“After twenty-five years, most men passed with indifference in front of the tomb of their assassinated father; but after a hundred years the spoliated generations will still feel hatred and rage at the sight of a field of which their family had been deprived.”

(Edmund Burke) Footnote 1

1 Introduction

From the misogynic to families in vendetta blood-feuds, hate-mongering populists and Rohingya-killing extremists, hatred not so much erupts as it lingers, often for a lifetime and, indeed, across generations. Moreover, hatred seems to involve more often than not some form of social typification, intergroup antagonisms or collective settings, rather than interpersonal forms of aversion. The title of this paper alludes to these two observations: first, that hatred tends to robustly linger and habitualize even in the face of long-faded harm and healed wounds (‘in hate we trust’); secondly, that it typically involves a certain collectivizing dynamic (‘in hate we trust’). I shall corroborate these two observations and show how the latter dynamic reinforces, and indeed explains, the habitualization of hatred.

I will do so in a somewhat unconventional manner, at least considering standard social-psychological and the available philosophical literature on the topic. Typical social-psychological discussions explore hatred as an affective mechanism underlying prejudices, dehumanization, political or moral exclusion and intergroup violence, with repeated references to the famous Milgram-experiments.Footnote 2 And work in philosophy of emotions has mainly concentrated on the distinction between hatred and cognate aversive emotions such as resentment, malice, anger, disgust or contempt (e.g., Ben-Ze’ev 2000; Brudholm 2010), or on the issue of whether hatred has any morally legitimate function (Murphy and Hampton 1988; Solomon 1995; Murphy 2016; Schmid forthcoming). In contrast, I shall exclusively focus on the peculiar affective-intentional nature of hatred. With few exceptions from the phenomenological tradition, this has been surprisingly little discussed. Yet, I contend that any understanding of its temporal and socio-psychological dynamics will ultimately turn upon grappling the distinctive structure of the affective intentionality of hatred. Incidentally, I think that any proper account of the normative (in-)appropriateness of hatred must also begin with just such an analysis; but this is something I cannot argue for here.

I will proceed as follows: I begin by describing hatred as a complex, but sui generis affective attitude (sect. 2), and dwell upon its affective-intentional structure (sect. 3). Next, I zero in on the tendency of hatred to overgeneralize its targets, and elaborate its formal objects and affective focus. Here, I show that the affective focus of hatred—i.e., what specifies the target as hateworthy—is essentially ‘blurred’ (sect. 4). I then show how the overgeneralizing tendency typically transmutes into a collectivizing mechanism and discuss the underlying socio-ontological structure in terms of what I call the ‘negative dialectics’ of hatred (sect. 5). I argue that we need (to maintain) our antagonistic attitude towards an outgroup to create and affectively reinforce ingroup alignment. In the final section (sect. 6), I address the question of what lends hatred its extreme affective weight—if not an affective concern for particular features of the target or some specific phenomenological properties of the attitude. I claim that hatred gets its affective powers for free, as it were, namely from a commitment and, in particular, a shared commitment to the attitude itself. Ultimately, I suggest that this is the reason why hatred becomes robustly habitualized and is best characterized as a shared habitus.

Before entering the discussion, however, two central caveats are in order: The first concerns the distinction between interpersonal or person-focused hatred, on the one hand, and social-identity- or group-based, or what I call ‘collectivizing’, hatred, on the other. In this paper, I will only focus on the latter kinds. In doing so, I don’t want to deny that hatred may be directed at individuals nor that person-focused hatred may sediment itself into a robust, maybe life-long sentiment that defines the affective biography of individuals. Think here of a betrayed and repeatedly ridiculed lover or a massively exploited servant, who hates his master. Arguably however, the most prevalent cases of hatred, exhibited by many millions repeatedly in the course of history, are precisely instances of collectivizing hatred, and it is also these that seem to be the least malleable. In any case, collectivizing cases most saliently exhibit the distinctive affective intentionality of hatred, or so I shall argue. In particular, I shall show that there is a sort of overgeneralization tendency already inherent in the interpersonal cases that makes hatred not only prone to habitualization; moreover, when the target of hatred is located in a broader social or intergroup context, this tendency eventually crystallizes as a form of stereotypical essentialization that I specify in terms of ‘collectivization’. Roberts arrestingly describes this complex process in his brief analysis of hatred:

“The abused wife may give a long list of reasons for hating her husband, but they come to trait ascriptions: He is a nasty, inconsiderate, selfish person. And so may the Palestinian, or the Nazi, for hating the Jew; the offenses, if they were there at all, have by now been compiled into a trait: being a Jew. Thus repeated episodes of anger, in which badness is ascribed to the object on account of culpable misdeeds, may lead to hatred by a kind of accumulation and abstraction: The badness of the object gets distilled from his offenses and acquires the independence and perduration of a personal essence.” (Roberts 2003, 251)

The paper can be read as an elaboration of this description, and in particular of the sense in which hatred ‘essentializes’, ‘abstracts’ and ‘perdures’. Moreover, I hope to establish that the reason why misogyny, racism, vendetta-cultures or genocides are the paradigmatic realizations of hatred is no historical contingency, but inherent in the very affective-intentional nature of hatred.

The second, related, caveat concerns the notion of ‘collectivization’. Like other affective states, hatred too can, under suitable circumstances and given certain robust requirements of affective sharing,Footnote 3 be realized jointly. In other words, I take it that a number of individuals can properly speaking hate together, such that the very instantiation of hatred can be jointly realized among individuals. However, in this paper, I will not say anything about collective hatred in this sense. Rather, I will restrict the use of ‘collectivization’ to the general tendency of hatred to blur the socio-ontological status of the target, transposing the hated properties from individuals to proxies or groups, and to what I shall elaborate in terms of the negative dialectics involved therein.

2 Hatred as an affective-intentional attitude

Hatred is an extreme and extremely messy affective phenomenon. Arguably, it lies at the most extreme end of a spectrum of antagonistic emotions humans are capable of. Indeed, hatred is massively interrelated with, cannot be understood without, and is usually not experienced without a whole package of previous or concomitant experiences of “affects of aggression” (Demmerling and Landweer 2008, 287–310; Landweer forthcoming). Thus, a sort of antipathetic affective muddle characterizes typical experiences of hatred, including, to a greater or lesser degree of prominence, anger, malice, indignation, resentment, contempt, malicious joy over harm to the hated object (Schadenfreude), vengeance, and often jealousy or envy as well as visceral feelings or more mediated expressions of disgust (see also Aumer et al. 2015). However, hatred can neither be reduced to any one nor to a summation or intensification of these affects of aggression; rather, it must be distinguished from these in various respects. Even if embedded in experiences of these cognate affects, hatred has sui generis sources, objects and aims. For instance, there are the important differences between anger and hatred regarding their objects: anger is typically a reaction to a specific harm or particular actions, whereas hatred, as we shall see, is directed at global personal traits, ideologies or persons as proxies for social groups or categories (sec. 4 and 5). Anger and hatred also have different goals or teloi: corrective measures to alleviate specific negative features or actions in anger versus destruction, annihilation or a desire to dissociate from the object in hatred. Furthermore, hatred typically involves an asymmetric power relation between the parties, which is the inverse of many cognate affects of aggression. Thus, whereas contempt usually involves a power relation in which the subject is more powerful than the targets, often, hatred is directed towards those towards whom one feels powerless and is yet dependent upon (see, however, Solomon 1993, 264–265).Footnote 4

That hatred lies at the extreme end of a spectrum certainly does not mean that it’s the last affective resort if those other affects of aggression are of no avail. If hatred is an intensification of kindred affects of aggression at all, it is not an intensification or aggregation of the involved antipathetic feelings. Rather, what is at issue is an increasing differentiation of the type of antagonistic social relations those affects constitute and, in particular, of the generalization of their intentional objects. Hatred, I shall argue in this paper, lies at the end of that spectrum inasmuch as it is the most complex—dialectical—type of social antagonism and also involves the most generalized socio-affective antagonism, a sort of negative intensification of the affective relation to the Other writ large.

Hatred is also extreme in the sense that it is extremely rarely experienced or acknowledged as such. However much it may determine individuals’ lives and, in fact, the course of history, the explicit expression of hatred, action-tendencies or overt behaviour based on hatred, or the acknowledgement of a disposition to hate by those who bear it is very rare, at least when compared to more socioculturally or morally legitimate affects of aggression and, in particular, to indignation or anger. This is probably the chief reason why general treatments of hatred in psychology and in particular empirical studies are scarce. Notwithstanding so-called hate-crimes and the increasing prevalence of hate-speech in (social) media, studying interpersonal hatred empirically requires very specific laboratory settings (see Sternberg and Sternberg 2008, 71–77; Aumer-Ryan and Hatfield 2007; Aumer et al. 2015).

Now, there is certainly no shortage of suggestions as to how to conceptualize hatred. Quite the contrary: widely differing definitions and taxonomies abound in psychologyFootnote 5 and philosophy. In particular, hatred seems to escape the standard categorization within the philosophy of emotions between episodic emotions, moods, and affective dispositions or sentiments like probably no other affective state, save its positive correlate, love. While there seems to be only minimal consensus that hatred is not an ordinary episodic emotion, but rather resembles dispositions or sentiments,Footnote 6 an increasing number of authors characterize it as an ‘attitude’ (Ben-Ze’ev 2000; Brudholm 2010).

In one of the recent accounts, Ben-Ze’ev, for example, characterizes hatred as a “global negative attitude” toward someone, which often, but not necessarily, turns into a “long-term sentiment” (2000, 380–381): It might also “stem from a specific and hence short-term state; here, hate will be limited to the specific case and thus will usually be brief” (383). Ben-Ze’ev, however, fails to give us an example of such cases, and one is left wondering whether such specific and short-term hatred is hatred after all, rather than for example anger, dislike or reluctance. Will I really hate Peter if I learn that he once told an embarrassing story about me to his partner and that they were laughing behind my back? Aren’t anger, resentment, possibly feelings of humiliation or betrayal or even contempt rather the more likely (and more warranted) emotional reactions in such casual incidents? Hatred, it seems, needs more robust traction to even get off the ground, let alone develop into and eventually be maintained as the global attitude it is. What is right in Ben-Ze’ev’s account, however, is his construal of hatred as a global attitude, and I will follow his suggestion. Specifically, hatred seems best characterized as an enduring affective-intentional attitude, which may or may not be explicitly expressed, consciously acknowledged or actively endorsed.

An important feature of this notion of attitude is that it cuts across the active/passive distinction, and also marks the distinctive intentionality and relationality of the phenomenon. Attitudes don’t just passively (affectively) register certain external facts or occurrences (as bodily sensations do), nor do they simply react to them. Having an affective attitude is a response, or a form of position-taking, to affective significances for a subject. In the case of interpersonal or social emotions, such as hatred, the affective significances concern other subjects. But what exactly ‘in’ others does the attitude of hatred register or evaluate? And in what sense is this attitude affective, if it is typically not accompanied by any felt (bodily) sensations? In order to answer these questions, we need to complicate the picture and elaborate on the intentional and formal object and the affective focus of hatred.

3 The affective intentionality of hatred

So far I have mainly considered the attitudinal or act-character of hatred. We have seen that hatred as an attitude exhibits a number of distinctive features: It is an extreme and extremely rare affective attitude, which is more robust than the emotional episodes it realizes or comprises. As an attitude it may span more than the life of individuals and be transmitted cross-generationally. It is also more complex than discrete emotional episodes as it is typically interwoven with kindred affects of aggression. Its affective phenomenology is hence characterized by a relative indistinctness vis-à-vis cognate affects of aggression, yet it is a sui generis affective-intentional attitude with its proper intentional target, object and telos.

That hatred is a robust intentional attitude doesn’t mean that it is not also an experience with its proper affective phenomenology. But according to the phenomenological account endorsed here, we cannot even begin to understand one without the other: We cannot grasp the affective phenomenology of hatred without understanding its intentional structure. (And this is true of all emotions). To elaborate the distinctive affective-intentional structure of hatred, then, is the task before us.

Affective intentional attitudes, like any other intentional states, are acts directed at objects. More precisely, according to a prominent account offered by Helm (2001), they are “felt evaluative” acts directed at objects of affective “import”. But the intentional structure of affective intentional attitudes is more complex than that of non-affective ones. It not only involves specific act- and object-components. Emotions as intentional phenomena, or with Peter Goldie’s by-now classic capture, as “feeling-towards” (Goldie 2000), also include (i) (bodily) feelings, (ii) expressive components, (iii) a telos, and (iv) an intriguingly complex internal differentiation with regard to their objectual components.

Let me briefly review these in turn, before dwelling on what hatred is directed at, or its objectual side:

(i) The bodily feeling components of hatred, or bodily changes that are felt, notoriously elude description.Footnote 7 Aristotle, for example, famously maintains that hatred is always cool in the sense that it lacks the pangs and pains of other passions, and notably the feeling of pain that accompanies anger for example. He tells us that hatred is not accompanied by pain, but it is not quite clear whether this means that hatred is not (bodily) felt at all or that it is not felt as painful (Aristotle, 1382a1-1382a16). Whatever exactly Aristotle’s view was,Footnote 8 I contend that hatred as an attitude need not be, and typically is not, accompanied by any discrete bodily changes. This is not distinctive of hatred. After all, I doubt that every complex, especially moral, emotion is necessarily accompanied by any distinct bodily feelings (think of envy, Ressentiment, loss of hope, etc.).

But what about other experiential components that are not visceral feelings, sensations or awareness of bodily changes? Or is hatred not necessarily felt or experienced at all? First, notice what one is typically (not) aware of or acquainted with in hatred. Typically, one does not experience or cannot identify the onset or the formation of hatred. Furthermore, one often hates not only persons who are long deceased or groups that don’t exist any more but also those who have never harmed one personally or whom one has never personally encountered and is not even expecting to encounter in the future. Furthermore, one often doesn’t know the reasons or motives that have led one to hate. The fact that one often doesn’t ever encounter the targets of one’s hatred helps obscure one’s very motives. Finally, unlike in anger or resentment, one is typically unable to precisely identify not just the perpetrators of an alleged harm but also even the harm or threat that hatred is supposed to evaluatively react to. All this is surely not enough to fully explain why the very experience of feeling hatred is often hardly noticed or not present at all (cf. Ben-Ze’ev 2000, 383). However, these factors lend some credence to the often-repeated claim that hatred is a ‘cool’ affect, lacking passionate engagement with ensuing bodily reactions. This is often contrasted with kindred but different antagonistic attitudes such as anger or furious rage. Yet, I contend that the often-employed heat-metaphorFootnote 9 is altogether not helpful. Merely adding some independent affective ingredients (disgust, anger, etc.), or some awareness of one’s bodily conditions, won’t help either to characterize the affective nature of hatred. Rather than relying on any such “add-on theory” (Goldie 2000), I will argue that we can only understand the feeling dimension of hatred in terms of analyzing its specific affective-intentional directness and, in particular, its peculiar affective focus. Before doing so, let me say a few words about the expression and telos of hatred.

(ii) The expressive components are obviously manifold. They range from private tirades of hatred to public hate-speech and from aggressive or violent behaviour to mass-murder, but they may be manifest in only very subtle forms of social avoidance or even be fully covert. In either case, as Sternberg and Sternberg (2008) rightly suggest, a distinctive expressive dimension of hatred consists in dynamic narrative structures. They describe the targets as anathema and eventually present them as allegedly planning hostile actions or succeeding in realizing them. Such narratives are realized in different types of stories, such as the “stranger”, the “impure”, the “greedy enemy story”, the “barbarian” or the “enemy of God story” (ibid., 83–98). The stories serve a rationalizing and justificatory function for the haters whose self-esteem is threatened. Sternberg and Sternberg construe their claim that “hate arises from stories” (ibid., 78) as a “developmental” one, rather than viewing the story-character of hate as an expressive feature. However, I allege that their claim can also be read as the necessary, structural-expressive component of hatred. Whether or not hatred translates into action or manifest discourse, it always expresses itself publicly or in the ‘solitary’ hater either explicitly in certain stories or in the form of narratively structured ruminations.

(iii) Affective attitudes do not merely passively evaluate persons or register evaluative properties. Similar to conative attitudes (desires, wishes), they are also goal-directed and normally motivate us to act in a way so as to elicit changes in other people or make the world conform to them. I have briefly touched upon the fact that hatred does not involve corrective or retributive measures, such as changes in the negatively evaluated features of the target, restoration of justice or reparations for harm caused. Rather, the typical telos of hatred is either the social or moral exclusion of the target or, in extreme cases, its annihilation through murder, genocide, etc. At minimum, there will often be a wish that the target personally suffers or is in some way harmed, ridiculed, or humiliated. But none of these are necessary. Hatred is again special, as its telos is far from clear-cut. Indeed, it is not clear whether it always has a proper goal. Hatred may linger without ever resulting in any particular action or behaviour.

(iv) What is hatred targeting? What is the object of hatred? And what is the act evaluating and affectively responding to? These three questions all concern the broadly speaking objectual side of hatred. But we have to carefully distinguish these three aspects of the object of hatred. According to a widely accepted view in the philosophy of emotions (see esp. Helm 2001, 2017), the objective side—towards which emotions are directed or to which they are ‘felt’ evaluative responses—can be considered as (a) their target, (b) their formal object and (c) their focus.

(a) The target of an emotion is any correlate object that a given emotion is directed at and evaluates and that elicits an affective response. When alt-right activists air their hatred, it is, say, immigrants that are their targets. But the target and the way hatred is intentionally directed upon it, is different than for other emotions and sentiments and are special in various ways. First, hatred has an exceptionally clear target-orientation. As Kolnai, one of the most intriguing phenomenologists of hatred,Footnote 10 observes, targets of hatred are more “clearly contoured” or “carved out” (schärfere Heraushebung des Gegenstandes durch die Hassintention) compared to other sentiments, and in particular to love (Kolnai 1935, 123). Kolnai points to the “essentially substantival” (“entity-directed”), as opposed to the “adjectival” (“feature-directed”), character of the intentional directedness of hatred (Kolnai 1998, 591). Hatred is not directed at specific or even global features of the target (e.g., a person’s actions, plans, intentions, values). Rather, as we shall see, hatred intends the target as a whole, or globally. But however global its focus, hatred tracks only separate, discrete and extreme, allegedly representative blocks of (hateworthy) properties in the other, certain “historical points of orientation” for the intentional act to latch onto (Kolnai 1935, 124). Using contemporary metaphors, we may contrast the brute on−/off-focusing of hatred with love’s evaluative disclosure of its objects, which can be seen as “analogue”, ever more fine-grained, continuous process (cf. Nozick 1989). Secondly, and correlatively, acts of hatred exhibit a certain intensification of the intentional directedness upon their targets.Footnote 11 Thus, the intentional act is also characterized by a more “pointed attention (zugespitztere Hinwendung) (…), a distinctive, historical attunement to the object” (als einmalig, als historisch charakterisierte Hinspannung)” (Kolnai 1935, 123). In hating, then, one is affectively fully invested, as it were, in the target as such and without caring much about its specific features. As we shall see, this is reflected in a peculiar commitment to the very attitude of hatred (sec. 6). Finally, unlike for other affective states, only very specific types of objects qualify as potential targets of hatred. Thus, one should be wary of the pitfalls of every-day talk of hatred that suggests that targets could be any entities. When people say that they hate spinach, their computers, cats, or their hometown, what they try to convey is strong aversion, passionate dislike, disgust, or moral anger or contempt rather than hatred proper. Recall that the telos of hatred is the (wish for the) moral or social elimination, humiliation and/or total destruction of its target, none of which is what people seem to mean when they say that they hate spinach, dogs, or their hometown. It is perfectly normal to register one’s strong feeling of aversion to these by referring to it as hatred; but we would be baffled, and possibly find it pathological, to learn that somebody is as globally, deeply and personally invested in her aversion—as hatred, I will argue below, indeed requires—so as to wish to see a cat suffer or all spinach disappear from the face of the earth. More importantly, as we shall also see later, hatred involves a generalizing and essentializing evaluation of its target as hateworthy or evil, viz. an evaluation of a sort that cats, spinach, etc. are simply not the appropriate targets for. Accordingly, hatred proper never targets things, non-human animals, situations, or non-social facts, but rather only individual persons, social groupings or evaluative properties that are in some sense or other person-centered or person-dependent. Although this is often pointed out in the literature, there is little precision when it comes to properly delimiting the potential types of targets of hatred. In the face of this, I suggest the following delimitation. Hatred targets either one of the following types of entities or some combination thereof: (i) persons; (ii) social groups; (iii) ‘types’ of persons, classes, or personalities or proxies for social groups and communities, and/or (iv) ideologies, ways of life, religions, abstract values, cultures or some other social facts.

Now I shall argue that, even if hatred straightforwardly targets concrete individuals, there the formal object and the affective focus of the attitude is such that there is an inherent tendency to overgeneralize. This is so both in interpersonal and collective cases. But when it comes to intergroup antagonism and conflict or collective contexts, this overgeneralizing tendency is transmuted into a process of social typification and eventually the individuals thus targeted become ‘collectivized’. Consequently, in such collective contexts, the target is focused in such a way that it tends to oscillate between its status as an individual and as a proxy for social groupings or social facts. This, I claim, is so even if the target of such collectivizing hatred is actually a particular individual. To corroborate this arguably bold claim we need to understand what it is in the target-objects that hatred tracks, or focuses on as affectively relevant. In other words, we need to understand what the formal object and focus of hatred is.

(b) Ever since Kenny’s introduction of the notion (Kenny 1963, esp. 193 f.), it has become customary in the philosophy of emotions to distinguish between the intentional object of an emotion understood as the target and its formal object (cf. de Sousa 1987, 121–123; Helm 2001). The formal object is still an intentional object, but it is not the object simpliciter that a particular emotion is directed at. Rather, it is the correlate that the type of emotion tracks. It is in virtue of the formal object that a particular emotion construes the target (e.g., refugees) as having a particular type of evaluative property (e.g., odiousness). Accordingly, it is the formal object that individuates the type of emotion across its multiple instances. Hatred can be directed at this or that person, at a group, an ideology, etc., but the type of evaluative property that it picks out across those objects is identical, viz. their hateworthiness.

(c) If the formal object of hatred is hateworthiness in the abstract, what makes the particular target really hateworthy for the hater, or what gives the hateworthiness its subjective affective weight? What makes it so that the evaluative property of the target is not just considered abstractly or cognitively, judged to have a particular (negative) value, but matters to the subject, or has import for her? Helm gives a convincing account of how to specify formal objects of an emotion in terms of the import they have for the subject, namely by introducing the notion of an emotion’s ‘focus’. Thus, the affective weight that a formal object has for the subject is determined by how the target is carved out against the background of all that matters to the subject, and this is indicated precisely by an emotion’s focus. The focus of an emotion can thus be characterized as the background object of concern that links the evaluative property to the target and is hence definitive of the formal object of the emotion. Viewed from the perspective of subjective salience, the focus is what renders intelligible how and why the target is affectively significant for the subject, or why the emotion has the formal object it has. I may judge X (target) to be dangerous (formal object) for Y, without fearing X, not simply because I am for whatever reasons not affectively touched by X (this may or may not be the case), but because Y is of no import to me: it is not something for whose well-being I care and on whose behalf I am eventually prepared to act (cf. Helm 2009). To illustrate the tripartite distinction less technically: My worry that you put a dent in the pristine chrome-frame of my vintage bike by having an accident and my furious anger at the bicycle-handicraftsman, who, due to utterly careless handling, damages the frame, have different targets (you; him) as well as different formal objects and hence are different types of emotions (viz. worry regarding your inexperienced riding skills; anger at the experienced handicraftsman’s carelessness). But the focus of my worry and my anger is identical (the flawlessness of my chrome-frame). Thus, the focus is what makes intelligible how my anger and worry are related, and it also marks the intentional target as affectively salient, as something that my emotions are responsive to. Ultimately, then, the feeling of import that the focal object of an emotion has for the subject is what makes my intentional attitude towards that object a felt evaluation, instead of being a mere evaluative judgment.

But as shall prove crucial in a later step of my argument (sec. 5), the focus is not merely something that makes the affective import of the formal object salient (as the optical metaphor suggests); moreover, having a particular emotion essentially is being committed to its focus. Helm calls this an emotion’s “focal commitment” (Helm 2001, 2009). I shall argue that the bulk of the affective weight that hatred has is precisely and often exclusively generated not by the commitment to any particular focus but by the very entertaining of the intentional-affective attitude itself. Before showing that, however, we need to get clear about what the focus of hatred is in the first place. I will try to convince you in the next section that hatred has in fact no distinctive focus at all; rather, it is essentially ‘blurred’, as it were. And this is so, even if hatred has an exceptionally clear target-orientation, which is on a par with grossly visceral forms of fear or disgust.

4 The ‘blurry’ affective focus of hatred

The obvious reason for distinguishing target, formal object and focus of affective attitudes is that they do not overlap in any straightforward way. This holds for all emotions and sentiments, but it is especially salient in hatred. In hatred, not only are the target and focus not identical, the focus is typically not even informative of or doesn’t make intelligible—as it indeed should—how the formal object (hateworthiness) relates to the target. And this is, I want to argue, distinctive of hatred.

To begin with, consider an all-too familiar example: hatred towards refugees incited by a populist leader. Here the focus is typically ‘blurred’. It will typically not be anything related to individual refugees, nor to often very heterogeneous groups of refugees. The focal background of the attitude will rather be, say, the allegedly endangered ethnic or sociocultural homogeneity of the host country. But the focus here seems uninformative as to how the targets (individual refugees or refugee-groups) are related to the formal object (hateworthiness).

What makes the focus blurry, or less metaphorically speaking uninformative, is twofold: first, the formal object is indeterminate in the sense that it is all-too global (literally ‘not focused’); secondly, the very targets are not fixed but shifting—namely between individuals, groups, generalized social types, or proxies for groups. As we shall see, these two aspects are closely interconnected, and indeed just two aspects of the distinctive affective intentionality of hatred.

It will be helpful to start with the issue of the determinacy of the object of attitudes. Recall that I have characterized hatred as an affective-intentional attitude. Now, according to a plausible phenomenological proposal, there are attitudes that do not apprehend, conceptualize or even just register a given affective domain, event, person or state-of-affair as determinate. (These may be affective or cognitive, propositional or non-propositional attitudes). In short, not all attitudes have a determinate object. Notice that in certain cases such indeterminacy is not merely a lack or ‘privation’ of determinacy, but part and parcel of the very sense and essence of the given intention, which present a pure ‘something’ or ‘somebody’, as in ‘I need something’, or ‘something is stirring in him’ (cf. Husserl 1901, V, § 16; see also Mulligan 2017, 229).

I want to suggest that hatred is essentially an indeterminate intentional-affective attitude in this sense. But there is more to it. The indeterminacy of the affective focus of hatred is distinctive and has to do with its specific formal object and target. Hatred is indeterminate in the two already mentioned, interrelated ways: First, regarding its formal object, the indeterminacy is due to the fact that the formal object involves an all-too global evaluation of its target. Secondly, the properties that are thus ‘fuzzily’ picked out are also indeterminate. And they are indeterminate, I shall argue, insofar as they are attributed to an indefinitely shifting target, namely a target whose status is oscillating between (overgeneralized) individuals and (stereotyped) social categories, social groups or types. This feature will also allow for grasping the essentially fungible character of the targets of hatred. By this I mean not only that they are shifting between individuals and social types but also that they typically only have the status of being (stereotyped) proxies for social groups. Targets of hatred are hence replaceable by any other individual exemplifying the same stereotyped negative properties.

Again, these two aspects are not only interrelated but in fact just different aspects of the intentional indeterminacy distinctive of hatred. Indeed, I contend that we can only properly understand the often mentioned (and little explained) global character of the formal object of hatred if we get a firm grip on the overgeneralizing and shifting intentionality regarding its target. But let me begin by elaborating on these two aspects in turn, before showing how the latter explains the former.

(i) Virtually all philosophers writing on the topic agree that one of the distinctive characteristics of hatred that distinguishes it from cognate antagonistic affective phenomena such as anger, contempt, disgust, etc. is that its negation is all-encompassing in scope or ‘global’ (e.g., Ben-Ze’ev 2000, 383; Solomon 1993, 265; Nussbaum 2016, 50). In everyday parlance, as Ben-Ze’ev rightly points out (2000, 383), hatred is taken to be an “extreme” rather than a “global negation” of the target. But a closer look reveals that hatred entails an extreme devaluation and eventual negation of the target precisely because it is negatively evaluating not just aspects of it but globally. The extreme negation of hatred is founded on or emerges from a global negation. But it is not at all clear what global negation actually entails, or what the scope of ‘global’ here is. If we understand hatred as an attitude targeting and negatively evaluating persons globally or a personality or character ‘as a whole’, hatred seems to belong to a class of reactive attitudes, including shame, contempt or admiration and esteem, which have recently been discussed in terms of ‘globalist attitudes’ or ‘character-oriented reactive attitudes’ (Mason 2003; Bell 2011; Helm 2017, 189–204).

However, there are important differences between such globalist reactive attitudes and hatred. First, it is highly debatable whether hatred should be conceived of as a reactive attitude at all, even given more liberal notions of reactive attitudes than Strawson’s (1962) (see Helm 2017). Hatred is not only no appropriate response to harm, threat, or the breaching of social or moral norms, it is not even supposed to serve such a function as attributing responsibility or marking blame- or praiseworthiness as reactive attitudes do.Footnote 12 Furthermore, unlike typical reactive attitudes such as contempt or resentment hatred need not be in any relevant sense reciprocated. There need not be any explicit acknowledgment by the target (testified for example by verbal or behavioural expression) nor even any felt reaction at all on her part (e.g., feelings of guilt in reaction to being blamed). To be sure, hatred always concerns and expresses some relation of enmity. But it need not necessarily be taken up by the target, and thus is not ‘reactive’ in that sense either. Indeed, as we shall see, often, it could not even be reciprocated in principle, since the targets are imaginary collectives (cf. Szanto 2017). Finally, and most importantly for our purposes, it is not simply the (whole) character of a person or the personality that is at issue in hatred.

To see why, consider Solomon’s characterization of the global negative evaluation in hatred, involving “dislike of the person over-all, mixed with respect for his various skills, abilities, manners, personality; possibly respect for everything particular about him; but not him” (1993, 265). This characterization does not strike me as generally true. For one, even if hatred is directed at such fuzzy targets as personality traits, it is certainly not evaluating any concrete skills or abilities of persons. And for another, hatred is not simply an extreme intensification of ‘dislike’, as implicitly suggested by Solomon. Hatred is not only no extreme form of dislike; if there are any progressive steps in hatred, it has, first, to do with the above-mentioned narrative deployment of “hate-stories” (see Sternberg and Sternberg 2008, chap. 4) and, secondly, as we shall see, with the degrees of social generalization of the target.

What is right, however, is Solomon’s insistence on the difference between particular features or actions of a person, even if we take their total aggregate, on the one hand, and the totality of that person, on the other. But what does ‘the totality of a person’ mean? Is it the very personhood of the person, and is hence hatred an extreme form of depersonalization? In the light of the discussions of specific depersonalization mechanisms involved especially in hate crimes, this could be part of an answer. But it is not sufficient. To be sure, there is a certain depersonalization or dehumanization involved in hatred (see Berkowitz 2005), but the relevant sense of depersonalizationFootnote 13 is not some sort of ripping persons from their personhood or a denial of their personhood-constituting properties (whatever these may be) (see Haslam 2006). Rather, as we will see, it has to do with the overgeneralization and stereotypic social categorization of the targets. So, if depersonalization is involved in hatred, it is a depersonalization qua deindividualization.Footnote 14

Note also that in hatred the relation between person-directedness and personal blameworthiness, or, more generally, personal accountability and moral responsibility, is inversely correlated than normally. Normally, when reactive emotions, such as resentment or anger, testify to or disclose some personal failures, misdeeds, etc., they are attributed precisely to the person from whom the harm originates. At the same time these emotions indicate—or as Strawson (1962) would have it, constitute—the accountability of those persons, and eventually seek some form of retributive justice for the victim (at least an apology on the part of the perpetrator). In hatred, too, the attribution of evil is ready at hand and exclusively person(s)-centered. But accountability and eventual retribution for harm on the basis of particular actions, omissions or failures attributed to that person are precisely not at stake. One doesn’t hate somebody because he has done something (wrong, harmful, detestable or even evil) or is such-and-such (disgusting, morally bad), and is accountable for being such-and-such or doing this or that. Hatred might very well originate from or be triggered by such specific misdeeds or characteristics (recall Roberts’ description). But eventually one comes to hate somebody because it is that (evil) person who does or simply is such-and-such.

But how does this square with the above-mentioned, distinctively fungible character of the targets of hatred? After all, it now seems that the target is precisely that (determinate) person who does or is such-and-such. However, I argue that this only seems so, and does not contradict the fungible character of hatred but, on the contrary, helps explain it. For targeting a person as a whole in this specific global sense can only mean that the person is overgeneralized and hence can be substituted by any other who does or simply is such-and-such, where ‘such-and-such’ is essentially indeterminate or underdetermined.

(ii) To corroborate this claim, we must turn to the second aspect of the indeterminacy of the affective focus of hatred. I claimed that the target of hatred is indeterminate not only because its formal object picks out the target all-too globally but also because the target-properties that are thus ‘fuzzily’ picked out are themselves indeterminate. And that they are indeterminate means, to repeat, that they are not fixed in their socio-ontological status, but are rather indefinitely oscillating between individuals, groups, generalized social types, or proxies for social groupings. Various philosophers have pointed out this socio-ontological indeterminacy under different guises. Ahmed, for example, characterizes this mechanism as a peculiar alignment of the particular with the general: “Hate may respond to the particular, but it tends to do so by aligning the particular with the general; ‘I hate you because you are this or that’, where the ‘this’ or ‘that’ evokes a group that the individual comes to stand for or stand in for.” (Ahmed 2014, 49) In a similar vein, Kolnai distinguishes “more personally and more abstractly, topically directed hatred”. He argues that notwithstanding a “certain personal touch” that accompanies every instance of hatred, there is always also a collective or proxy aspect of the target. For Kolnai, every form of hatred, including what he calls “collective hatred” and “political hatred”, is directed at “the elimination of a unique, historically given human being” (1935, 141). However, even if hatred is directed at a concrete individual, inherent to the intentionality of hatred is also always

“a decision ‘about’ a third, something beloved, indeed implicitly about the ‘world’, ‘humanity’, the ‘people’ meant, such that the enemy is never hated, irrationally, as ‘that person there,’ but rather at the same time also as a ‘representative of a direction’ (Vertreter einer Richtung), i.e. as a force in the sense of a decision (combatted by us), or in terms of an emotional directedness (Gefühlsrichtung), a ‘verdict about the world’ (‘Weltentscheidung’).” (Kolnai 1935, 141)

But Sartre has probably the most radical construal of this mechanism. In a key passage in his analysis of affectively charged intentional attitudes towards others (such as desire, guilt, hate and sadism) Sartre characterizes hatred in Being and Nothingness as the ultimately irrational and futile existential-ontological project of subjects to re-institute their freedom and constitute themselves as ‘being-for-itself’ in contradistinction to the freedom of Others. They do so by “treating the Other as an instrument”, and “pursu[ing] the death of the Other” (Sartre 1944a, 410).Footnote 15 According to Sartre, in its “wish to destroy the transcendence” of the Other (her freedom), the intentionality of hatred ‘travels through’, as it were, particular qualities of another to the whole person (the Other). Eventually, the Other whom I hate assumes the role of an ontological placeholder for all others. This is how Sartre acutely puts it:

“What I hate in the Other is not this appearance, this fault, this particular action. (…) one hates right through the revealed psychic but not the psychic itself; this is why also it is indifferent whether we hate the Other’s transcendence through what we empirically call his vices or his virtues. What I hate is the whole psychic totality in so far as it refers me to the Other’s transcendence (…) a consequence of these observations is that hate is the hate of all Others in one Other. What I want to attain symbolically by pursuing the death of a particular Other is the general principle of the existence of others. The Other whom I hate actually represents all Others.” (Sartre 1944a, 410)

Even if we don’t agree with Sartre’s general existential-ontological twist, I contend that he has a key point in case. Thus, even if we don’t agree—as I for my part don’t—that it is ultimately “the general principle of existence of others” that is the target of particular instances of hatred, and even if we deny that it is the target of any instance of hatred, I believe that he is right that the intentionality of hatred essentially overgeneralizes its target. But if the target of my hate is not you, him or her, or any particular other or group, but, rather, the capitalized Other’s proxy, then my affective attitude clearly overshoots its mark, as it were, and hence becomes out of focus.Footnote 16

But what exactly happens in such overgeneralization? Is some sort of abstraction or subsuming of species (e.g., this odious person) under some general type (e.g., the hateworthy Other) at work? The answer is far from straightforward. Yet, we will not grasp the specific affective intentionality of hatred unless we have a firm handle on the nature of this mechanism.

We have a first hint in Aquinas’ important conceptual distinction between “occurring generally” and “generality as such” with regard to hatred. As Aquinas rightly points out, the generalization involved in hatred is not “generality as such”. The reason is that affects or sentiments cannot react to the general as such, “since what is general is precisely abstracted from what we sense: matter in its particularity”. However, just as our perceptual capacities (not our perceptions though) “relate to things generally”, hatred typically concerns general occurrences (ST I-II, q.29, a.6):

“(…) the object of sight is colour in general (though not colour as general) (…) and in the same way sheep don’t like wolves in general, since it is not only certain wolves that threaten them but the whole class of wolves. Contrast anger which is always directed against some particular harmful action. The senses perceive something which is open to abstractive generalization, but they don’t perceive its generality. So we cannot dislike what everything has in common, but we can dislike something common to many (…).” (Aquinas 1991, ST I-II, q.29, a.6)

Whether or not it is true that animals can dislike or even just fear really classes of those entities that threaten them, distinctive of human hatred is precisely targeting individuals in such generality. The anti-Semite or misogynist doesn’t hate a particular Jew or woman nor Judaism or femininity as general categories, but all Jews or women, i.e., the whole class of those who exhibit those general properties (falsely or correctly) attributed to them. As already mentioned, such generalizing hatred might originate in specific encounters with or in the face of specific evil deeds of particular individuals, say, that torturer X, who tortured one’s brother in the war. But, it seems, even in such extreme cases, the affective intentionality will travel through, as it were, that specific person to be directed at those torturers, those enemies or those war criminals.

In the most systematic phenomenological account of how hatred progresses via step-by-step generalizations from interpersonal to group-focused hatred, Hadreas (2007) employs the useful notion of “pseudo-eidetic generalization” to characterize the peculiar logic of the overgeneralization at stake. I cannot do justice to his exemplarily fine-grained analysis; let me just point to Hadreas’ core claim. Congenial to Aquinas’ distinction between abstractive generality and the generality of common occurrences, Hadreas argues that the generalization “is not the usual subsumption under increasing degrees of abstraction, that is of species under genus and lower genera under higher genera” (2007, 64). The peculiarity of group-focused hatred is that “the classes or types that are targets of malevolent hatred, oddly enough, neither apply to particular individuals as individuals, nor do they allow inclusion under broader generalities.” (ibid., 85) Hadreas then shows how hatred attributes negative properties to pseudo-essentialized and pre-determined classes of individuals and thus excludes the negatively branded targets from membership in any other class, to wit for pseudo-essential or pseudo-logical reasons. Here is how Hadreas describes this perversion of the logic of extension and intension characteristic of hatred:

“By exploiting not only the logical underpinning of essential generalization, as well as taking advantage of the exclusive ‘or ’, [group hatred] removes the class of hated individuals from degrees of generalization. It thrusts the exclusive logical ‘or’ onto the linkage between increasingly general terms. (…) With intense hatred, the hated figure is either a Palestinian or a human being. They cannot be both. And since we must have connotation before we can specify denotation, a removal from a ladder of extension follows in its wake of the severance from the ladder of intension.“ (Hadreas 2007, 85–86; see also 82)

What emerges here is the picture of hatred as a process of overgeneralization that ultimately places the negatively branded individuals into ‘imaginary’ groups (cf. Ahmed 2014, 49), which at the same time become the object of outgroup-focused hatred. To complete this picture, we now need to look at the logic of group-directed or, as I shall call it, ‘collectivizing hatred’ and the ingroup/outgroup dialectics at play.

5 Collectivizing hatred and the negative dialectics of hatred

The sociality of hatred is peculiar just like its overgeneralizing tendency. Indeed, the two are correlated. Thus, unlike ordinary social emotions, such as admiration or envy, paradigmatic instances of hatred don’t concern interpersonal relations, but rather have a collectivizing tendency, or so I argue. By ‘collectivizing’ I mean here, initially, the above-described overgeneralizing tendency to blur the socio-ontological status of the target, transposing the hated properties from individuals to proxies or groups, and vice versa. But there is more to the peculiar collectivizing tendency of hatred, and this is what I shall describe as the ‘negative dialectics’ of collectivizing hatred.

To be sure, the fact that hatred can and indeed typically is directed not at individuals but at groups has been pointed out by almost all authors working on the topic, philosophers and psychologists alike. Probably the first to notice this was Aristotle. For Aristotle, the distinguishing feature between hatred and anger lies precisely in the possibility of the latter being concerned with social groupings: “Anger is always concerned with individuals—Callias or Socrates—whereas hatred is directed also against classes: we all hate any thief and any informer.” (Rhetoric, 1382a1-1382a16) I don’t think that Aristotle is right about anger here. After all, why could one not be angry at the financial system or with politicians, but this is not at issue here. In any case, Aristotle is certainly right in claiming that hatred is typically directed at social groupings and generalizes them as social categories or “classes”.Footnote 17

Drawing in part on Aristotle, in his classic study of prejudice, the social psychologist Allport maintained that hatred actually more often (and more “easily”) targets groups than individuals. According to Allport, this is due to certain de-individualization processes and particularly to the fact that individuals more readily sympathize with individuals than with abstract groups (especially if they are members of the hated outgroup). Individuals have a “body-image” that, however differently construed for example in ‘cross-racial’ settings (Hugenberg et al. 2007), still very much resembles our own and makes sympathizing for example with the pain of others easier. Groups, however, lack such a body-image. Moreover, according to Allport, there is no need to gauge group-based stereotypes against reality. And we have even more motivation to hold fast to stereotypes if we allow for exceptions for some outgroup members (1954, 363–364).Footnote 18 It is also often mentioned that at the birth of hatred stands a certain ingroup/outgroup demarcation, or that this distinction is an “important root condition” of hatred (Sternberg and Sternberg 2008, 27).

As I shall now argue, however, ingroup/outgroup demarcation is not just a “root condition”—or, conversely, a typical result—but rather a constitutive feature of paradigmatic instances of hatred. Hatred certainly facilitates and, if maintained, further fuels ingroup/outgroup demarcations (cf. Alford 2005; Halperin et al. 2012). But I want to make a stronger claim here, namely that hatred, through its collectivizing tendency, essentially co-constitutes the respective ingroup and outgroup. Collectivizing hatred thus co-constitutes not just an antagonistic relationship but also the very boundaries between the relata, and hence co-individuates those. To put it in a formula: We hate (them), therefore we are (distinct from them). More than just affectively reinforcing ingroup-attachment and affective ingroup cohesion, or a sense of belonging, hatred has indeed the social-psychological and social-ontological power to constitute a community of fellow-feelers—the community of haters. This resonates with Sartre’s view that the reduction of the Other’s freedom to a Being-in-Itself in hatred is involved in the very constitution of oneself as a For-itself (1944a, 410ff.). But Kolnai may have been the first to observe that in every, “however trivial,” instance of hatred, including interpersonal ones, the “world breaks into the association of those evil there and these good here—to whom ‘oneself belongs’”. (Kolnai 1935, 133).

To avoid misunderstandings, I’m not claiming that antagonistic affective attitudes, or hatred in particular, alone are involved in the constitution of ingroup/outgroup demarcations or in intergroup conflict. And that is why I refer to these processes as co-constitutive. After all, there are certainly other than affective mechanisms involved here (symbolic, discursive, socio-economic power-struggles, etc.). I’m not prepared—as Sartre is—to ascribe any distinctive existential-ontological powers to hatred on the individual or interpersonal level either. What I do wish to maintain, however, is that individuals and groups necessarily require for their very self-constitution as haters or hate-communities an affective tie that correlates the antagonistic parties. Hatred requires an affective attachment to one another, however much this ‘attachment’ is an antagonistic one. We need the other, whom we create according to our own stereotypical model, for our very self-constitution as (a community of) haters.

And this is what I want to call the ‘negative dialectics’ of hatred. The dialectics is ‘negative’ not just because it entails an outgroup demarcation but also because the (prima facie positive) attachment to one’s own ingroup and the reinforced ingroup homogeneity is itself just a function of an antagonistic affective attitude (viz. hatred). Moreover, the target is often an imaginary other, constructed on the basis of the overgeneralizing, stereotyping tendency of hatred. Ahmed appositely glosses this process as follows:

“Hate is involved in the very negotiation of boundaries between selves and others, and between communities, where ‘others’ are brought into the sphere of my or our existence as a threat. (…) some demarcations come into existence through hate (…). If hate is felt as belonging to me but caused by an other, then the others (however imaginary) are required for the very continuation of the life of the ‘I’ or the ‘we’. (…) what is at stake in the intensity of hate as a negative attachment to others is how hate creates the ‘I’ and the ‘we’ as utterable simultaneously in a moment of alignment. At one level, we can see that an ‘I’ that declares itself as hating an other (…) comes into existence by also declaring its love for that which is threatened by this imagined other (the nation, the community and so on).” (Ahmed 2014, 51)

There are two correlated aspects in this negative dialectics that should be emphasized again: first, the dialectics between the haters’ positive affective binding or attachment to one’s own group (the community of haters), or the alignment of the ‘I’ and ‘we’, on the one hand, and the negative outgroup demarcation, on the other; secondly, and just as importantly, there is an affective attachment, albeit a ‘negative’ one, to the target. The result of this dialectical process is what Kolnai describes as the essentially “dynamic relation, [or] the existential unity (Daseinsverbundenheit) between the emoter (the hater) and the target” (1935, 111).

To illustrate how the blurriness of the affective focus elaborated in the previous section enters into a dangerous liaison with collectivizing hatred and plays out in the negative dialectic, consider again the case of xenophobic hatred targeting refugees. I mentioned earlier that the focal background object of hatred against refugees is some alleged harm issuing from the targets, typically some (self-)deceptively invoked—and often historically outright false—state of affairs or property, such as the so-called ‘Judeo-Christian’ European heritage, the allegedly unambiguous liberal tradition of Western Enlightenment,Footnote 19 or some similarly invoked sociocultural homogeneity of the supposedly endangered community of haters. I have claimed that this affective focus is uninformative as to how specific targets (individual refugees or particular groups of refugees) are related to the formal object (hateworthiness). One reason for this is that the focus itself is overgeneralized. To be sure, political, discursive and narrative framing often aims at rendering the focus more clear-cut so as to fuel hatred by trading on more specific negative affective reactions (such as fear, self-righteous anger, etc.). This may happen for instance by means of specific representations in the media (see Chavez 2001). Often particularistic or protectionist aspects also come into play, elicited by the rather concrete fear of losing jobs due to foreign workforce competitors. But none of this helps render the affective focus of hatred more informative. For how should the fear of losing my job, even if appropriate, give me any (affective or rational) reason to hate ‘the refugees’? To be sure, we may have here a socio-psychological explanation.Footnote 20 The fear of losing my job may thus explain the sources of my hatred. However, nothing in the affective focus of fear informs how the hateworthiness is related to any specific targets. And this is so even if aided by specific discursive, narrative or political strategies by which one may come to (morally or legally) delineate or ‘sort’ migrants into distinct categories, for example into those refugees who rightly deserve asylum-status and those who are self-inflicted sufferers or ‘undeserving trespassers’ (cf. Holmes and Castañeda 2016). For, in either case, what charges the focus with affective power is not something that affects me alone or others for whom I care. Rather, it is an in turn overgeneralized ‘us’ (say, we English, Europeans, Christians, Liberals, hard workers, etc.). And this ‘we’ is brought into position in the first place by affectively delineating it from those hateworthy others.

6 Hatred as a commitment to hate and habitus

At this point in the argument, we seem to face a serious problem. Hatred is arguably one of the most extreme affective attitudes. But given that its affective focus is blurred one may wonder where it derives its extreme affective weight from, a weight that can motivate extreme actions to the point of genocide. After all, as we have seen, emotions derive their affective weight from the subject’s concern with the focal background object of an emotion. Ironically, however, given the extreme effects of hatred, the focal targets of one’s hatred are of no real concern. So, if not from specific hateworthy properties of the target and their affective significance for the subject, where then does hatred receive its affective impact?

I want to argue finally that hatred gets its affective power for free, as it were—namely from the commitment to the attitude. And this commitment is importantly related to the just described negative dialectics. The idea, in a nutshell, is this: We are committed to (maintaining) hatred because it establishes or reinforces our identity as distinct from others, i.e., our social identity. I shall finally argue that this is the reason why we trust in hatred, or why it becomes robustly habitualized.

Consider again the concept of a ‘focal commitment’. In ordinary emotions, the focal commitment of emotions entails, on pain of irrationality or grave inconsistencies and all things being equal, that my emotions track changes in the objects they evaluate. Moreover, emotions directed at the same object of import (i.e., those with a common focus) ought to fit into an overall evaluative pattern (cf. Helm 2001, 67–71). My emotions ought to change if the relevant evaluative properties of the intentional object change or if my assessment of those as being evaluatively thus-and-so changes. If you are hoping that the person who insults your friend later apologizes, it would be unwarranted (irrational) to be dissatisfied to hear that the person eventually in fact apologized, or not be relieved upon learning that the alleged insult was just a joke and your friend also took it as such. Since the focus of your emotions singles out what matters to you, it commits you to such changes.

Now, given that the focus of hatred is essentially indeterminate, it might seem that hatred doesn’t commit the hater to anything (specific). The hater, it seems, is not committed to certain changes in her affective attitude, given relevant changes in the behaviour, reactions, values, or any evaluatively relevant features of the target. And this, I argue, is indeed the case. Moreover, I argue that this is also the reason why hatred is essentially characterized by a structural “inertia”, as Sartre puts its (1936/37, 62), and is prone to become a habitualized attitude that one often maintains even after the decease or elimination of its targets.

But, again, where does hatred, conceived as an intentional affective attitude, get its affective weight, if not from the concern rendered salient by the focus? I want to argue that it is distinctive of hatred that haters are committed not to anything specific that the focus singles out as mattering to the emoter—as this is essentially blurry—but simply to the attitude as such, and, in particular, to sharingFootnote 21 this attitude with one’s fellows. In other words, haters are not committed to the focal concern of the attitude; rather, their concern is only with entertaining the attitude, and with entertaining it together with others.

What is decisive for hatred, then, is not so much (if at all) the specifics of the evaluative target or the intentional objects of the attitude but rather the holding or having of the attitude itself. It doesn’t matter so much whom one hates or why exactly, but rather that one hates. Accordingly, what makes hatred an affective attitude—rather than merely a cognitive or evaluative one—is that the subject not simply holds the attitude, and under certain circumstances is disposed to feel something or to act upon it, but that she cares deeply about holding the attitude itself.Footnote 22 Importantly, this care or concern structure is not some higher-order evaluative property of the attitude, but is built into its very intentional structure. The attitude is affective—a “passion”, and not just a “conception of the world” (Sartre 1944b, 11)—in the sense that bearing the attitude itself has import for the subject. It is of affective importance for the hater to hate.

The personal-affective investment in hatred has also been highlighted by a number of phenomenologists and psychologists. As Kolnai keenly observes, hatred is characterized by a certain “personal commitment to” or “investment in” the target (persönlicher Einsatz) (1935, 110; cf. also 101, 115). He rightly views this as not just an add-on feature but an essential and distinctive intentional characteristic of hatred: “The central and unifying act of hatred seems to me to be a self-imposed—or quasi-imposed and accepted—commitment to hostility. Hatred connotes a tinge of free will more than do fear (fright, dread) or disgust.” (Kolnai 1998, 592) It is in this sense that Kolnai conceives hatred as a “centered” and “deep” sentiment and its robust biographical nature as a “historical aspect of human life” (1935, 102). And it is also in this sense of a deep and robust personal commitment that we should understand his otherwise somewhat obscure claim that hatred endows the hater with a sort of existential meaning, or is an “essential and co-determinant element for one’s very organization of life” (wesentliches, mitentscheidendes Element der Lebensgestaltung selbst; ibid., 102). Thus, hatred affects not just its targets in various very real ways (they may be avoided, excluded, killed, etc.), but also the subjects who bear it. We need not go as far as claiming that there is such a thing as a ‘hater personality’, similar to the infamous ‘authoritarian personality’ (Adorno et al. 1950; see Allport 1954, chap. 25). Yet, hatred deeply characterizes and shapes the haters themselves.

We find similar views among psychologists, who often stress the dialectical and social aspect of such commitment. Gaylin characterizes “true haters” as “obsessed with their enemies, attached to them in a paranoid partnership. It is this attachment that defines true hatred.” (Gaylin 2003, 4–5; see also Alford 2005) Similarly, some point out the “addictive” feature of hatred (Sternberg and Sternberg 2008, 64), and Sartre not only views anti-Semitic hatred as a “passion” but the anti-Semite, for Sartre, also exhibits a “profound sexual attraction” to his targets (Sartre 1944b, 33). Or, as one of the foremost intergroup theorists formulates it somewhat more soberly: “People who hate (…) are emotionally invested in the target of hate. You do not hate a person or group unless you feel strongly dependent or connected to that person or group.” (Halperin 2016, 37; cf. also Yanay 2002) Note that this does not, pace Solomon, entail any mutual dependence or “mutual responsibility” for the hate-relationship; I fail to see why “you can’t really hate someone who is indifferent to you (rather resent him).” (Solomon 1993, 265).

Importantly, creating and maintaining the dependence on the target is not just a personal but typically a shared commitment.Footnote 23 This is even evidenced by so-called ‘lone-wolf terrorists’ acting out of hatred: Think of some recent alt-right terrorists, such as Anders Behring Breivik, who clearly conceived of himself as a lone warrior and martyr, but apparently still saw the need to devise a 1600-page ideological patchwork pamphlet, “2083: A European Declaration of Independence”, thus embedding his paranoid crusade in an allegedly shared (‘European’) doctrine (see Ranstorp 2013); or take the Austrian Unabomber of the 1990s, Franz Fuchs, who persistently claimed that he acted in the name of the—in fact fictitious—“Bavarian Liberation Army”. Surely, certain psychopathological dispositions, paranoia, etc. will factor in here. But the shared commitment—imaginary or not—seems to play an equally central role (cf. Post 2010).

Accordingly, hatred may not only be a decisive element of individuals’ biographic fabric, it essentially co-determines social identities. And in its social identity conveying or, better, re-creating feature lie further powers that lend hatred its affective weight. This happens by way of the negative dialectics sketched above. Thus, I am invested in hating not only because I’m (negatively) attached to my target but also because it re-attaches me (positively) to my ingroup and reinvigorates my own social identity and the identity of my group. What gives my aversive concern with the target its force is the affective reward that I gain by aligning myself with my fellow-feelers (my ‘community of haters’), or by my re-invigorated sense of belonging. Eventually, it is in our shared commitment to antagonistically demarcate ourselves from ‘them’, in and though hate, that the haters’ very ‘organization of life’ is endowed with existential meaning and assumes clearer affective contours. In hating together, our social identity becomes accentuated and, indeed, co-determined.

To be sure, I don’t mean to suggest that there is a necessary connection between individual social identity and hatred, nor that all groups require hatred to reinforce ingroup homogeneity or group identity. As social psychologists have long pointed out, there is no necessary correlation between ingroup attachment or favoritism, on the one hand, and outgroup derogation or hostility, on the other; indeed, ingroup love is compatible with (mildly) positive attitudes as well as hatred towards outgroups (Allport 1954; Brewer 1999). The features occasionally mentioned in the literature that do in fact reinforce outgroup hatred and at the same time are correlated with ingroup attachment include: perceived threat or outright intergroup conflict (Duckitt and Mphuthing 1998), and especially in intractable conflicts (Halperin 2016), or feelings of moral superiority (Brewer 1999). Somewhat paradoxically, shared goals or values with the outgroup might actually increase outgroup hostility, as they might not only invigorate competition but also threaten clear intergroup differentiation, positive ingroup distinction, and hence the basis for ingroup identification (Brewer 1999, 2000). But no consensus has yet emerged regarding the role of contact, direct interaction or cooperation. Whereas some point out that the prospect (and fear) of such closer integration with the outgroup might sometimes reinforce hatred (Brewer 1999, 2000), others suggest that it is precisely the lack thereof that amplifies hatred as there will be even fewer opportunities to positively reappraise the target (cf. Halperin 2016, 46–47). One study that corroborates my own shared commitment model of collectivizing hatred suggests that socio-communicative sharing of negative or traumatizing experiences among ingroup members clearly fuels hatred (ibid.). More work is then needed to determine precise criteria regarding when and for which groups the negative dialectics in hatred is really the key social identity conveying mechanism.

But what seems clear is that the negative dialectics doesn’t operate in a vacuum. It is socioculturally or politically mediated and facilitated. In collectivizing hatred, there is always a public “cultivation of hatred” at work (Gay 1993). As we have seen, socially ‘scripted’, traditionally patterned negative narratives about the target facilitate our “commitment to ostracism” (Sternberg and Sternberg 2008, 66). Such hate-stories are often inherited, sometimes passed on from generation to generation, become sedimented as collective memories of (perceived) victimhood or traumata (Bar-Tal et al. 2009) and are often fueled by political discourse. Moreover, hate-communities share certain “norms that facilitate discrimination” (Sternberg and Sternberg 2008, 66) such as honor codes. Indeed, hatred doesn’t just rely on honor codes; hatred itself “has a shared code of honor that is lacking in other hostile emotions” (Solomon 1993, 265). Cases in point are family feuds and vendetta cultures (cf. Hardin 1995, 115–123). But most cases of ethnic cleansing and genocide also originate in the invocation and political exploitation of such narratives and the fuzzy, yet robust soil of traditional sociocultural norms of semi-institutionalized hate-communities (cf. Brudholm and Lang 2018). The lack or gradual loss of one’s own personal concern for the target can thus be readily reactivated by drawing from the pool of shared codes of ostracism and aversive narratives. This is nicely encapsulated in the epigraph to this paper from Burke.

In hatred, then, it seems not only that Frijda’s “The Law of Change” and the correlative “Law of Habituation” are out of action; they are, in fact, functioning in reverse. According to the former, “emotions are elicited not so much by the presence of favorable or unfavorable conditions but by actual or expected changes in favorable or unfavorable conditions”; according to the latter “continued pleasures wear off; continued hardships lose their poignancy” (Frijda 2007, 10). Though these laws are for most emotions overwhelmingly plausible and empirically well supported, in hatred, it is rather the ongoing attachment to a devaluated object or the commitment to the aversive attitude towards it—however much its specific properties might change—that keeps the affective power of the sentiment alive, and particularly so if shared with others. It might not be true that the ‘pleasure’ of hating is reinforced by habituation rather than wearing off with time, as some implicitly suggest (cf. Sartre 1944b; Alford 2005). Yet, if my argument is sound, at least the affective poignancy of the concern for the target is facilitated, indeed co-constituted, by habituation.

Thus, the inertness and robustness in hatred is not so much owing to the fact that ‘you can trust your enemies’, as the familiar saying goes. Rather it is the attitude of hatred itself that you and we can trust. For one, its affectivity comes cheap, if not for free. We don’t need to be concerned with tracking any specific negative properties in the target nor with alleviating or correcting those, since hatred is not targeted at moral repair. After all, the ultimate telos of hatred, murder or genocide cannot sensibly be conceived in terms of such morally retributive measures. Surely, destruction may be the political aim that haters pursue. But, even so, hatred may curiously persist in the face of having eliminated its target. Precisely because haters aren’t really concerned with any particular target, the attachment to our fellow haters and ingroup cohesion, and hence commitment to hatred, may persist even in the wake of, say, a ‘successfully’ executed ethnic cleansing.Footnote 24 Moreover, we can trust hatred, since it may surface even if there have never been any actual bearers of the (alleged) hateworthy properties. It’s enough to be committed to the attitude, however imaginary its objects. As Brudholm and Schepelern Johansen put this point: “to build up intense hatred, the object has to be perceived not simply as an abstract evil, but as an evil and a fearful threat to us.“(Brudholm and Johansen Schepelern 2018, 94) The intensity of hatred is precisely owing to perceiving the target as evil to us. Notice, however, that a shared evaluation of those properties as evil doesn’t make those properties less ‘abstract’ in the sense of being more focused, nor does the affective intensification of hatred presuppose that. If my analysis is on the right track, a shared commitment to construe those properties as evil alone is enough to endow hatred with affective intensity.

Whether life-long, intergenerational or less robust, hatred is an affective attitude that reaches deep into the narrative-biographical fabric of one’s personal and social identity. My identity is partly defined not just by my preferences, desires and positive attachments, but just as much by my aversive attachments and antagonistic alignments, or by whom and with whom I hate. Hatred, then, is an attitude very much in the commonsense meaning of a ‘way of life’, or indeed “a mode of being in the world” (Alford 2005, 252). As Gaylin aptly remarks, “the true purpose of an enemy will be to serve the modus vivendi, the lifestyle, of the hater” (2003, 176). And as a shared way of life, hatred is ultimately best characterized as a habitus, i.e., roughly, a socially and historically inherited and transposable, robust system of dispositions with certain internalized cognitive and motivational contents (cf. Bourdieu 1980, 52–63) and, we shall add, a certain ‘affective style’. Hatred is a habitus that disposes one to devaluate others in ways that conform to one’s and one’s peers’ previous experiences and already held, and often biased, epistemic or moral convictions. But there is not only a certain “habitus of hatred”, as Kolnai (1935, 141) and Husserl (Hass als habituelle Gefühlsrichtung; Husserl 1920/1924, 8) already observed. It’s not only that certain forms of hatred exhibit a certain habitus. Hatred is ultimately nothing but a habitus, a veritable machine of hazardous social distinctions and demarcations. And what makes this ‘affective machine’ even more dangerous is that, in terms of its affective mechanism, it’s a near-perfect perpetuum mobile.

7 Conclusion

I have suggested that the affective intentionality of hatred is distinctive in a number of ways. However, the distinctiveness of hatred is in no way owing to some especially salient or intensive affective phenomenology. This seems paradoxical, given that hatred is an extreme affective attitude. But not only is there nothing special about what it is like to feel hatred; typically, haters don’t feel anything particular, since their affective focus is blurred. Hatred indeed exhibits a personal-existential affective investment in the attitude that is atypically strong compared to other emotions. It draws one globally into the aversive relation, and this mirrors the global evaluation of the target of hatred. At the same time—however clearly carved-out the target and however deep the personal entanglement with it— the (formal) object and focus of hatred are indistinct. This, I argued, correlates with an indeterminacy regarding the attribution of hateworthy properties: typically, and especially in contexts of intergroup antagonism, they are at once attributed to individuals, proxies and social groupings or types. But, short of a clear affective focus, haters simply commit themselves to the aversive attitude, and it is this commitment from which the attitude derives its indeed extreme affective weight. Moreover, they turn to their fellows’ commitment to hate. In hating overgeneralized, unspecified others, and thus in default of concrete targets that affectively really matter, we commit ourselves to the attitude together. As I have further argued, we do so partly because this lends our respective sentiments additional affective power, to wit, a sense of togetherness with our fellow haters. This, in turn, leads to a sort of affective double bind, making hatred spin in a void, as it were. I have spelled this out in terms of a negative dialectics, according to which we need (to maintain) our antagonistic attitude towards an outgroup to create and affectively reinforce ingroup alignment. And thus hatred becomes robustly sedimented as a shared habitus. In being habitually committed to maintaining our hatred, even if we may not be ‘addicted’ to hate, we can readily trust it—for better or worse.