out’ by ‘extremists’: part of the right to individual security is arguably cultivation of and protection by the State of a strong political sphere, robustly able to entertain diverging perspectives and counter-cultural viewpoints. In view of Schmid’s view above that the central marker of extremism should be anti-liberal content, perhaps the ‘others’ could reasonably be simply all those who are members of a liberal democracy. This fits particularly well for Article 19(3) in light of the fact that the ECHR exclusively deals with rights as held by individuals but under the ICCPR the term ‘others’ may relate to persons individually or as a group.109 Construing the issue as the harmful influence of extremism on society at large, the relevant ‘right’ becomes a community right to a secure civil society space. This being so, the ‘others’ could be regarded broadly as all within the community seeking access to that space.

Thus it is straightforward, if instinctively troubling by its very simplicity, to find that extremism legislation serves the legitimate purpose of respecting the rights of others. However, if ‘rights’ and ‘others’ can be anything it becomes exponentially more imperative for extremism to be tautly defined. Thankfully, the ease of passing this second hurdle within Article 19(3) is tempered by the limiting nature of Article 19(3)’s third criteria of necessity, to which this paper now turns.

Need to Restrict Extremist Expressions

Having concluded that legislation restricting extremist speech could potentially be legitimately directed at protecting national security or respecting the rights of others this section addresses the final criterion of Article 19(3), questioning whether such legislation is needed to achieve either of these aims. To do so, it is vital first to unpack how ‘necessity’ is defined and applied under the ICCPR to adjudicate the legitimacy of restrictions on qualified fundamental rights: though it is a cornerstone of human rights law that interferences with qualified rights may extend only so far as is ‘necessary’ this phrase requires some clarification.

Necessity under the ICCPR

The established stricture of international human rights law dictating how far States may restrict non-absolute fundamental rights is simply that measures imposed in pursuit of a legitimate purpose must not be more restrictive than are ‘necessary’ to achieve that purpose. It is widely agreed that necessity is an adaptable standard that takes account of both individual circumstances and the nature of the fundamental right at stake; however, what precisely necessity entails and how it is or should be practically applied is not altogether straightforward.

Necessity is firmly twinned within the ICCPR with the notion of proportionality. The UN has specifically affirmed that Article 19(3) ICCPR deliberately intended the proportionality of rights-restrictive action110 to be considered. It has further specified that to be necessary/proportionate, restrictions to the right to freedom of expression must be ‘directly related’ to the need they claim to serve and the ‘least intrusive instrument’ for the purpose, taking into account the form of expression at issue as well as the means of its dissemination. There must be a ‘direct and immediate’ connection between the free expression at stake and the threat its restriction is designed to counter.111 The UN Siracusa Principles also take a very narrow interpretation of ‘necessity’, stipulating as general guidance that States must prove limitations to qualified rights do not ‘impair the democratic functioning of society’.112 In view of the central importance of freedom of expression to democracy, this requirement places States under an onerous obligation with little scope for any restriction to be justified. The UN Rabat Plan of Action further reiterates that free expression restrictions must be targeted and recommends that they should respond to a ‘pressing social need’.113

Aimed at guiding States on how to balance Articles 19 and 20 ICCPR, the Rabat Plan’s principles and recommendations arguably apply irrespective of the provision under which the expression in question is in fact categorised. Thus, two further points made therein are also relevant. First, given that the strength of harm evaluation impacts on the extent to which interruption measures are warranted, the six factors identified above for assessing whether a specific instance of speech ought to be prohibited or punished as incitement logically have broader applicability to assessments of necessity. Second, it is noteworthy that a clear distinction is drawn between three types of expression: expression which constitutes a criminal offence; expression below that threshold but justifying civil or administrative sanctions; and, expression lower again but that raises a concern in terms of tolerance, civility and respect for the rights of others. This sliding scale gives useful scope for nuanced approaches, especially as the Plan advocates viewing criminal sanctions as a ‘last resort’.114

Interpretations of necessity under Article 19 ICCPR are further informed by the regional human rights mechanisms’ robust attitude towards freedom of expression, particularly that of the ECtHR. The UN accords entirely with the long-established ECtHR position that the right to freedom of expression must only be restricted in very narrow circumstances, and the ECtHR has in recent years taken an increasingly protective attitude towards free speech: analysing 35 years of case-law Voorhoof demonstrates how ECtHR judgments now ever-more frequently uphold the right to freedom of expression and regularly emphasise the value for democratic society in tolerance, broadmindedness, pluralism and participation in public debate.115 A limited number of judgments handed down since Voorhoof’s paper do buck the trend he identifies; however, as the majority of these cases concern statements made online there is rationality and strength in suggestions by commentators that the shift in approach may be regarded as attributable and limited to teething discomfort in the ECtHR over the internet’s de facto provision of enhanced free speech opportunities.116 Overall, the ECtHR’s approach indicates that there is very little scope for States to justify restrictions to free speech, particularly to the level of criminal sanction, and suggestions of wider benefits to society flowing from enhanced free expression protection117 adds to the reasons that Article 19(3) ICCPR should continue to be interpreted narrowly.

Assessing Necessity

Concluding that a ‘strong’ standard of necessity is and should be applied where the right to freedom of expression is at stake, however, leaves untouched the question of how exactly how necessity/proportionality assessments are or should be undertaken. Three central decisions of the UN Human Rights Committee concerning freedom of expression and national security relevantly emphasise the impressionistic nature of ‘necessity’ assessments. In Tae Hoon Park the Committee was quite evidently unimpressed by South Korea invoking as a security threat the ‘general situation’ in North Korea and the nature of North Korea’s communist ideology.118 There, as again later in Keun-Tae Kim it stated disapprovingly that the ‘precise nature and extent’ of the alleged threat posed by the applicants’ activities (respectively, membership of a student organisation discussing Korean unification and distribution of political material siding with North Korea) had not been adequately identified, nor was there any evidence in either case of a need for criminal prosecution.119 Some years later, in Jeong-Eun Lee (concerning conviction of a student for belonging to a political association regarded by South Korea as ‘enemy-befitting’) the Committee rather tetchily stipulated that necessity required aversion of a ‘real and not only hypothetical’ danger.120 What may be gleaned from these cases is that the Committee applies a consistently exacting standard of necessity but shows no detailed assessment framework for its decision-making process.

This is unhelpful when attempting to construct an overarching theory of necessity, and indeed generally for transparency and accountability. The link between necessity and proportionality lends itself to a balancing test much as when individual rights are pitted against one another. Building on this, Gerards interestingly has advocated for a novel two-fold test: first examine the suitability of particular means for the ends pursued; then only if the chosen means appear to be both adequate and necessary does balancing come into play.121 Such an approach could positively act to emphasise the need for measures to be underpinned by objective justifications: as the ECtHR has recognised, especially where free expression abhorrent to liberal democracy is concerned there is a vital need to prevent personal conviction from influencing judgments about what is actually dangerous.122 The remainder of this section therefore relies on Gerard’s approach when asking whether extremism legislation does or is ever likely to be able to meet the stringent standards for necessity under Article 19(3), considering first the need to restrict extremist speech to protect national security.

Necessity to Protect National Security

The international law framework established above effectively provides that to comply with Article 19(3) ICCPR’s necessity requirement any restrictions to extremist speech premised on protecting national security need to directly respond to the harm to the State likely to be otherwise inflicted and be at the minimal adequate level of severity. Having examined how extremism links with violence in terrorism, identified national security as being concerned with preventing violence and considered a theory of extremists as ‘enemies within’ presenting a constant and insidious threat to national security, this section now asks to what extent it is or might be necessary to restrict the freedom of expression enjoyed by extremists to serve the purpose of national security.

Risks and Responses

The need to restrict expressions of extremism depends on the extent to which they pose a ‘direct and immediate’ risk to national security. Assuming for present purposes that some need to restrict to protect against extremism can be shown, determining the ‘precise nature and extent’ of any security risk posed by extremist speech, and thus the level of restriction on free expression justified, requires examination of the speech’s content and context. Though each individual situation will be unique, the particularly vocal nature of ‘extremists’ could be a relevant consideration: Schmid notes that ‘extremists’ often actively recruit to their cause; intensity and fervour heightening their agitation and leading them to gain disproportionate prominence.123 When coupled with Buyse’s finding that when one interpretation or discourse gains the ‘upper hand’ and stifles alternative voices the danger of violent escalation is larger,124 this circumstance could be used to justify greater intrusion into extremist speech.

States, however, must be careful that in the vigor of their counter-extremism approaches they do not illegitimately focus on eradicating extremist speech from public dialogue to the exclusion of advancing other equally or more needed security measures. The UN Terrorism Rapporteur’s 2016 Report criticised States for paying attention to only those issues ‘most appealing’ to them and ‘shying away’ from more complex matters.125 States must also not exaggerate risks: arguably, extremism measures should always be wholly exceptional as the small numbers of people that eventually get radicalised and take terrorist action make intrusion into this core democratic right generally disproportionate. Indeed, extending this line of reasoning and echoing the point above that indirect and stretched connections between speech and violent action may mean that only Article 20(2) impositions on extremist speech are justified, domestic legislation directly outlawing and penalising incitement to terrorist acts – a product of States’ obligations under international law to combat terrorism – arguably suffices without extremism legislation at all to cover any risks to security posed by extremists in society.126

Cram and Malik further argue that it is fallacious to assume that curtailing extremist speech negates existence of its underlying ideology and goals. They comment that extremists may be willing to make superficial concessions in service of their aims but ‘moderate’ speech may mask an uncompromising worldview that is shared only with select audiences. Restrictions on extremist speech have no impact on these private extremists, dimming the urgency of restricting fundamental rights to protect.127 Civil sanctions might also be regarded as less severe than criminal penalties but they have the insidious potential to impact on individuals and groups otherwise untainted by the label of extremism: for example, closing a community centre to prevent gatherings of certain extremists will have a knock-on suppressive effect on non-related, entirely innocuous activities. Ultimately, therefore, whilst it might be possible to make a case for some need to restrict extremist speech, the need ever for severe sanctions, particularly to the level of criminal punishment, is highly suspect.

More Speech, not Less

Adding to this, there are strong empirical indications that limiting the range of expression in society might not effectively counter the threats posed by extremism: on the contrary, banning or criminalising public expressions of extremist views might actually heighten the risk of violence ultimately occurring. As a matter of law, if interfering with freedom of expression proves to be unsuitable to achieve the aim of protecting national security even legislation imposing only low-level restrictions on extremist expression might be disproportionate: Gerards argues that in necessity assessments any negative consequences of action for fundamental rights are of vital relevance and measures harming fundamental rights but failing to benefit anyone or to achieve the desired results will not be reasonable, effective or necessary.128

There are four reasons that, at least for some individuals, silencing extremist messages exacerbates rather than alleviates potential damage. First, suppression of extremist views can cause alienation and disaffection: identified factors in the road to radicalisation and terrorist violence.129 Second, a focus on eliminating extremist speech can prevent meaningful engagement with those most at risk of radicalisation130 and take extremist conversations underground, preventing vulnerable individuals from being exposed to contrary views.131 Third, censorship gives credence to claims by certain extremists that one of the most core freedoms of democracy is denied to those who argue against it, increasing the ‘level’ of extremism some individuals will be willing to subscribe to132 - notably, the UN’s ‘Plan of Action’ suggests that extremists ‘provoke States into overreacting then exploit ill-conceived government action for their own propaganda ends’.133 Fourth, repressive action taken by the State against extremist speech feeds two dangerous illusions: that clashing with authority is a ‘testament to truth’, and ‘sinister attribution error’ whereby everything negative is construed as a plot.134

The negative effect of State narratives of extremism leading inescapably to violence which attach to restrictive measures also has a potentially exacerbating effect on the risk of extremist speech escalating to violence. With regard to the UK, Kundnani contends that the official narrative of extremism as an ‘embedded problem’ unhelpfully ‘promotes and reinforces false stereotypes, distorts public discourse and fosters social divisions’, all of which increases the likelihood of violence ultimately occurring.135 Referencing Buyse’s conclusion, actively marginalising people by labelling their voice extremist and banning or criminalising expression of their views in public democratic dialogue is likely only to ‘harden boundaries between groups, instil fear and push individuals towards embracing violent action as a vehicle to advance their access to equality, development and freedom of [ideology]’.136

Cumulatively, therefore, the need to and wisdom of silencing the voices of extremists to secure the nation appears seriously contraindicated. There is strength in the argument that any threat to national security posed by extremism should be fought with human rights, not by compromising them.137 NGOs notably consistently advocate for more rather than less speech; a joint statement criticising the UK’s Counter-Extremism Bill, for example, called for ‘soaring confidence in our values and the society we seek to build together’ and the defeat of terrorism through freedom not fear.138 The UN, OSCE, OAS and ACHPR freedom of expression representatives additionally chose to emphasise in a 2015 joint statement that ‘censorship is not an effective response to extremism’.139 States have positive obligations under international law to ‘ensure that all groups in society have access to opportunities to make their voices heard’140 and to obstruct free speech where it poses a direct threat of terrorism:141 these imperatives could arguably be interpreted broadly and deployed (if individual circumstances warranted) to cover more indirect risk-filled expressions of extremism and prevent harm. In view of this, this paper tentatively suggests that Article 19(3)’s test of necessity is not or is unlikely to be met by extremism legislation premised on protecting national security.

Necessity to Respect the Rights of Others

In light of the above conclusion, this paper finally asks whether the test of necessity under Article 19(3) alternatively is or can be met by conceiving extremism legislation as aiming to respect the rights of others. ‘Respecting’ rather than ‘protecting’ suggests prima facie that there is an even lesser need for stringent curtailment and criminalisation; however, recalling the operative paradigm of extremism begetting terrorism it may still be viable under this heading to justify some restriction on extremist speech.

Restricting Intolerance

Different conceptual understandings of liberalism alter the point at which it is deemed necessary to exclude a portion of free speech from civil society space in order to respect the rights of others. This paper earlier supported Schmid’s notion that the version of legality in respect of extremism may be allowed to be restricted, if at its core, it is anti-liberal, and this conclusion returns in relevance here. When assessing necessity, the liberal State usually neutrally weighs rights against each other, making a content-empty assessment of whether a particular contested expression should prevail. Nehustan, however, forcefully rejects this approach. He contends that such neutrality undercuts the liberal State’s positive duty to promote human flourishing by protecting certain values and ways of life. Backing the ECtHR tradition of taking a more content-based approach to regulating speech, Nehustan encourages all liberal States to be actively biased; to take the content of expression into account when balancing it against competing rights and provide less protection to anti-liberal expressions.142 Under Nehustan’s conception of liberalism it arguably becomes necessary to take active steps to restrict those espousing extremist ideologies in order to prevent them poisoning civil society space.

Nehustan’s vision of ‘perfectionist liberalism’ perceives the State not as a neutral spectator in rival notions of ‘the good life’ but as an active participant responsible for creating the environment in which a particular conception of ‘the common good’ may be achieved.143 ‘Perfectionist liberalism’ as promoted by Nehustan tolerates only a certain level of illiberal behaviours and opinions running counter to the liberal ‘goods’ of autonomy, freedom and equality.144 Nehustan does not contend that the content of expression alone is sufficient to justify its curtailment and he asserts that complete prohibition of offensive expressions should be ‘extremely rare’, but he does perceive content as relevant and proclaims that intolerance should not be tolerated: that which is offensive to the liberal ear should be curtailed by the State in order to protect ‘worthwhile’ freedom.145 For Nehustan the ear offended is also relevant: he contends that liberal States should be more willing to restrict expressions when defending liberal offended persons than when defending illiberal offended persons.146

Applying the UN Rabat Plan terminology to Nehustan’s philosophy, it arguably serves a ‘pressing social need’ to restrict extremist speech since protecting the middle-ground in society respects the rights of all to have access to a free and open democratic discourse. In an earlier echo of this, Post asserts that censoring expression could be justified when allowing it would alienate all other citizens from participating in public discourse: silencing a minority serves to protect the very existence of public discourse.147 When taking into consideration again the volume of extremist voices, the need to ensure a full range of views can be expressed could make it acceptable for concerns over extremists occupying a disproportionate part of the public sphere to inform States’ pursuit of extremism legislation.148 Running counter to liberal values, extremist expressions warrant little protection.

Nehustan’s construction of liberalism, however, arguably goes too far in countenancing liberal State interference with individual freedom. The UN-identified role for States is to provide an adequate framework for persons to achieve their full potential. It places a ‘special emphasis’ on strengthening pluralistic societies including fully protecting the right to espouse minority or dissenting views.149 Established international law protects with particular fervor speech at the fringes of democracy: it is quite clear from Handyside and Voorhoof’s extensive analysis of more recent ECtHR judgments that even under a content-based approach to expression distasteful and bigoted views are strongly protected. Extremists are precisely the type of minority whose free expression is in need of protection and a secure democracy has no need to fear any disagreement with its precepts, nor even challenges to its very foundation. Permitting extremist views to form part of civil society dialogue arguably strengthens rather than undermines the democratic sphere. The potential negative impact on commitments to extremism caused by restriction is again relevant, as is the argument that rights are respected through an increase in, as opposed to a reduction of, free speech. Thus, this paper argues that it is likely unsustainable to claim any need within the terms of Article 19(3) ICCPR to restrict expressions of extremism to respect the rights of others.

Respecting Rights through Security

In concluding this section, this paper contends that whether liberalism can justify restrictions on expressions of extremism or not, the idea of ‘respecting rights through increasing security’ is in any event a fundamentally flawed concept. The UN posits security as needed instrumentally to enable individuals to pursue autonomous life plans or safeguard natural rights150 but invoking a ‘right to security’ to justify rights-restrictive counter-extremism measures is highly problematic. Examining the philosophical basis for a ‘right to security’, Lazarus argues that the tendency to ‘securitize’ rights must be resisted: where security is used as the prism to view other rights, she submits, the liberty it is invoked to promote is in fact subtly downgraded, stripped of richness and nuance.151 Similarly, Gearty contends that increasing security so that rights may be freely exercised fatally undermines the integrity of the human rights enterprise. He states that the easiest security response to an invisible threat (note again here the extremist as an ‘enemy within’) is to suspect entire communities of being evil. Placing whole populations under suspicion, however, he notes, transforms even ‘moderate’ individuals from human rights subjects into objects of control. A focus on ‘security’ thus insidiously ‘redefines’ rights such that their repression is deemed necessary to ensure their overall survival.152

Ironically, therefore, in the process of promoting security anti-liberal objectives of extremists are facilitated rather than frustrated. Pap further argues that as society perceives ‘security’ not as an objectively determined social condition but a socio-psychological construction, desire for ‘more security’ incrementally creates a tyrannical state. In consequence, ‘prevention’ becomes a ‘blank cheque’: if no violent terrorist attack occurs, governments can credit their counter-extremism security measures; any attack that does occur substantiates political will to further increase rights-restrictive regulations.153 Taken together, these views support the conclusion that a politically-asserted ‘need’ for security does not equate to Article 19(3) necessity. It does not lawfully follow that because a minority holding extreme views is suspected of connection with terrorist activity that their entitlements to core rights and freedoms can be swept away for the sake of a putative gain to the security of the majority.


Internationally, security discourse has since September 2001 pushed extremism to the forefront of global and domestic political agendas, establishing this as the ‘threat’ which needs to be confronted to prevent future violent terrorist action. But liberal States enacting increasing numbers of domestic laws impacting on fundamental rights has concerning implications for the rule of law. It insidiously empowers authoritarian States to justify repressive laws, which may then be nefariously applied outside the protection provided by strong due process rules and procedures, and questions hover over whether laws that curtail expressions of extremism comply with the established precepts of international human rights law. Restricting public declarations of extremist ideology (whether spoken, written or symbolic) clashes with the core democratic right to freedom of expression as enshrined in Article 19 ICCPR, and political claims of ‘necessity’ do not equate to legal justification. This paper has thus asked from a theoretical standpoint if and how domestic laws banning or criminalising extremist speech could comply with Article 19(3)’s tripartite test for the legitimacy of rights-restrictive measures. Article 19(3)’s criteria are cumulative: all must be met for rights-restrictive measures to be lawful; however, interrogating each in turn this paper has uncovered difficulties at every stage.

Satisfying the first criteria of legality appears an almost insurmountable obstacle. UN documents and State practice evidence that suppressing extremist speech is premised on preventing terrorism, with which promulgation of anti-liberal hate-filled extremist ideologies has been empirically linked. Given that the underlying and overriding motivation for extremism legislation infringing on the right to freedom of expression is a matter politically affecting the international community it would be highly beneficial to adopt an international law definition of extremism. Codifying extremism internationally could lessen the risk of problematic or illegitimate domestic definitions and frustrate both authoritarian and liberal State attempts to subvert approved political rationale into illegitimately repressive laws and activities. Whether codifying domestically or internationally, however, legally defining extremism is something of a labyrinthine endeavour: States and academics alike have struggled to articulate a sufficiently narrow and precise version of this concept. Legitimate restrictions to the right to free speech can be carved out only by carefully crafted terms; problematically, however, exactly identifying the markers of extremism that connect with the threat of terrorism without including purely fringe or offensive but non-violent views appears a near-impossible feat when taking into account the need also to avoid discrimination. This being so, there are strong grounds to suggest that only expressions of extremism that amount to incitement to hate, violence or discrimination and thus fall within the significantly narrower spectrum of expressions that must be expunged from society according to Article 20(2) can be legitimately curtailed. However, though this paper is very hesitant to suggest that any sufficiently narrow and precise definition of extremism could be formulated to be legal under Article 19(3), it does tentatively suggest that Schmid’s detailed list of extremism features could be a valuable resource.

Assuming that extremism can be legally defined under the terms of Article 19(3) and focusing on extremism as a root of terrorism, this paper has found it relatively simple to prove that restricting extremist speech is genuinely aimed at serving a legitimate purpose; either protecting national security or respecting the rights of others. Concerns for the rule of law under this second criterion of Article 19(3) in fact mostly arise precisely due to the relative ease of justification; there being few legal constraints on the definitions of ‘national security’, ‘rights’ or ‘others’.

Nonetheless, complexities with meeting the threshold for necessity under Article 19(3) stall any cavalier conclusions of lawfulness. The vital importance of the right to freedom of expression for democracy runs as a golden thread throughout approaches to its restriction and particularly where necessity is reviewed the high value placed upon free speech imbues assessments of the need to limit it with an unspoken bias against curtailment. This hesitancy is only heightened by the multiplicity of issues surrounding the definition of extremism and the breadth of scope available to claim legitimate purposes are served. Questions of necessity are always highly context-specific and in determining whether any restriction to extremist free speech is needed the actual expression in issue will be relevant as will the sanction applied. As a matter of theory, however, it is highly dubious whether extremist speech can ever justifiably be limited to protect national security or respect the rights of others. As far as national security is concerned, the line from extremist speech to terrorist action is filled with other factors and branches and empirical evidence strongly counter-indicates the need to restrict extremist speech to prevent terrorist violence, suggesting that in fact the very restrictions intended to prevent may have an opposing emboldening effect. The need to restrict freedom of expression to respect the rights of others is also suspect. Though some forms of liberalism would support suppressing the rights of those at the edge to ensure the voices of those in the middle are not ‘drowned out’ it is precisely those whose speech is counter-cultural or even odious that urgently need the protection of human rights law: the minority needs to be protected against the majority, not the other way around. Tolerating the presence of extremist speech does not mean agreement with its content and robust disagreement with extremism increases the opportunity for those who do disagree to participate in dialogue, strengthening freedom of expression for all in society.

Ultimately, having highlighted and explored the nuances involved in enacting extremism legislation this paper takes a cautious approach to making definitive findings. It does not argue that extremism legislation can never satisfy the high standards of international human rights law but it does conclude that to be lawful domestic extremism legislation needs exceedingly tight definitional terms and in view of the fundamental importance of free speech for democracy very careful consideration must be given to the nature, reason for and extent of any restrictive measures and sanctions. Extremist ideologies may be abhorrent to liberal society but this does not mean that those professing such views have no legal right to say their piece.