Can Hatred Speak ? On the Linguistic Dimensions of Hate Crime *

Starting from the pressing issue of hate crime and hate speech, as addressed by several EU Framework Decisions and recently by the International Criminal Tribunal for the former Yugoslavia (ICTY), the article focuses on the complex and often contested relation of language and violence. Whereas recent accounts are usually concerned with the violent dimension of speech and language, the article approaches the question from a different methodological angle and asks in which way hatred and violence might be understood as a form of speaking and address. This approach is based on the more general thesis that it is possible neither to gain an adequate understanding of speech and language without considering their violent force, nor to gain an adequate account of interpersonal violence without considering its linguistic dimension. In order to support this view, it will be argued that it is precisely the symbolic-linguistic character of hate crime that is responsible for its particular injurious force as well as for its dehumanizing effects. Moreover, hate crimes are not only linguistically in character, they are also directed against the possibility of language and speech itself, insofar as they aim at making us speechless, depriving us of the possibility of speaking out and being heard. This leads to an account of hate crime as a form of multi-addressed violence that not only calls for a strong concept of responsibility, but also requires a differentiated response by all social institutions and authorities.


Introduction
In March 2016 the ultranationalist Serbian politician Vojislav Šešelj was cleared of all charges of war crimes and crimes against humanity by the Third Chamber of the International Criminal Tribunal for the former Yugoslavia (ICTY). 1 The judgement was, however, by no means consensual.On the one hand, the Chamber acknowledged that Šešelj repeatedly and systematically incited hatred against the non-Serbian and Muslim population, propagated the creation of Greater Serbia, and appealed for the expulsion and forcible transfer of Croats and Bosnian Muslims.On the other hand, two of three judges, Judge Jean-Claude Antonetti and Judge Mandiaye Niang, did not find sufficient evidence to establish a direct link between Šešelj's statements and the atrocities committed by his troops, which called themselves "Seselj's men".2Moreover, the majority took the view that the creation of "Greater Serbia" and the appeals for expulsion were "more of a political venture than a criminal project" (ICTY 2016b), more of an "expression of an alternative programme that would never be put into practice" (ICTY 2016d).In doing so, the majority clearly questions an immediate connection between saying and doing, between the uttering of certain words with a particular meaning on the one hand, and the act one performs and the effects one achieves in and by saying something on the other hand (cf.Austin 1975: 120).
In contrast, the third Judge Flavia Lattanzi firmly opposes this position and suggests a direct correlation between Šešelj's speeches, their humiliating and dehumanizing effects, and the war crimes and crimes against humanity committed by his men.According to Lattanzi, there is ample evidence to conclude that he incited the crimes charged in the Indictment […] through all of his inflammatory speeches, calling clearly and directly for the expulsion and forcible transfer of the non-Serbs, in addition to his speeches in which he denigrated and dehumanized the Croats comparing them to "primates" and "vampires" and qualifying them as cowards.In the same way, he called the Bosnian Muslims "balija" or "pogani" which he himself translated as "excrements".I hold that the use of these terms, together with his constant references to "genocide" committed by the Croats during World War Two and the saying "an eye for an eye, a tooth for a tooth" and "revenge is blind", Vojislav Šešelj also took the risk that murder, torture, cruel treatment and destruction would be committed […].
What the dissenting opinion of Judge Lattanzi clearly brings to light is not only the violent character of certain forms of speech, but also the dissent between conflicting views of language: between language as a mere means of expression and communication, and language as an effective instrument to injure and to inflict pain on others. 3This tension is also reflected by recent attempts of the European Union to combat "certain forms and expressions of racism and xenophobia by means of criminal law" (Council of the EU 2008/913), while at the same time ensuring the fundamental rights to freedom of expression and freedom of the press.Thereby, the EU -following demands by the OSCE Representative on Freedom of the Media 4 -aims to eliminate the existing judicial imbalance between so-called violations of honour on the one hand and racist and xenophobic speech on the other hand.Whereas legislation against defamation, such as libel or slander, is often (mis)used by state authorities in order to limit freedom of speech and freedom of the press, as well as to silence political dissent, forms of hate speech, inflammatory speech, or incitement of hatred continue to go unrecognized or are restricted to very specific cases in criminal law. 5 Accordingly, the European Union legislation pursues a double objective: 1. decriminalization of acts of defamation or honour crimes on the one hand, and 2. enforcement of criminal prosecution of hate crime and inflammatory speech on the other hand.
It is noteworthy that in the EU this trend is closely connected to efforts to combat terrorism and to find effective legal ways to prosecute Islamist hate preachers.This connection is inter alia motivated by the assumption that expressions of racism and xenophobia as well as terrorism are violations of fundamental rights and of "the universal values of human dignity, liberty, equality and solidarity, respect for human rights and fundamental freedoms on which the European Union is founded" (Council of the EU 2008/919; see also 2008/913).In doing so, the Council focuses not only on the violent and injurious force of language and speech, but also on the increasing virtualisation and medialisation of violence by digital media.This does not only include newly arising forms of cyber-bullying, harassment, or hateful comments on social media, but, remarkably, also the internet as a "virtual training camp" for potential terrorists. 6he report of the European Agency for Fundamental Rights (FRA) Making hate crime visible in the European Union: acknowledging victim's rights (2012) points to a similar direction.On the one hand, the FRA-report adopts a human rights perspective and argues that hate crime is an "abuse of fundamental rights" (cf.FRA 2012: 10) thereby constituting a crime in its own right.On the other hand, it explains the unique character of hate crime as well as its particularly destructive and injurious force by reconstructing hate crime as a kind of "speaking" that communicates specific messages to its addressees.
What these various examples -the Šešelj-judgment, the Council Framework Decisions, and the FRA-report -clearly bring to the fore, is the complex and often contested relation of language and violence.While in Western philosophy this relation is traditionally considered as a merely external one or even as one of strict opposition, more recent accounts stress the intrinsic entanglement of language and violence.Whereas these attempts are usually concerned with violent dimension of speech and language, in this contribution, I want to approach this issue from another methodological angle.Therefore, I will not take language as my starting point and ask in which way linguistic utterances may constitute acts of violence and a criminal offence.Rather, I will follow the FRA-report and ask in which way hatred and violence might be understood as a form of speaking.In doing so, I hope to explicate the particular violating and injurious force of hate crime and hate speech as well as the intrinsic relation of language and violence.This is based on the hypothesis that it is possible neither to gain an adequate understanding of speech and language without considering their violent force, nor to gain an adequate account of interpersonal violence without considering its linguistic-symbolic dimension.This does not imply that language can be equated with violence.The aim is, rather, to reject unilateral and reductionist explanations and to envisage a more subtile differen-tiation of the relation of language and violence that takes their complex entanglement into account.
In a first step, I will give a schematic overview of the problem of linguistic violence and vulnerability in Western thought (2) and, thereafter, propose a fundamental critique of the classical opposition of language and violence (3).In order to illustrate my point, I will analyse the FRA-report and outline the specific characteristics, functions and effects of hate crime (4).As a result, it will be demonstrated that it is less the brutality of the deed itself, but rather its linguistic-symbolic character that is responsible for its particular injurious force as well as for its dehumanizing effects (5).Moreover, hate crimes do not only involve a fundamentally linguistic aspect, rather they are also directed against the possibility of language and speech itself, insofar as they aim at making us speechless, depriving us of the possibility of speaking out and being heard (6).Finally, this will lead us to an account of hate crime as a form of multiaddressed violence that not only calls for a strong concept of responsibility, but also requires a differentiated response by all social institutions and authorities -by courts of law as well as by politics and civil society (7).

Linguistic violence and vulnerability
The problem of linguistic violence and vulnerability and thereby the complex relation between language and violence is at the centre of many recent debates.From insulting utterances and injurious speech, religious and political forms of propaganda, hate speech, and incitement of hatred, to the rapidly growing problem of cyber-bullying and cyber-harassment, we are confronted with a wide variety of symbolic and medial forms of violence in our everyday life that have to be acknowledged and dealt with.Moreover, it turns out that not only the violent dimensions of speech and language but also the communicative and symbolic dimensions of violence have to be analysed, if we want to understand and respond especially to newly arising forms of violence, e. g. global terrorism as an insidious "strategy of communication" (Waldmann 2011).
However, if one takes a closer look at the history of philosophy, one realizes that the challenges posed by linguistic violence and vulnerability have been systematically underestimated or disregarded.Although language has been one of the main issues of 20 th -century philosophy, and violence has been at the centre of interdisciplinary research at least since the 1960s, remarkably, the problem of the constitutive connection of language and violence has been largely neglected until the end of the 20 th century.
There seem to be good reasons for this omission: For if we take a look at the history of philosophy, it appears that the conceptual pair of language/violence is closely connected to the classical oppositions that characterize Western thought: culture/nature, mind/body, civilization/barbarity, reason/madness, rational/irrational, sublime/brute, meaningful/meaningless, eloquent/mute etc.If one accepts this binary logic, the nonviolent and unconstrained character of rational speech is opposed to the irrationality and senselessness of brute force and violence.Thus, it would be impossible to assign sense, meaning, or value to violence as well as brutality and oppressiveness to language (cf.Hirsch 2001).Violence, understood in this way, is both a threat to language and to reason, a threat to any reasonable, rational and modest speech.It not only puts an end to every dialogue and communication, but also destroys the very possi-bility of sense and meaning, whereas speech, on the other hand, represents the very promise that nonviolent forms of cohabitation and cooperation might be possible.
Thus, according to Hannah Arendt, violence begins where speech falls silent, while true human action is "transacted in words" (Arendt 2002: 340;1989: 26).If language can nonetheless be forceful and violent, this is not attributed to language itself, but either to the fact that it partakes in a prior physical violence that is the actual source of its injurious force, or to the possibility that the very foundations of cooperation and action oriented toward understanding are undermined by strategic reasoning (cf.e. g.Habermas 1984).
In contrast, poststructuralist philosophers like Jacques Derrida and Michel Foucault -taking Friedrich Nietzsche as a starting point -elaborated on the relation of violence, language, and speech already in the 1960s and 70s.While Derrida focused on a kind of "original" or "transcendental" violence inherent to any identifying act of naming as its condition of possibility (cf.Derrida 1978Derrida , 1997)), Foucault was interested above all in the procedures of exclusion, regulation, and delimitation by and in which the speaking subject is engendered and whichprior to any utterance -determine what can and what cannot be said (Foucault 1972).
In the 1970s and 80s, in the wake of Foucauldian discourse analysis and John L. Austin's speech act theory, sociolinguistics and feminist linguistics addressed the open and hidden forms of violence in our everyday communication.By demonstrating that even in our most 'innocent' conversational procedures, like change of speaker, interruption, limitation of speaking time or choice of topic, subtle forms of power and violence are exercised, it could be shown that forms of linguistic violence are equally operative within the public and political space as well as in private discourses (cf.Lakoff 1973;Trömel-Plötz 1984;Frank 1992;Wodak 1997).
However, it was not before the 1990s that the relation of language and violence was systematically scrutinized in its whole complexity.Judith Butler's Excitable Speech: A Politics of the Performative (1997) has been an especially important stimulus for the theoretical elaboration of the problem of linguistic violence and vulnerability.Based on Austinʼs speech act theory, Louis Althusser's notion of subjectification as ideological interpellation, and Derridaʼs concept of iterability, Butler demonstrates that we are vulnerable to verbal utterances because we are fundamentally linguistic beings which rely on linguistic acts of address and recognition in order to maintain their social existence.

Critique of the classical opposition of language and violence
In fact, there is strong evidence for the assumption that language and violence cannot be separated from each other readily.This even holds true for cases where language or other forms of representation, at first sight, merely serve as a means to represent and describe acts of violence: For example, in the case of so-called "happy slapping", mobile phones are used to record and spread videos of violent acts via social media.However, a closer look clarifies that the representation of the "real", physically violent act is not just a secondary or derivative phenomenon.Rather, it turns out that the "original" act of physical violence is also a means in the service of a symbolic violence exercised on the victim and potential victims by the infinite reproduction of the images, an effect that is evident also in the strategic employment and distribution of "execution videos" by the so-called "Islamic State".
What comes to light here, besides the ethical problem of the representation of violence and suffering, is the violence that can be found in the representation of violence itself that has so far been barely acknowledged (cf.Sontag 2003;Butler 2009;Posselt 2012).In fact, the way in which we describe, name, or frame certain events or actions makes a decisive difference in how we perceive and evaluate their violent character (see e. g. the controversial term "genocide" and its strategic employments).Consequently, not only the prior or subsequent legitimization of violence, but also the perception and experience of violence is already discursively structured and mediated.
The intrinsic entanglement of language and violence can also be demonstrated by the fact that already the very definition of violence poses a central challenge for any theory of violence.
For any definition of violence (which is at first sight a purely linguistic act) is itself a violent act of positing, insofar as it delimits from the beginning what can be considered as violence and what cannot -with all its far-reaching social, moral, juridical, and political consequences.
Heitmeyer and Hagan strongly emphasize this point in their International Handbook of Violence Research by underlining that the "problems of violence research begin at the outset, with the attempt to determine exactly what should be classified as violence" (Heitmeyer/Hagan 2003: 4).These problems become not only manifest in the fundamental ambivalence of the concept of violence, insofar as violence can involve both the destruction (as, for example, in the case of war) and the creation of order (as, for example, in the case of founding a state), but also in the "many disagreements about the authority of definitions of what violence is, or is said to be", which turn any attempt to determine definitively what constitutes violence into a high-risk and ultimately violent undertaking (ibid.).
Taking these findings seriously, we have to concede that questions of language and violence do not exclude, but rather imply each other mutually.But how exactly do we have to understand that relation and what does this mean for our conception of language on the one hand (A) and of violence on the other (B)?
(A) If we look at contemporary theories that address these questions starting from language, we usually get a twofold answer: 1.We can not only describe, express or articulate violence by verbal utterances, but also carry out acts of violence: inflict pain on others, humiliate and mortify them.2. At the same time, there seems to be a violence proper to language itself.For the structures, norms, and conventions of language not only determine what can be said and in which way; they also delimit what is speakable and what is excluded from the realm of discourse (see Herrmann et al. 2007).In the first case, linguistic violence is ascribed to the performative and act-like character of language; in the second case, linguistic violence is located in the norms and conventions of language itself.While the first approach focuses above all on personal and concrete forms of linguistic violence, the second position implies an abstract concept of violence, for there is no subject to which the act of violence can be ascribed and also the addressee of the violent act remains undetermined.Thus, two forms of linguistic violence can be differentiated: 1. violence by language that can take the form of addressing or representing someone as something (e. g.Althusser 1971;Butler 1997;Hall 1997); 2. violence in language, violence operative in language itself that can take the form of an identifying, equalizing violence as in Nietzsche or Adorno (1973), a transcendental violence as in Derrida (1978), a discursive violence as in Foucault (1972), a symbolic violence as in Bourdieu (1991) or an epistemic violence as in Spivak (1988).However, a more detailed analysis shows that these two positions cannot be strictly separated from each other; rather, they presuppose each other mutually (cf.Butler 1997;Herrmann/Kuch 2010;Posselt 2011).
For it is possible neither to describe linguistic violence simply as a singular event, as an individual speech act with a clearly assignable speaker independent of the linguistic structures and conventions that render it possible, nor as a form of violence that is simply inscribed in the symbolic order of language itself, independent of the realization and actualization of the linguistic rules and conventions by the individual speakers.
(B) If we approach the question of linguistic violence starting from violence, we find an equally ambivalent complexity in the concept of violence.This can be illustrated by the German term for violence: "Gewalt".For the German "Gewalt" comprises two different meanings that can be traced back to the Latin terms vis/violentia and potentia/potestas: 1. Gewalt refers to the negatively connoted immoral and illegal violation of persons (violentia).
2. Gewalt refers to the authority in relations of power and domination (potestas) (which becomes manifest in the German expressions "Gewaltenteilung", "Staatsgewalt").In short, while violentia -understood as a notion of action -refers to the illegitimate, criminal, destructive force that someone inflicts upon another, potestas -understood as a notion of competence -signifies the legitimate, justified and institutionalized power of control (cf. Neidhardt 1986; Corbineau-Hoffmann/Nicklas 2000; Krämer 2010).Thus again, two forms of linguistic violence can be differentiated: 1. linguistic violence as a violating force directed against persons; 2. linguistic violence as a regulating, abstracting force, as a violent act of naming and predication, as an equation of the unequal, as Nietzsche puts it, which is the prerequisite for any concept formation (Nietzsche 1988b).
Nietzsche even takes a further step, thereby bringing the symbolic dimension of violent and injurious acts clearly into sight (cf.Posselt 2010).According to Nietzsche, not only is communication violent, insofar as every act of communication entails the attempt "to extend one's power (Gewalt) over the other"; but already "the injuries of the other are the sign-language of the stronger" (Nietzsche 1988a: 298).In short, as every imposition of a sign on the will of another is already injurious, equally every injury itself is readable as a sign.What comes to light here -besides the communicative and symbolic dimension of violence -is the close entanglement of language and body, of verbal and bodily vulnerability (cf.Scarry 1985;Felman 2003; Hermann/Kuch 2010).To the same extent that physical violence possesses a symbolic dimension (as becomes clear, e. g. in humiliating forms of punishment), linguistic violence implies a bodily component.
These considerations have far-reaching consequences that can be summed up as follows: The relation of language and violence is at the core and at the outset of any thorough analysis of language as well as of violence.Questions of language and violence do not exclude each other; rather, they stand in a relation of interdependence.Hence, we must equally abandon the notion of a pure, non-violent language as well as the notion of a raw, meaningless violence.
There is no violence "as such".Violence is always already bound up with processes of appropriation and interpretation by "victims", "offenders" and "third parties" that define and represent an action and/or experience as violence.This is, moreover, complicated by the fact that any attempt to determine conclusively the concept of violence already performs a violent act of positing, insofar as it aims at delimiting once and for all what counts as violence and what does not.

Hate crime and hate speech: A fundamental rights perspective
In what follows, I will illustrate and substantiate these rather fundamental considerations by the example of the problem of hate crime as outlined by the aforementioned report of the European Agency for Fundamental Rights (FRA) Making hate crime visible in the European Union.Although hate crimes are a widely acknowledged problem in the European Union, there is neither a uniform definition of hate crime nor a systematic official data collection, which would allow an objective assessment of the current situation.Hate crimes are, as the FRA-report claims, in contrast to their severity and destructiveness, almost invisible.Currently they are not specifically registered and evaluated by state authorities on a common basis.Moreover, hate crimes are often not reported or even regarded as crimes at all, with the effect that the offenders remain unprosecuted and victims have no access to redress.This also shows itself on the legal level: Only a few states within the European Union consider bias-motivated crimes as a specific criminal offence, even though racist motives are sometimes considered as aggravating circumstances (FRA 2012: 7-10).
In order to address these difficulties, the authors of the report propose a human rights perspective on hate crime that provides an adequate account and useful definition of hate crime, broadens the scope of official data collection, and strengthens the right of the victims to effective remedy.To this end, the report defines "hate crimes", in a first approach, as "[v]iolence and crimes motivated by racism, xenophobia, religious intolerance or by a person's disability, sexual orientation or gender identity" (FRA 2012: 7).Thus at a first glance, what distinguishes hate crimes from "ordinary" crimes seem to be the motivation "behind" the deed, i. e., the fact that hate crimes are motivated by prejudices, intolerance, or hatred against others because of their assumed race, religion, sex, gender identity, disability etc.In doing so, the report follows the assumption that hate crimes or so-called bias-motivated crimes do not only violate the physical integrity of their victims, but also infringe fundamental human rights, namely the right to human dignity and to non-discrimination, as laid out by the European Convention on Human Rights (ECHR), Article 14,7 and the Charter of Fundamental Rights of the European Union (CFREU), Article 21.Furthermore, the violation of the fundamental rights is not simply an effect of hate crime, but rather inherent to it (cf.FRA 2012: 13).Consequently, as the FRA-report argues, hate crime constitutes a crime in its own right that requires more than the mere inclusion of "any given bias motivation as an aggravating circumstance in the criminal code" (FRA 2012: 11).It is also for this reason that hate crimes require particular attention and sanctions by state authorities: For what is attacked by hate crimes is not only the individual, but also the social group to which it belongs, as well as society at large, including its fundamental democratic values.
This, however, raises the issue of "[w]hy exactly the bias motivation of the offender makes such a difference" (FRA 2012: 18)."If a person abuses another, why does it make a difference whether the offence was motivated by prejudice, as is the case with hate crimes, or by other motives, such as anger, indifference or greed" (FRA 2012: 15), as is the case with ordinary crimes?In order to clarify this question, the authors propose the following scenario: ISSN 1615-3014 So what is the key difference between these two situations?In both cases, a person's right to physical integrity is violated.However, while in the first instance the violence arises out of the argument itself and is therefore "situational", linked to the specific circumstances of the argument, in the second case C attacks D because C perceives D as being gay and generally disdains gay people.Consequently, when C attacks D, while at the same time expressing his or her contempt for C, "C's bias motive adds an insult to the assault" (FRA 2012: 19).In other words, C's prejudices against gay people add a verbal act to the physical attack, a verbal injury to the bodily injury, by communicating to the victim and the bystanders that the right of the victim (to physical integrity, non-discrimination etc.) matters less -simply because D is gay or, rather, is perceived as gay.At the same time, this perception is used as a legitimation and justification for the very offense.Consequently, a hate crime occurs, according to the report, whenever "a person is victimised -killed, raped, abused, assaulted, threatened, insulted etc. -for being X" or, rather, for "being perceived as X" (FRA 2012: 19), while this perception in turn is used as legitimation for the attack.In short: "Victims of hate crime are abused for what they are perceived to be." (FRA 2012: 19) Furthermore, it is decisive that this perception or feature is not taken randomly, but rather "defines a 'difference in value' that affects the formation of […] society" (FRA 2012: 20).Hence, discrimination should not be mixed up with mere unequal treatment."Discrimination does not mean to treat others differently in a random fashion; instead, it means to enact and wield a distinction powerful enough to structure society."(ibid.)Discrimination, thus conceived, does not just mean to perceive and treat a person in a different way than others, but rather to attribute to a person a certain characteristic "that undermines the social status of that person" (ibid.).In other words, although discrimination leads to unequal treatment, the latter does not necessarily imply discrimination.If a robber only robs elderly women, s/he is not necessarily driven by hatred for elderly women, but because s/he thinks that they are easy victims.Consequently, any discrimination involves two interrelated aspects: 1. an introduction of a distinction that categorizes and denominates persons according to real or alleged features, 2. a utilization of this distinction as a justification to deny the person certain rights.This also explains why hate crimes often severely traumatize their victims: "because they cannot attribute their victimization to bad luck.Instead, they are forced", as the authors point out, "to accept that their social identity was targeted and that they remain at risk of repeated victimisation" (FRA 2012: 20).Consequently, hate crimes address not only the individual, but also any other person "likely to be [similarly] perceived".To take up again the aforementioned example: C's conduct has the potential to spread fear and intimidation to other persons whom he might label as gay.Therefor C's behaviour may affect other persons at risk of similar labelling.These persons may become aware that they too could have been or could be victimised, which is a process known as a vicarious traumatisation effect.(FRA 2012: 22) In other words, hate crimes are contagious, as it were: They are able to disseminate and to communicate a certain message that "transcends the context of the individuals directly involved", thereby affecting an entire social group that might be similarly labelled.Equally, the perpetrator "should not be assessed in isolation but as a representative of widespread attitudes" (FRA 2012: 22) that are prevalent in a society and that can be easily invoked and actualized: "In this way, hate crime 'speaks', and matters not only to the immediate victim.It also matters to persons who sympathise with the offender and whose biased attitudes the offender confirms and reinforces.It also matters to 'others' who understand that they are at risk of similar labelling and victimisation" (FRA 2012: 22).

The linguistic dimension of hate crime
What this passage clearly illustrates is the linguistic-symbolic dimension of hate crime.If hate crime "speaks", as the authors of the report claim, then hate crime constitutes a mode of speech or address.This raises the question of whether the linguistic aspect is an additional feature of hate crime or rather inherent to it.Indeed, the report allows for both interpretations.If we highlight the formulation that "C's bias motive adds an insult to the assault" (FRA 2012: 19), the linguistic quality seems to be a merely additional feature of hate crime, insofar as language adds a linguistic impetus to the exerted physical violence in form of an insult or verbal abuse.Accordingly, hate crime could be defined as a physical attack plus hate speech, whereas hate speech would be tantamount to hate crime minus physical attack.Hence, hate speech would be simply a subtype, special form or preform of hate crime.However, if it is true that hate crime "speaks" and that it is precisely the linguistic-symbolic moment that constitutes hate crime in its specificity, then language and speech can no longer be conceived as merely additional features of hate crime, rather they are essential to it.Thus, hate speech would be the general type and hate crime merely a subspecies of it.
This seemingly paradoxical relation between hate crime and hate speech points to the complex entanglement of language and violence as outlined above.As a result, any approach that presupposes a merely external relation of language and violence faces a series of difficulties.Firstly, it can explain neither the violent and injurious character of speech nor the linguistic and symbolic character of violence.Secondly, if speech, and the bias motivation expressed in it, were only an additional feature of hate crime, it would be difficult to consider hate crime as a criminal offence in its own right.Thirdly, by explaining the severity of hate crime solely by the motivation of the offender, the juridical prosecution of hate crime runs the risk of establishing a so-called "Gesinnungsstrafrecht", i. e., a criminal law that judges the accused less by their deeds, than by their attitudes, political convictions, or religious beliefs.Indeed, the authors of the report seem to be well aware of this danger when they stress: At the same time, the report argues -following the jurisdiction of the European Court for Human Rights (ECtHR) -that state "authorities have the additional duty to take all reasonable steps to unmask any racist motive and to establish whether or not ethnic hatred or prejudice may have played a role in the events" (ECtHR, Bekos and Koutropoulos v. Greece, cited in FRA 2012: 17).8Thus, contradiction seems inevitable: On the one hand, hate crime is defined as bias-motivated crime whose underlying motives must be unmasked and made visible; on the other hand, not the motivation itself, but only its explicit expression and articulation should be taken into account when assessing the criminal offense.
This contradiction can only be avoided, if we consider language and speech not as additional features of hate crime, but as inherent to it.At first, the fundamental linguistic-symbolic dimension of hate crime becomes obvious in its active attributive role.For, as we have seen, in hate crime, a person is attacked not randomly, but precisely for being perceived as X.In other words, hate crime identifies, categorizes, and labels persons according to real or supposed features such as sex, race, class, sexual orientation etc.This act of labelling a person as someone or something is in itself already a linguistic act of positing, an act of denomination and determination that attributes a social status to a person.This tendency to differentiate and to discriminate can go as far as to restrict the use of the term "human" exclusively for the designation of one's own social group.Claude Lévi-Strauss points this out with regard to his ethnographical studies when he states that often [h]umanity is confined to the borders of the tribe, the linguistic group, or even, in some instances, to the village, so that many so-called primitive peoples describe themselves as "the men" (or sometimes -though hardly more discreetly -as "the good", "the excellent", "the well achieved"), thus implying that the other tribes, groups or villages have no part in the human virtues or even in human nature, but that their members are, at best, "bad", "wicked", "groundmonkeys" or "lousy eggs".
Furthermore, as we have seen, hate crimes do not only inflict injuries on others, they also communicate a certain message on several levels and to different addressees: to the attacked individual and bystanders, to the social group the individual belongs to and to sympathisers of the offender as well as to society at large.In this vein the authors of the report claim that hate crime is not just an act of brute force, but a meaningful act that requires special attention: "Any analysis of hate crime has to take into account the meaning of actions and how this meaning impacts on social identities.Hate crimes cannot be understood unless one listens to what they say, with all forms of hate crimes essentially conveying a common message."(FRA 2012: 21) In its most general formulation this message is that the "the victim's rights matter less" (FRA 2012: 7), because certain features can be attributed to him or her.In other words, if it is justified for C to beat up D, because C perceives D as gay, it follows that C denies D's right to physical integrity as well as to non-discrimination and effective remedy.However, to deny a person certain fundamental human rights just because certain features can be seemingly attributed to him or her, implicitly means to deny him or her certain human features.In this way, hate crime and hate speech always dehumanize their counterpart at least to a certain extent.9 This is because we are essentially social beings that rely on linguistic acts of address and response for their social existence.In other words, we are not born as autonomous subjects capable of language and speaking; the individual is rather introduced to the social and symbolic order by acts of interpellation and address and thereby constituted as a recognizable subject in the first place.However, precisely "because personhood is initially constituted by others in terms of language, […] people remain vulnerable to symbolic interaction" (FRA 2012: 20).Consequently, the acts of interpellation and address by which we are conferred certain subject positions have both enabling as well as injurious effects, both subjectivizing as well as desubjectivizing consequences.This is also the reason why even severe forms of injurious speech can have enabling effects up to a certain point.For we are not only subjected and determined by the names we are called, but they also give us a certain possibility of social and discursive existence (cf.Butler 1997: 2).
Hence, personhood and subjectivity remain constantly in the making and are always fragile and precarious.This also has implications for the traditional concept of violence.While most definitions of violence stress the violation and injury of a bodily self, it turns out that violence is equally directed against the identity and continuity of persons as agents of significant and binding actions.It is in this vein that Emmanuel Levinas writes that "violence does not consist so much in injuring and annihilating persons as in interrupting their continuity, making them play roles in which they no longer recognize themselves, […] making them carry out actions that will destroy every possibility for action" (Levinas 1979: 21).
Moreover, as already pointed out, the message that hate crime conveys is not only addressed to the victim and bystanders, but also to the social group the victim belongs to, to society at large as well as to sympathisers of the offenders.This is possible because due to its linguisticsymbolic structure hate crime is able to break with its context and to transcend the concrete scene of address.Moreover, if hate crime "speaks" and conveys a common message that must be listened to and for which civil society and governmental institutions must find an adequate answer, this is because the action and the message it conveys "is not 'drafted' in private but in common language" (FRA 2012: 22) on which it relies and which it employs.
What comes into play here, along with its linguistic dimension, is the iterative and citational structure of hate crime.In his deconstruction of the classical concept of writing in Western philosophy, Derrida demonstrates that no sign or mark (written or spoken) could work as a sign, if it could not be repeated or cited, if it were not able to "break with every given context, engendering an infinity of new contexts" (Derrida 1988b: 12).In this connection Derrida speaks of a general iterability (derived from Latin iter "again," and Sanskrit itara "other") that "ties repetition to alterity" (ibid.7) and that is valid not only "for all orders of 'signs' and for all languages in general but moreover, beyond semio-linguistic communication, for the entire field of what philosophy would call experience" (ibid.9).This also holds true for the most singular and "'event-ridden' utterances" (ibid.19), the performative utterances or speech acts, as Derrida points out in his critique of Austin's speech act theory.No performative could succeed, according to Derrida, "if its formulation did not repeat a "coded" or iterable utterance, or in other words, if the formula I pronounce in order to open a meeting, launch a ship or a marriage were not identifiable as conforming with an iterable model, if it were not then identifiable in some way as a 'citationʼ" (ibid.18).In addition, Derrida stresses the forceful, transformative as well as material-bodily dimension of performative speech.Just as the performative "does not describe something that exists outside of language and prior to it", but rather "produces or transforms a situation", equally "[c]ommunicating, in the case of the performative", cannot be reduced to the transference of an intention or a propositional content, but "would be tantamount to communicating a force through the impetus of a mark" (Derrida 1988a: 13).
Both aspects -the iterative-citational as well as the transformative-performative structureare equally constitutive for hate speech and hate crime.Both sides have to be taken into account, if we want to understand their specific injurious force.Thus, if hate crime and hate speech have the potential to harm and injure others, even beyond the concrete scene of address, this is not only due to the intentions that motivates them.Rather it results precisely from the fact that hate crime and hate speech -both as speech and as act -invoke past injuries and traumas that are historically sedimented in the norms, structures and conventions of language and social institutions.Or, in the words of Butler: Any racist speech or action "works through the invocation of convention; it circulates, and though it requires the subject for its speaking, it neither begins nor ends with the subject who speaks or with the specific name that is used" (Butler 1997: 34).
This also explains why "the physical harm resulting from violence motivated by prejudice is often less significant than the powerful accompanying sense of violation and humiliation" (FRA 2012: 20).One only needs to think of the humiliating practices of scrubbing off the sidewalks to which Jews were forced by the Nazis in Austria after the so-called "Anschluss" in 1938.Here, the actual physical violence seems to be relatively insignificant in comparison to the exerted symbolic violence.The severely harmful and traumatizing effects of the inflicted violence as well as the injuries that the victims and the whole Jewish population suffered by the implicit threat of reiteration was apparently grounded less in the physical violence than in its linguistic-symbolic character.
In other words, precisely by conceiving hate crime as a specific kind of speech and address, its injurious and dehumanizing force comes into sight.However, this explanation is not intended to relativize the often extreme physical violence that hate crime inflicts on its victims or to exempt the offenders from their responsibility.On the contrary, precisely by taking into account the both act-and speech-like character of hate crime, its specific violating and destructive character (which takes effect far beyond the concrete scene of address) as well as the accountability that goes with it (and which is not reducible to the motivations of the offender) comes into view.
This also provides us with an answer to the above-posed question as to whether hate speech is a subtype of hate crime or whether it is an additional feature that turns "ordinary" crimes into hate crimes.It now becomes clear that both answers fall short of the mark.Neither is it possi-ble to subsume or reduce hate speech to hate crime, nor is language a merely additional feature of it.Instead, hate speech as well as hate crime have to be considered as speech and as action, i. e., as a speech act in the strict sense of the word that is intertwined with certain actions and activities as well as an act that expresses a particular meaning.Furthermore, if hate crime "speaks" and conveys a common message that must be listened and answered to, remarkably language itself seems to provide the vocabulary to illustrate the effects of physical violence.While it is often argued that we do not have a proper language to describe forms of linguistic violence and injury, but are "forced to draw [our] vocabulary from physical injury" (Butler 1997: 4), it now seems to be the other way round: We have to draw our vocabulary from language to account for the specificity of hate crime.

Hate crime as violence against speech itself
This leads me to a further important aspect: Hate crime does not only imply a fundamentally linguistic-symbolic dimension, rather it is directed against the possibility of language and speech itself, against the possibility of speech to make sense and to perform significant actions.This is, indeed, a seemingly contradictory claim.For how is it possible to claim on the one hand that hate crime "speaks", and on the other hand, that it undermines the possibility of speech itself?This question points to a fundamental paradox inherent to the relation of language and violence.For it can be argued that violence is both the condition of possibility and the condition of impossibility of speech: It is, so to speak, the condition of possibility of speech, insofar as neither sense nor meaning would be possible without an act of naming that makes something recognizable, identifiable, or addressable as something in the first place (cf.Dreisholtkamp 1999: 146).It is the condition of impossibility, insofar as it undermines the very possibility of speech itself, namely, inasmuch as it aims at making us speechless, depriving us of the possibility of speaking out and being heard.This paradox becomes particularly manifest in hate crime and hate speech.On the one hand, there is no meaningful speech and action at all without some act of identification and labelling; on the other hand, a "speech" in which every meaning is univocal, every answer prefabricated, and every dissent impossible would be, as it were, no speech at all, but a mere exchange information (cf.Posselt 2011).
In fact, we experience acts of violence -physical or symbolical -always as a way of making us speechless.Violence muzzles us; it deprives us -at least for a short moment -of our voice and our word, of the possibility of response, talking-back, and counter-speech.This does not mean that hate crime or hate speech deprives us of our physical ability to speak.However, it can mean that it dictates words and fixes meanings, that it imposes certain interpretations, that it forces us to speak in a certain way or from a certain position, as, for example, in the extreme situation of torture during interrogation (cf.Liebsch 2007: 32;Lyotard 1988).Or it can mean that it expels us entirely from the realm of representation and visibility, with the effect that our utterances become more similar to meaningless noise than to articulated words and legitimate claims (cf.Rancière 1999: 29) -as Aristotle argues in his Politics with respect to slaves and barbarians.For, although slaves are indeed human beings, they only slightly differ from animals, as Aristotle argues, insofar as slaves participate in speech and reason (logos) "so far as to apprehend it, but not to possess it" (Politics 1254b).This is also reflected in Primo Levi's famous claim that wherever violence is done to humans, there is also violence done to language (Levi 1988).In the extreme, this may mean to lose language itself, as Levi famously puts it in his recalling of his detention in a Nazi concentration camp: Then for the first time we became aware that our language lacks words to express this offence, the demolition of a man.
[…] Nothing belongs to us anymore; they have taken away our clothes, our shoes, even our hair; if we speak, they will not listen to us, and if they listen, they will not understand.They will even take away our name […].(Levi 1979: 21) In other words, if humans are essentially speaking beings that require language for their social survival, and if "only in the individual does language receive its ultimate determinacy", as Wilhelm von Humboldt emphasized (Humboldt 1988: 63), then any violence that is directed against another person is also directed against language and speech itself.If we further take into account, as outlined above, that human beings are fundamentally social and linguistic beings that rely for their social survival on discursive acts of address and recognition, it becomes clear that dehumanization in its worst outcomes cannot be reduced to the deprivation of specific human features or particular fundamental rights -as the right to freedom, freedom of movement, thought, or expression.On the contrary, as Hannah Arendt argues, "the abstract nakedness of being nothing but human" would be the "greatest danger" for those who are subjected to it (Arendt 1979: 300).Rather, it means, as Arendt puts it, to be deprived "of a place in the world which makes opinions significant and actions effective", which amounts to "the loss of the relevance of speech […] and the loss of all human relationship" (Arendt 1979: 296f).This loss can manifest itself on all social levels: On the individual level, the subject as a bodily and speaking being is harmed, silenced and made invisible, on the group level, the dialects, sociolects and idioms of a group are suppressed and delimited, and on the societal level, language as a social institution is undermined in its ability to sustain the public space of appearance and the web of human relationships (cf.Arendt 1989).
Indeed, as outlined above, the main objective of the FRA-report is to make "hate crime visible", since victims and witnesses of hate crimes very often do not recognize hate crimes at first sight, when they finally do, they "are reluctant to report them" (FRA 2012: 7), and if they report them, these crimes frequently remain unacknowledged.However, the invisibility and non-traceability of hate crime is not a mere coincidence.Rather, as the previous considerations demonstrate, it is a constitutive feature of hate crime itself, insofar as it aims at disenfranchising and silencing those it attacks.Moreover, the objective of hate crime is not only to deny a person or a group certain fundamental rights, but precisely to make impossible any claim to these rights.Hence, the physical violence of hate crime does not only aim at the physical destruction of its victims, but also at the possibility to tell about it, to bear witness to it and to articulate the experience of that violence as violence.However, since not only hate speech but also hate crime depends for its functioning on its message being understood by its victims, it must acknowledge a minimum of humanity in its victims.Thus, hate crime -even in its most cruel, destructive and dehumanizing forms -requires and maintains the subjectivity of the other (cf.Seitz 2016).10This is, according to Emmanuel Levinas, the "logical absurdity" of hatred: The one who hates seeks to be the cause of a suffering to which the despised being must be witness.To inflict suffering is not to reduce the Other to the rank of object, but on the contrary is to maintain him superbly in his subjectivity.In suffering the subject must know his reification, but in order to do so he must precisely remain a subject.(Levinas 1979: 239) However, if hatred aims at the destruction of the other while requiring that the other is maintained in its subjectivity, then this paradoxical structure might at the same time open up the possibility to respond to it.For, as we have seen, every act of address or interpellation does not only subject the other, but also subjectivizes him or her by assigning a specific subject position.Thus, even though violence is directed against the possibility of speech and discourse itself, violence is not the end of speech.Or in the words of Levinas: "Violence does not stop Discourse", for it "is produced only in a world where I can die as a result of someone and for someone" (Levinas 1979: 239).

7
How to respond to hatred?
This brings me back to my initial question concerning the relation of language and violence as well as to the underlying question as to whether hatred can speak.I have chosen the approach via the problem of hate crime, precisely in order to demonstrate that not only language and speech comprise a fundamentally violent dimension, but that even the most brutal crimes need to be reconstructed as a form of "speaking", if we wish to understand their particularly destructive and devastating force.However, if hate crime or hatred "speaks", this is not because it "uses" language in the strict sense of the word, but because it is structured as a form of address that implies an act of labelling and denomination.As such a multi-addressed violence, hate crime breaks with its "original" context and conveys a generally intelligible message that is directed against the victim as an individual and as a member of a social group as well as against society in general.This is possible because, as a form of address, hate crime evokes, reiterates, and actualizes past injuries and traumas that are historically sedimented and memorized within language (in its structures, norms, and conventions), in the institutions of a community as well as in the bodies of the individuals.
However, if hate crime always also implies the subjectification of the other as well the general structure of iterability, then this also opens up the possibility of reframing and reiterating its message in a different way.This also applies to the fundamental rights in question.In fact, human dignity and the right to non-discrimination can never be taken for granted, but are in need of their continuous confirmation and reaffirmation: "While the law deems human dignity inviolable, in reality it is fragile and precarious, in need of protection and defence."(FRA 2012: 21) Hence, hate crime, understood as a multi-addressed violence, requires an equally differentiated response and reassertion.
This assertion is addressed to the victim, whose right to non-discrimination and human dignity is reaffirmed; to the offender, who is made responsible and called upon to respect the victim and the victim's rights; and to the public.In our societies, it is courts that must perform this symbolic function, as they have a monopoly on defining criminal offences.(FRA 2012: 22) However, this "monopoly on defining criminal offences", especially in relation to language, is not without dangers, as Butler points out in Excitable Speech.For if the courts have the final say on which utterances constitute a criminal offence and which do not, what is sayable and what should be excluded from the realm of the expressible and the audible, the injurious force of language is not only confirmed, but also any possibility of resistance and counter-speech is delimited from the outset.Hence, Butler is sceptical about the tendency to delegate the final decision about injurious speech to juridical or political authorities.Instead, she asks for "a theory of linguistic agency that provides an alternative to the relentless search for legal remedy" (Butler 1997: 15).This is not to say that the legal definition and prosecution of hate crime and hate speech is neither possible nor necessary.But it points to the fact that any attempt to define hate crime and hate speech "can only be preliminary and subjected to history, because the injury is not inflicted by the words themselves, but by the entire scene of address" (Butler 2006: 260, my translation), as Butler emphasises in her afterword to the German edition of Excitable Speech.However, this "entire scene of address" was central to our attempt of reconstructing hate crime as a multi-addressed violence that, due to its iterative structure, transcends its context and takes effect on several levels.
If we now return to the Šešelj judgment mentioned at the beginning and revisit the reasoning of the tribunal, it becomes obvious that it is precisely the description of the context and the construction of the scene of address that is at stake.Accordingly, the Chamber by a majority, Judge Lattanzi dissenting, "could not rule out the reasonable possibility that [Šešelj's] speeches were made in a context of conflict and were meant to boost the morale of the troops of his camp, rather than calling upon them to spare no one" (ICTY 2016d, my emphasis).Equally, the majority did not find enough evidence to prove beyond all reasonable doubt that, in calling upon the Serbs to "cleanse" Bosnia of the "pogani" or the "balijas", Vojislav Šešelj was calling for ethnic cleansing of Bosnia's non-Serbs.The majority, in fact, believes that the evidence supplied by the Prosecution is not sufficient to exclude the possibility, in view of the context, that in making this appeal, Vojislav Šešelj was rather participating in the war effort by galvanising the Serb forces.
(ICTY 2016d, my emphasis) These formulations are particularly intriguing.For, even though the majority is not denying the performative dimension of speech and the connection between saying and doing, it construes the "context" in such way that any effects and any responsibility that go beyond the concrete situation is dismissed.In contrast, in her "partially dissenting opinion"11 , the third judge Lattanzi reconstructs the context and the scene of address in a completely different way.
She not only stresses the dehumanizing effects of Šešelj's speeches, but also refers to the historical context in which these utterances are embedded and which is clearly evoked by Šešelj himself (ICTY 2016c).
In fact, the previous considerations have repeatedly highlighted the historical and iterative character of hate speech and hate crime.For every performative requires for its functioning that it can be reiterated, that it evokes and affirms the historical sedimented norms, structures and conventions that ultimately enable it.This becomes particularly apparent, when we recall ISSN 1615-3014 22 that the one who performs hate speech or commits hate crime never acts 'alone', but as a representative of social structures, attitudes and dispositions that his or her (speech) acts invoke and actualize.Hence, hate speech and hate crime are never simply expressions of an individual, but always also a reflection and reiteration of "inbuilt tendencies and predispositions of societal structures" (FRA 23) that make those acts possible.
However, this does not mean to excuse the offender or to diminish his or her responsibility.
On the contrary, it rather calls for a strong concept of responsibility that is linked to the iterative structure of hate speech and hate crime as a form of "speaking".Precisely because speaking is always a citation, precisely because every speech act invokes and reiterates the norms, conventions, traumas, and exclusions that are sedimented in language and in the bodies, we are responsible for our utterances.In other words, neither the citationality of speech nor the anonymity of social structures and institutions can release us from our responsibility.Rather, we are responsible for our utterances and subsequently for our actions, as Butler points out, "precisely through the citational character of speech.The speaker renews the linguistic tokens of a 'community' reissuing and reinvigorating such speech.Responsibility is thus linked with speech as repetition, not as origination" (Butler 1997: 39;cf. Posselt/Flatscher 2016: 249).
Against this background, it is hard to conceive how Šešelj could be cleared of all charges and absolved from any responsibility.However, it would be too simplistic and not without risk just to blame the courts.For, if responsibility is not only linked with authorship, but also with reiteration, as the analysis of hate crime and hate speech clearly demonstrates, then this responsibility has to be assumed by all social actors -from the courts across politics to civil society.This requires at least a double task: On the one hand, we need to scrutinize, evaluate and prosecute those forms of acting and speaking that humiliate, disenfranchise and silence others -thereby posing a fundamental threat to any democratic society (a matter to be resolved primarily, but not exclusively by the courts).On the other hand, we must take care not to censor and restrict language, but to maintain it as a resource for future modes of speaking and acting in concert.Only if we succeed in meeting both challenges equally, we will be able to respond to hatred effectively.

13
Person A, in the midst of a heated argument, beats up and injures Person B. In a second incident, C publicly injures D in a manner very similar to the way A treated B […], except that C's motivation is different from A's.By uttering homophobic slurs, C makes it clear that he is attacking D because he is gay or, more precisely, because C believes D is gay.