I. What are securitising requests and how are they distinct from securitising moves and successful securitisation?

The concept of securitisation originates from the work of the Copenhagen School, and specifically from that of one of its key members, Ole Wæver. After the Cold War, security policy both widened and deepened to include non-traditional security issues (for example, environmental degradation, economic decline, and migration) and referent objects of security policy other than the state (for example, the environment, large banks).Footnote ³² Securitisation theory is able to capture analytically many of these new forms of security because it holds that security is a social and political construction, whereby issues become security threats only when they are considered to be that by relevant actors. In other words, it does not matter as such whether threats are objectively present (that is, real).

Today, securitisation scholarship comprises a number of competing strands of securitisation, many of which have extended, refined, and clarified important aspects of the Copenhagen School’s original theory.Footnote ³³ In part, this theoretical work is driven by numerous inconsistencies and even contradictions in the Copenhagen School and/or Wæver’s work. For purposes of this article, it is important to note that there is some confusion over when securitisation ‘succeeds’. The Copenhagen School appears to vacillate between tying the success of securitisation to the acceptance of the securitising move by an audience (for example, in the case of terrorism in the United Kingdom, when the general public has accepted that there is a terrorist threat) and the adoption of extraordinary measures and/or exceptional means (for example, in the UK terrorism case when exceptional policy change occurred). Thus, on the one hand, they argue that securitisation is defined by ‘three components (or steps): existential threats, emergency action, and effects on interunit relations by breaking free of rules’,Footnote ³⁴ suggesting that extraordinary emergency measures are a necessary part of securitisation because rules are tantamount to ‘the normal way’.Footnote ³⁵ On the other hand, some passages of the seminal Buzan et al.’s Security: A New Framework for Analysis suggest that securitisation is not decided by the use of extraordinary emergency measures in that same way. Instead, it argues that ‘we do not push the demand so high as to say that an emergency measure has to be adopted, only that the existential threat has to be argued’;Footnote ³⁶ while in his single authored work, Wæver has repeatedly stated that ‘“security” [is] a speech act’.Footnote ³⁷

As a consequence, some scholars have argued that if the linguistic speech act element, which is seemingly so central to original securitisation theory is taken seriously, then securitisation succeeds at the point of audience acceptance (for example, in the case of the securitisation of terrorism in the United Kingdom, securitisation succeeds when the general public accepts arguments put forward by the executive that there is a terrorist threat), and adoption of security measures does not come into it. Vuori, for example, writes that ‘security measures and their public securitisation are theoretically and at times even practically separate from each other, and thereby the application of security practices cannot be a sufficient criterion for the success of securitisation’.Footnote ³⁸ In this vein he goes on to identify five types of ‘securitisation’: ‘(1) securitisation for raising an issue on the agenda; (2) securitisation for deterrence; (3) securitisation for legitimating past acts or for reproducing the security status of an issue; and (4) securitisation for control’ in addition to (5) ‘securitisation for legitimating future acts’, which he attributes to Wæver.Footnote ³⁹ Vuori overlooks, however, that the Copenhagen School actually differentiates between securitisations that, we may say (merely) exist (that is, at the point of audience acceptance) and those that are successful (meaning when, in addition to a relevant audience accepting a securitising request, security measures are adopted).Footnote ⁴⁰ In the case of UK terrorism, for example, this refers to the various new laws and police powers that came after. The reason why the school draws this distinction is that the former leaves one unable to sort the ‘important cases [of securitisation] from the less important ones’, because ‘many actions can take [the securitisation form] on a small scale’.Footnote ⁴¹

In successful securitisation, then, at least for the Copenhagen School, securitisation and security measures are not separate processes at all, while the securitising move is a part of securitisation (successful or otherwise). Moreover, in successful securitisations, utterers of securitising moves and securitising actors are one and the same; indeed, securitising actors use language to – in Vuori’s terms – legitimate past or present action, to deter and to control, in ways that they then act on. Or, as I have argued elsewhere, bona fide securitising actorsFootnote ⁴² use language either to warn off an ‘aggressor’ (loosely: any agent at the source of a threat) or to promise protection to a referent object, or, in some cases, to do both.Footnote ⁴³ This said, however, it is possible for other actors to speak security. Securitising actors (mostly) speak security in order to move or advance further in the direction of securitisation,Footnote ⁴⁴ for example, by seeking acceptance from a designated audience for the necessary policy change.Footnote ⁴⁵ Other speakers of security, however, have neither the intention nor the capabilities to perform securitisation.Footnote ⁴⁶ Instead, their purpose in speaking security is to request securitisation from another – usually more powerful – entity.Footnote ⁴⁷ Vuori refers to this as agenda-setting ‘securitisation’, but he is an exception, most other scholars indiscriminately use the term securitising move for all instances of speaking security, even if many, at least implicitly, recognise different logics informing security speech.Footnote ⁴⁸ It is reasonable to suggest that this conceptual oversight stems from the fact that for many scholars securitisation is satisfied/succeeds at the point of audience acceptance, regardless what follows on from this. If, however, we accept that successful securitisation involves policy change we can see that security speech serves different purposes. Securitising requests are separate from securitising moves because they are not uttered with the intention to legitimate, explain, or announce behaviour that will follow or that has gone before; instead, they are addressed to those in a position to conduct successful securitisation (for example, to those that can enact policy change).Footnote ⁴⁹

It is difficult to say at what point a securitising request succeeds. At least three options present themselves, one is at the point of audience acceptance (that is, when the request has been heard), arguably, however, the request is only granted when the designated securitising actor securitises, ergo securitising requests succeed at the point of securitisation. This logic, however, implies that only securitisation could remedy the problem identified in the initial securitising request, when it is possible that politicisation by the targeted actor might solve the problem. All the same, securitising requests and the conceptual separation off from securitising move, enables us to see that securitising moves can be influenced by these primary speech acts. Indeed, an important issue to observe is that while securitising actors need to be in a position of power to perform successful securitisation, anyone can utter a securitising request. That does not mean that everyone has access to the same platforms, or that everyone will be heard equally well, but simply that anyone who has a voice, or who can write, can in principle utter a securitising request.Footnote ⁵⁰ Anyone, that is, possessing the physical capability to speak or write and is not silenced by structural constraints, including fear.Footnote ⁵¹ This inclusiveness is the reason why this study on the ethics of speaking security commences from a focus on securitising requests as opposed to securitising moves by securitising actors. A holistic ethics of speaking security recognises that ordinary people can – in principle – facilitate securitisation. This said, any general rules regarding the ethics of speaking security generated by such a study also apply for securitising moves, as both securitising requests and securitising moves are forms of speaking security, neither do securitising actors have special rights exempting them from all possible ethical standards.

This article now turns to the question of whether anything goes; that is, should people be permitted to say or write what they want, or whether there should be protections, that is to say legislation that criminalises some kinds of securitising requests, notably those that target or single out racial, religious, or ethnic minorities?Footnote ⁵² In other words, and given the empirical context against which this article is set, are laws banning public expressions of Islamophobia justified?

Before this article dives into the discussion, a word on the meaning of criminalisation is necessary as there is some debate over the proportionality of certain punitive measures (especially non-pecuniary ones, such as imprisonment) for using denigrating speech.Footnote ⁵³ Although this article does not start from the premise that securitising requests should be criminalised, it does not engage with the debate over what type of punitive measure is most appropriate (despite the fact that this could sway one’s thinking either way). Instead, the research starts from the lowest common denominator, namely, it considers the issue criminalised if it is written into criminal law regardless of the nature of the punitive measures adopted.

II. The case for protections against harmful and offensive securitising requests

Lest there be any doubt, it is important to be clear that when contemplating protections against security speech, this article is not concerned with legislating against all possible kinds of securitising requests but merely against a few select types. Freedom of speech is an essential value in liberal democracies; among many other things we can see the value of free speech on matters of security in the fact that everyone’s potential ability to put important issues on the agenda can have positive consequences. Thus, quite regardless of whether or not they result in successful securitisation, once phrased in security terms issues are often more likely to be heard, discussed, and acted on. As a case in point, consider the recent warnings from scientists that the widespread overuse of antibiotics (especially broad-spectrum ones) is considered a threat to global health security, public health and – in some countries (notably the United States) – to national security, because it leads to antibiotic and microbial resistance. In England these warnings have led to new guidelines for general medical practitioners concerning the frequency of prescriptions for antibiotics.Footnote ⁵⁴ So, in contemplating protections against securitising requests, this article is concerned only with a specific type of such requests, namely ones that single out or target racial, ethnic, or religious minorities. This article distinguishes between two different ways of doing so: (1) in plain, but not offensive securitising language; or (2) in an overtly offensive way, for example, by using dehumanising language (for example, by comparing people to animals or diseases), or by employing racially or culturally motivated slurs, including ridicule.Footnote ⁵⁵ An example of the first kind comes from the United Kingdom after the 7/7 London bombings, when Muslims were repeatedly identified as threatening because they were considered likely to become radicalised and engage in terrorist acts.Footnote ⁵⁶ As Stuart Croft notes, the connections between ‘Islam and terrorism; sharia law and barbarism; sharia law, Islam and a threat to British values’ were primarily made by ‘powerful social actors [and] major media organizations’.Footnote ⁵⁷ For example, by polling Londoners’ comfort levels when sitting next to Muslims on public transport in case they might blow themselves up.Footnote ⁵⁸ Precisely through such means, the relevant actors requested government to get tough on Muslims, and for example, not build a new mosque near the 2012 Olympic Park, to disallow Muslim faith-based schools, and instead to create a Britain based on British values.Footnote ⁵⁹

An example of the second way of making security requests was mentioned with Bachman’s neologisms of ‘Rapefugees’ and ‘RefugeISIS’ in the introduction, and PEGIDA’s request for elected politicians to protect Germany and Germans from such threats. For a further example, consider the case of British media personality and Sun newspaper columnist Katie Hopkins who, during the ongoing migrant crisis in Europe, called illegal migrants ‘cockroaches’, ‘a plague of feral humans’, ‘spreading like norovirus on a cruise ship’, and requested that the British government and EU governments needed to adopt Australian-style security measures. In her own words:

Although the boundaries between these two kinds of securitising requests are somewhat fluid – and clearly offence may be taken whatever the language – there are good analytical reasons for differentiating between the two. In particular, arguments for permitting securitising requests of the first kind do not preclude the need for banning securitising requests of the second kind. The primary reason for why one might want to prohibit both kinds of securitising requests is because of the probable consequences they may have for those that are the subject of security speech (that is, the minorities in question). Thus, those identified as security threats (meaning, the relevant minority) might feel increasingly insecure in society, fearing vigilante attacks. In the UK, the Tell MAMA (Measuring Anti-Muslim Attacks) project not only reports a ‘200 percent’ increase in anti-Muslim hate crime in 2015, but identifies ‘the misrepresentation of Muslims in Britain from certain media sources, politicians and public figures’ as a primary cause.Footnote ⁶¹ Although securitising requests do not necessarily lead to securitisation or hate crime, they can nevertheless instil change in society. In particular, the securitising requests that concern us here may render society as a whole more xenophobic and racist. Xenophobia and racism, in turn, are harmful because they – to use Didier Bigo’s apt phrase – further ‘insecuritise’ minorities within society.Footnote ⁶²

The linkage between speech and insecurity is also stressed by Jeremy Waldron in his 2012 book The Harm in Hate Speech on the need for protections against hate speech. Waldron argues that hate speech deprives people of security because it erodes what John Rawls referred to as ‘well-ordered societies’,Footnote ⁶³ by denying certain groups the fundamentals of such societies, namely ‘that all are equally human, and have the dignity of humanity, that all have an elementary entitlement to justice, and that all deserve protection from the most egregious forms of violence, exclusion, indignity, and subordination’.Footnote ⁶⁴ In Waldron’s view, one important measure in assuring these fundamentals – and thus maintaining and even creating well-ordered societies – is by legislating against hate speech and defamation.Footnote ⁶⁵ ‘[H]ate speech laws aim not only to protect the public good of dignity-based assurance, but also to block the construction of this rival public good that the racists and Islamophobes are seeking to construct among themselves.Footnote ⁶⁶

All these are important points; they are relevant here because they show that securitising requests of the first kind closely resemble hate speech. Accordingly, while securitising requests of the first kind do not use offensive language, the use of the language of security (that is, the identification of a minority as a ‘security threat’ because of who they are) alone is degrading. The use of security language changes, indeed reduces, the standing of affected minorities in society. Securitising requests of the first kind also reveal that ‘hate speech’ can be motivated by something other than hatred, or in other words, by something other than racism and xenophobia. Instead ‘hate speech’, or better ‘harmful speech’, can be motivated by fear, specifically fear of the minority posing a security risk. Of course, if such fears are entirely unwarranted (that is, because no credible empirical precedent involving the minority exists) it becomes inseparable from racism, while it is also the case that securitising requests motivated by fear often end up producing racism and xenophobia. Nevertheless we ought to recognise that expressions of harmful speech, including Islamophobia are based not exclusively on hatred but also on fear.

It should be obvious that if all of the above reasons on why there should be protections against harmful securitising speech hold for securitising requests of the first kind, then they also hold for securitising requests of the second kind, as the latter are a more serious case of the former. That said, there are additional reasons for curtailing securitising requests that use offensive language when they single out or target racial, ethnic, or religious minorities. Yet, it is not always easy to know when something is offensive. In hate speech legislation, offending is seen as a step down from instilling hate and is not generally covered.Footnote ⁶⁷ One reason for this is the extremely subjective nature of offence. Specifically, thresholds for when offence is taken differ widely with age, culture, religion, and also with who the speaker is. Plus, it is possible to offend without intending to do so. For example, the Caucasian Oscar-nominated actor Benedict Cumberbatch had to apologise profusely for causing offence when he referred to black actors as ‘coloured’ in an interview with a US broadcaster.Footnote ⁶⁸ The comments were made in the context of Cumberbatch bemoaning the absence of equal opportunities in the industry.

With offence hard to capture, it is important to recognise that it might be extremely difficult to provide legislation criminalising securitising requests that use offensive language.Footnote ⁶⁹ However, in the interest of (national) security, the case for criminalising these is strong, because they are doubly dangerous. Not only do they ‘insecuritise’ minorities, erode well-ordered society and potentially lead to hate crime, they also run the risk of rendering the general population insecure because offence (especially against religious beliefs) may cause retaliation. Conversely, if offending in this way is punishable by law, the offence is less likely to cause retaliation and/or to occur in the first place.

In recent years we have seen revenge attacks in response to securitising requests of the second kind (that is, those that use offensive language or visuals) in the aftermath of the so-called Muhammad cartoon crisis. Lene Hansen has argued that especially in the age of the Internet, images are prolific and influential and ought to be included into a theory of securitisation.Footnote ⁷⁰ This is because ‘to constitute something as a matter of security is to make a political intervention’Footnote ⁷¹ and images (especially iconic images) can have political implications.Footnote ⁷² This political intervention can take a number of forms. Hansen suggests that images can very readily depict ‘the Other as demonic, barbaric, evil, and menacing’ while they can also be used to belittle the Other ‘constituting him/her/it as insignificant, weak, small, cowardly, backward, or feminine, as someone “to be laughed at rather than hated or feared”’.Footnote ⁷³ ‘A depiction through demonization’, Hansen further argues, ‘constitutes a threat to be conquered while a strategy of belittling makes the threat manageable’.Footnote ⁷⁴ Either way, images, like words, may thus be utilised to call on powerful actors to securitise any given issue or (perceived) threat.

Hansen nevertheless holds that images ‘securitise’ usually in conjunction with spoken or written words that give a specific meaning to the image. Words are necessary because the image itself can convey multiple meanings. For example, ‘the lit turban-bomb, the bushy eyebrows, and the piercing eyes [of Kurt Westergaard’s (in)famous bomb-turban cartoon] evoke on the one hand a violent, possibly terrorist-suicidal subject; on the other this is a disembodied face, devoid of spatial and temporal location with no other subjects present’.Footnote ⁷⁵ She further argues, however, that clear meaning was given to the cartoons, which were obviously belittling, in the context of several accompanying editorials in Danish newspapers which identified the West as existentially threatened ‘from “Muslim” imams as well as from the wider “Islamic worldview”’ and requested that ‘Western society’ and ‘Muslims’ deter this threat.Footnote ⁷⁶ One way of resisting the narrative suggested in securitising requests is through counter-securitisation.Footnote ⁷⁷

Many Muslims were greatly offended by the cartoons and how they were being contextualised in Denmark and abroad with many media outlets republishing the cartoons. Their publication led to widespread violent protests (including in Pakistan, Afghanistan, and North Africa), terrorist attacks (besides Charlie Hebdo also deadly shootings in Copenhagen in February 2015) and arguably a deepening (or indeed the initial creation) of a ‘clash of civilisations’.Footnote ⁷⁸ Given the climate in which these cartoons were published (that is, during the global war on terror) these consequences were not only probable but also foreseeable. Since individual newspapers clearly do not have that foresight and may very well be motivated by agendas (especially financial or political) other than one of greater overall security, it is reasonable to suggest that legislators should intervene and curtail the spreading of unnecessarily offensive securitising requests (in this case offensive images accompanied by offensive editorials) by criminalising such security speech. This suggestion is made on the basis of empirical evidence. Researchers have found that civil hate speech protections have a positive effect on ‘mediated outlets’ because the staff is trained to be sensitive to hate speech legislation.Footnote ⁷⁹ Indeed, the effects of what Gelber and McNamara call ‘the “laws” educative role’Footnote ⁸⁰ is such that it is reasonable to suggest that had unduly offensive securitising requests been controlled by legislation (that is, criminalised), it is a serious possibility that the massacre at the French satirical magazine Charlie Hebdo in which 12 staff were killed by 2 Islamist gunmen might have been avoided. At a minimum, it is likely that the magazine would have abstained from publishing the images and the accompanying editorial.

III. The case against protections of securitising requests whatever their nature

The case against protections of any kind of securitising request starts from the assumption that free speech is a fundamental value of liberal democratic societies that ought not to be curtailed. Most advocates of free speech believe that hate speech legislation is unwarranted and that the only things that ought to be curtailed are pornography involving minors and extreme libel.Footnote ⁸¹ When it comes to security we can see the value of free speech quite clearly; thus in Western liberal democracies anyone is allowed to criticise their government for making people (feel) insecure, whether by misguided foreign policy, exaggerating the terror threat, monitoring populations through technologies including surveillance, or for not providing adequate military hardware for armed forces in combat situations. If we start to curtail some kinds of securitising requests by legislation then where does it stop? In times of crisis, when unanimity is key and often assured,Footnote ⁸² might not the legislator be tempted to prohibit criticism of executives’ policies or indeed ‘desecuritising requests’? In the United States, for example, criticism of the official governmental line in response to the terrorist attacks of 9/11 was considered deeply unpatriotic.Footnote ⁸³ At the time, the prevailing mood was captured in George W. Bush’s often used phrase: ‘Either you are with us, or you are with the terrorists’,Footnote ⁸⁴ while at the same time, the PATRIOT Act and subsequent counter-terrorist activities, which allowed the extended surveillance of suspected terrorists by the newly-created Department of Homeland Security were widely regarded as reminiscent of McCarthyism.Footnote ⁸⁵

A second deontological argument against protections is that securitisation theory tells us that insecurity is a matter of perception and politics.Footnote ⁸⁶ Surely if one is scared of something or someone, one should have the right to say it, regardless of the consequences. This is particularly important in light of the fact that fear can be a direct result of governmental policy. For example, one may believe that the UK government has with its counter-terrorist policies created a ‘suspect Muslim community’,Footnote ⁸⁷ and thus driven hitherto moderate Muslims into radicalisation with the result that they now pose a concrete threat to oneself and other British people. The right to voice one’s opinion in this case is essentially the late Ronald Dworkin’s legitimacy argument in favour of free speech.Footnote ⁸⁸ Much simplified, he argues that in a democracy everyone is forced to live by some laws that they do not agree with (he has in mind laws about positive discrimination), because of the consequences of those laws (that is, admittance to jobs, etc.); if one is not allowed to voice one’s discontent against the law in question, then the legislator is illegitimate because – in a democracy – legitimacy arises from everyone having a voice. In his own words:

A third argument against protections of securitising requests – whatever their nature – is that securitising requests can fail out in the open.Footnote ⁹⁰ This is essentially an extrapolation of the argument by many opponents of hate speech protections that opinions are better out in the open than driven underground.Footnote ⁹¹ Allowing securitising requests to fail democratically would render a stronger, well-ordered society. Notably, historical evidence suggests that a society whose citizens live according to governing principles that they believe in is bound to be stronger than one where universal values are enforced top-down by the legislator.Footnote ⁹² Indeed, criminalisation of securitising requests can create further resentment against the feared group(s). Instead of being addressed, the issue could then be driven underground where it might gain momentum, perhaps in part, out of resentment against the legislator.Footnote ⁹³

Finally, it is not clear in what sense securitising requests cause factual, direct harm. In a 2013 critical reply to Waldron’s The Harm in Hate Speech, Dennis Baker and Lucy Zhao argue that Waldron fails to distinguish between theoretical harm and factual harm. They argue that no factual harm is caused when putative victims simply feel that their dignity is compromised by denigrating speech.Footnote ⁹⁴ They also remind us of what precisely states seek to prevent by means of hate speech legislation:

In this context, it is important to recall that unlike with straightforward hate speech, the utterer of the securitising request does not want the ‘audience’ to do abhorrent and harmful things (that is, commit hate crimes). Instead they want the securitising actor (that is, the audience of the securitising request) to securitise the issue; and it is conceivable that – sometimes at least – abhorrent language is used primarily to stress the urgency of the matter and to get a powerful audience’s attention. In short, even overtly offensive securitising requests are not incitement to hatred, but rather requests for protection of valued referent objects driven by fear and hatred. This is important because it means that criminalisation is a disproportionate response because securitising requests (that is, security speech) does not in fact cause factual or direct harm, but merely theoretical or indirect harm, insofar as targeted groups perceive greater levels of insecurity or feel offended. Direct, factual harm is caused only when someone in the position to securitise follows up on the requests and securitises the issue. For example, it is well established that in the United Kingdom, Muslims were not only disproportionally affected by new terror legislation following the 7/7 bombing, but some of the measures (notably detention without trial) were directly harmful.Footnote ⁹⁶ Arguably, however, the use of security measures that are directly harmful is unjustifiable only if the securitisation itself is unjust.Footnote ⁹⁷ At most, we can thus debate whether unjust securitisation can or ought to be criminalised, while the criminalisation of securitising requests is always unjustified.

Direct harm is also caused when a counter-securitisation is launched (as is the case with revenge attacks), but these clearly are an unintended consequence of the initial securitising request and most certainly not the intention of those requesting securitisation. Finally, direct harm is caused when vigilante commit hate crime against minorities informed by the securitising request.

It is one thing to argue that we ought to be allowed to say what we want, but may we also say it in the way we want? In other words what possible arguments are there for not restricting securitising requests of the second kind (those that are overtly offensive to minorities)? Existing legislation on hate speech covers some but not all overtly offensive acts possible here, in spite of the fact that this legislation is growing. But there are problems with legislating against overtly offensive securitising requests. Waldron, for one, argues that ‘offense is not a proper object of legislative concern … [unlike dignity it is not about] objective or social aspects of a person’s standing in society’ but rather, it is about ‘subjective aspects of feeling, including hurt, shock, and anger’.Footnote ⁹⁸ At least three more arguments are relevant in this context, and all of them can be exemplified using the Charlie Hebdo case. The first is advanced as part of Charlie Hebdo’s editorial supporting publication of the cartoons, where it was argued that ‘the suggestion that you can laugh at everything, except certain aspects of Islam, because Muslims are much more prickly than the rest of the population – what is that, if not discrimination?’.Footnote ⁹⁹ In other words, all faiths are equally ridiculous, fanciful, and/or sacred, and if we are allowed to laugh at one of them, then by implication, we should be allowed to laugh at all of them.Footnote ¹⁰⁰

Second, peace researchers have suggested that humour can not only be a conflict generator, but also a tool for peacebuilding.Footnote ¹⁰¹ Recall Lene Hansen’s point that the bomb-turban cartoon can be interpreted in different ways. Specifically, ‘the cartoon relies upon a generic image of “the non-Western” as looking wilder, more rural, and less modern than a Western subject, yet with the possible exception of the turban, the imagery is not distinctly Muslim’.Footnote ¹⁰² Precisely because it is not distinctly Muslim, it could have presented those Muslims who consider the views of radicalised ‘Islamists’ un-Islamic with a means to distance themselves from such people, and to strengthen a joint French, Danish, etc. identity with their fellow countrymen and women. Such togetherness and unity could help counteract radicalisation with fewer young Muslims being caught between different worlds. Indeed, in the United Kingdom, counter-terrorism police are already working together with a Muslim comedian to counter-act radicalisation is secondary schools.Footnote ¹⁰³

Finally, it is fairly safe to assume that individuals who act in the abhorrent way that Saїd and Chérif Kouachi (the Charlie Hebdo gunmen) did in response to offensive and belittling cartoons would have done so anyway in revenge against foreign policy choices, etc. In the aftermath of the shooting it emerged that both individuals had become radicalised over a long period of time and one had previously been arrested on his way abroad to commit jihad; in other words, the ideological commitment was already there, the cartoons may have simply been the catalyst.

IV. Requesting securitisation ethically

While the arguments on both sides are strong, from the point of view of securitisation, one of them carries more weight than the others because it goes right to the heart of the difference between simply speaking security and securitisation proper (that is, security speech plus policy change). This is the argument that legal protections against harmful and derogatory security speech are a disproportionate response because the harm caused by such speech acts is indirect/theoretical, and accordingly, that only direct, factual harm, as that caused by securitisation may warrant criminalisation.Footnote ¹⁰⁴ Yet, even if criminalising harmful securitising requests (including banning expressions of Islamophobia) is to be rejected on the grounds of proportionality, the analysis also suggests that mere security speech is or should not be free from ethical considerations. Not only because indirect harm is hurtful to the minorities in question, but also because it detracts from the fundamentals of a well-ordered society.

Given that everyone can speak security by requesting securitisation, we can say that principles governing the ethics of speaking security ethically apply to all persons, and not only to securitising actors when uttering securitising moves. The existing and growing literature on ethics and security/securitisation, sometimes known as the ‘positive/negative debate on security’, has so far failed to engage with the ethics of security speech. Ethical concerns over securitisation and hence the need for moral criteria governing securitisation arises largely from the ill effects of the exceptional measures legitimised by securitisation.Footnote ¹⁰⁵ Or, in other words, the very issue of its ethics has arisen because securitisation causes direct harm and causing harm can supposedly only be justified if securitisation itself is just. Since securitisation theory is very much concerned with security speech, just securitisation research must consider also the ethics of security speech. An ethics of speaking security must start from the concept of securitising requests not only because securitising requests are much more inclusive than securitising moves, but also because securitising requests can influence securitising moves. This article proposes that two issues are important in this regard: when to request security and how to request security. In the following each is addressed in turn.

By drawing on the ‘just war’Footnote ¹⁰⁶ tradition, I have argued elsewhere that the single most important criterion for just securitisation is the presence of a just reason, because exceptional measures can only be justified in the presence of a real threat.Footnote ¹⁰⁷ Given that the use of exceptional measures is the objective of those requesting securitisation, we can reasonably require for persons requesting securitisation to be as certain as they can be that the issue (including the racial, ethnic, and religious minority) presents a real threat before they utter security speech. The connection between objectively present or real threats and ethics is made by a number of scholars writing on ethics and security, albeit usually not from the perspective of securitisation.Footnote ¹⁰⁸ Indeed, the invocation of real threats seems antithetical to the very idea of security as a social and political construction. This said, it needs to be remembered that the Copenhagen School, or at least Ole Wæver, does not hold that there are no such thing as objective existential threats, but rather that ‘security threats’ are socially and politically constructed. In his own words: ‘Lots of real threats exist, but they do not come with the security label attached.’Footnote ¹⁰⁹ With some notable exceptions,Footnote ¹¹⁰ securitisation scholars tend to be only interested in how issues achieve security threat status; and some might deny the possibility of ascertaining threat status as it is impossible to know all the morally relevant facts about a situation. This assumption is not disputed here. Consequently, instead of requiring that relevant persons need to establish the objective presence of threats in a ‘fact-relative sense’,Footnote ¹¹¹ a judgement about the status of a threat needs to be true in an evidence-relative sense. Evidence-relativity refers to the situation when the available evidence suggests decisive reasons that the beliefs people hold about a given situation are true.Footnote ¹¹² In practice this means that persons intent on requesting securitisation must aim to gather evidence of the threat level of someone/-thing approximating the actual facts before requesting securitisation. In the case of the suspected link between refugees and migrants and terrorism, for example, the PEGIDA leadership should have checked the facts on migrant and refugee involvement in the Paris terror attacks, which was dismissed as false soon after the attacks.

Persons requesting securitisation intent to bring about securitising moves and securitisation. If this is so, then requesting securitisation ethically also means that persons requesting securitisation need to be as sure as they can be that the issue in question is ideally treated as a matter of security and not simply as a political matter. This requires them not only to anticipate the probable consequences of an ensuing successful securitisation, but to measure – as best as possible – these consequences against not securitising. Relevant in this context are both the likelihood of securitisation successfully averting the threat, and also the probabilities of securitisation not causing further insecurity by generating counter-securitisation or additional/new security threats. The securitisation of terrorism and the resulting treatment of Muslims in the UK after the 7/7 bombings, for example, led to an increase in homegrown radicalisation and recently the issue of the return of foreign fighters, throwing doubt onto the initial decision to securitise or – at least – about the way it was done.Footnote ¹¹³ Importantly, even if there is a real threat, securitisation as opposed to politicisation is not always the best solution.

Moving on to the issue of how to request security ethically, it is helpful to consider Tariq Modood’s argument (made in the context of the Muhammad cartoon): ‘One relies on the sensitivity and responsibility of individuals and institutions to refrain from what is legal but unacceptable. Where these qualities are missing one relies on public debate and censure to provide standards and restraints’.Footnote ¹¹⁴ That Modood is correct in identifying, what we might call ‘a collective standard of communication’ is apparent from the fact that the above cited securitising requests by Petry, Bachmann, and Hopkins require little in the way of explanation as being unacceptable; instead they rather speak for themselves. A decisive feature about securitisation is that it legitimises the use of exceptional measures, yet the examples suggest that requesting securitisation ethically means to observe the collective standard of communication. In other words, even a real threat does not legitimise exceptionally derogatory and defamatory speech.

Moreover, the examples discussed above also show that in the interest of peace and security it is not always wise to say or publish anything and everything at all times. The Muhammad cartoon crisis, for example, was made worse by newspaper and magazine editors’ decision to exacerbate a sensitive issue by republishing the initial offending cartoons and by commissioning new, equally offensive ones. Waldron has a valid point when he argues: ‘Often, the best that [newspaper editors] could say for [republishing the cartoons] was that they were upholding their right to publish them. But a right does not give a right-bearer a reason to exercise the right one way or another’.Footnote ¹¹⁵ However, given that many politicians and security organisations (the police, border agencies, etc.) benefit from a politics of fear that often involves singling out minorities, it must be acknowledged that self-censorship faces many obstacles.Footnote ¹¹⁶

Although the principles of the ethics of speaking security have here been developed in the context of securitising requests, it goes without saying that these apply also to securitising actors and their securitising moves. There is simply no reason why securitising actors should be exempt from these basic rules of engagement. Rules that, if obeyed, could reduce the level of harm caused in society considerably.

In summary, we can now say that just security speech has three components: Speak/request security only when (1) there is a threat that is real in an evidence-relative sense; and (2) when securitisation is likely to be a better solution than politicisation or non-action whereby ‘better’ is measured in terms of the least amount of insecurity caused, as well as in terms of securitisation’s relative chances at successfully addressing the threat. Finally, when requesting/speaking security (3) observe existing collective standard of communication, and – in the interest of peace and security – exercise a degree of self-censorship.