The concept of violence in the work of Hannah Arendt

Abstract

Arendt claimed that violence is not part of the political because it is instrumental. Her position has generated a vast corpus of scholarship, most of which falls into the context of the realist-liberal divide. Taking these discussions as a starting point, this essay engages with violence in Arendt’s work from a different perspective. Its interest lies not in Arendt’s theory of violence in the world, but in the function that violence performed in her work, namely, in the constitutive role of violence in her thought. It argues that the concept of violence allowed Arendt to make important distinctions serving to catalyze the categories that constitute her political philosophy and, in particular, the categories of public and private. More specifically, it claims that the concept of violence in Arendt’s work is the a priori background against which both the public and private realms should be defined.

The topic of violence in Arendt’s thinking has often been explored in the context of the realist-liberal divide. Arendt maintained that violence is not part of the essence of the political. For her, violence is instrumental. As such, it stands “outside the walls of the city, as it were, in order to stand guard over them” (Klein 1971, 98).¹ This position has generated a vast corpus of scholarship, most of which falls into one of two groups. The first is the work mainly of theorists associated with the realist tradition (who draw on thinkers as different as Thucydides, Machiavelli, Hobbes, Weber, Sorel, Schmitt, and Fanon, among others), for whom Arendt’s insistence on the fundamental distinction between violence and politics reveals a failure to identify violence as a mode of being in the world (Frazer and Hutchings, 2008). In that context, Arendt can be considered part of the liberal camp, which understands politics in terms of reasonable or rational agreement—a position that political realists critique as a desire “to evade, displace or escape from politics” (North 2010, 381). Yet even a superficial reading of Arendt’s philosophy makes clear that she had no wish to escape from politics; on the contrary, her entire project was to retrieve a conception of the political that she held to be forgotten or rejected by liberalism.² Hence the second group, for whom Arendt does not stand on either side of the realist-liberal debate. For these scholars, the task has been to grasp Arendt’s original concept of the political, as distinct from both violence and liberal consent (Breen 2007, Dallmayr 2001, Emden 2008, Finlay 2009).

Taking these discussions as a starting point, I would like to engage with violence in Arendt from a different perspective. My interest lies in the function that violence performed in her work, or to put it differently, in the constitutive role of violence in her thought. I start from the observation that Arendt raised the question of violence in an empirical context, namely the brutality and turbulence that had characterized the century to that point. However, my aim here will be to understand how the discussion of violence influenced the inner process of her philosophy, rather than her specific remarks on twentieth-century politics. I argue that it is Arendt’s conception of violence which allowed her to create the categories that constitute her political philosophy and, in particular, the categories of public and private. In other words, my discussion of violence in Arendt’s work will be deconstructive in a “thin” sense, in that it will follow the thread of the concept of violence to understand the composition of Arendt’s political philosophy.³ In this context, I will argue that in Arendt’s thought violence plays a role similar to that which Derrida gives the concept of khôra in Plato’s Timaeus.

In “Khôra,” Derrida comments on Plato’s conceptualization of this somewhat enigmatic term as something that is neither sensible nor intelligible but constitutes a “third genus,” about which “one cannot even say […] that it is neither this nor that or that it is both this and that” (Derrida 1995, 89).⁴ That is, khôra belongs neither to the sensible nor the intelligible, nor yet to both; but one cannot understand the sensible and the intelligible without khôra. Khôra is a void that gives everything its place, an abyss or mise en abyme that creates the possibility of places and genres, a chasm thanks to which “everything would, at the same time, come to take place and be reflected” (Derrida 1995, 104). Put differently, the logical structure of khôra is a gap and matrix, and as such, it plays a fundamental role in the political reality: It is, in effect, what makes possible the differentiation between categories of people, the distribution of functions, and, more broadly, the distinction between the political realm and everything else. However, the structure of khôra is also the structure of philosophy itself, which—like the political reality—consists of distinctions between categories. It is from this third space or genus that Socrates conceptualizes the non-place of the sophists and the place of philosophy-politics. Plato’s khôra, as Derrida understands it, is thus the structure that generates both the political reality and the philosophical discourse, as well as the philosophical discourse on the political reality and on the philosophical discourse.

I would like to retain here the structure of a “third genus” which “does not designate any of the known or recognized or, if you like, received types of existent…” (Derrida 1995, 96) in order to show that violence, in Arendt’s work, functions in a way similar to what Derrida says of khôra in Plato’s work. By this I mean that violence, for Arendt, belongs neither to the political nor the private sphere, but, like khôra, constitutes a third genus located in neither of the two, and without which the distinction between the public and private domains cannot be understood. It is the discourse on violence, like the discourse on khôra, that gives life to both domains, not in the sense that one of these domains is violent and the other not, but because the discourse on violence as non-public and non-private generates the need to define and redefine these terms.

In the first section of this paper, I discuss Arendt’s understanding of violence as instrumental and her distinction between political power and violence. I show that violence is not a necessary part of the political sphere, although it can be used by power. In the second section I establish that, being instrumental, violence can be and is often used in the private domain, but is no more an essential part of this domain than it is an essential part of the public sphere. In the third part of the paper, I argue that violence, or pure instrumentality, is the concept against which the distinction between public and private in Arendt’s philosophy is conceived. As such, it constitutes a third genus that gives life, negatively so to speak, to the two necessary domains of human existence.

1 The instrumental nature of violence

As noted above, Arendt raised the question of violence in the context of the events of the 1960s (the civil rights movement, Vietnam, and the student revolt), as well as “against the background of the twentieth century” in general (Arendt 1970, 3). By the “background of these experiences,” as she repeats later (Arendt 1970, 35), she refers to events that affected the political, by which she means power. For Arendt, “power is inherent in all politics, and all government rests on power” (Klein 1971, 98). Here we must recall Arendt’s compartmentalization of human life into labor, work, and action in The Human Condition. Action is one of the basic modes of human existence (Arendt 1998, 7); indeed, it is the manifestation of human life: “A life without speech and without action… is literally dead to the world; it has ceased to be a human life because it is no longer lived among men. With word and deed, we insert ourselves into the human world” (Arendt 1998, 176).⁵ Human life is the prime condition of politics, and so the chapter in The Human Condition titled “Action” deals with the public realm. Nonetheless, politics is the manifestation not of action, but of power; it is power, not action, that “is the essence of all political entities” (Klein 1971, 99). This is because action is individual, whereas power resides in the group: “Power corresponds to the human ability not just to act but to act in concert. Power is never the property of an individual; it belongs to a group and remains in existence only so long as the group keeps together” (Arendt 1970, 44). Politics involves not merely acting, but “acting together” (Bernstein 2011, 9).

Habermas, who was among the first to grasp the importance of Arendt’s concept of power, understands it to be the “ability to agree upon the common course of action in unconstrained communication” (Habermas 1977, 3). To me, however, Arendt’s definition of power as an ability to act in concert does not reflect communication and agreement, but, more primordially, group formation (Ayyash 2013; Breen 2007; Emden 2008)—that is, the “capacity to come together” (Frazer 2014, 156). As she writes, power acquires its meaning from “the initial getting together rather than from any action that then may follow” (Arendt 1970, 52). To put it differently, power has a “potential” character (Arendt 1998, 200), in that it precedes all agreement and communication, which are already part of the course of action itself.⁶ What makes power powerful, as it were, is thus not the content of specific actions, or even the common will and agreement that they express, but the willingness to act in common. Therefore, when, in On Violence, Arendt quotes C. Wright Mills as saying that “all politics is a struggle for power; the ultimate kind of power is violence” (Mills 1956, 171; Arendt 1970, 35), she has no problem with the first part of the sentence. Politics is a struggle for power, not in the sense that power will be “possessed” by some and “exercised” on others (as per Voltaire or Dahl), but in the sense that group cohesion is the result of a never-ending process.

However, Arendt takes issue with the second part of Mills’ sentence. Violence has nothing to do with power because in no way is violence intrinsically related to the process of group formation. For Arendt, violence is a means to an end, while power “is an end in itself”; indeed, power “is actually the very condition enabling a group of people to think and act in terms of the means-end category” (Arendt 1970, 51). Put differently, power and violence belong to different categories: Power is a central concept in Arendt’s phenomenology of human life, while violence is essentially an instrument (Arendt 1970, 51). To explain the difference between these two categories, Arendt makes the unexpected distinction between “justification” and “legitimacy.” The former relates to the realization of goals or ends, and looks toward the future; the latter relates to the act of association which grounds any political community, and so derives its validity from events that happened in the past. Power, being an end to itself, is never “justified” (it aims at nothing that would justify it), but it acquires legitimacy from the coming together of the group (Arendt 1970, 52). Indeed, since power consists in the ability to act in concert, it is by definition legitimate but never justified. Violence, on the other hand, can only be “justified,” because it draws validation only from its use as an instrument to achieve future aims.

At this point, one could argue that Arendt’s definitions are somewhat arbitrary (though she claims they are not; see Arendt 1970, 46). Certainly, they “do not correspond to any standard uses of these terms [power and violence] by political theorists or philosophers” (Bernstein 2011, 7). Moreover, as Benhabib emphasized, Arendt’s “art of making distinctions often obscured rather than illuminated the phenomena at hand” (Benhabib 1996, 123; see also Canovan 1992, 172). There is no doubt that Arendt’s insistence on the incommensurability of power and violence is often unclear and raises difficult questions, especially when she finally says that they are opposites (Arendt 1970, 56). I will return to this opposition later on. For the moment, I would like to put these questions on hold and try to understand the implications of the instrumental nature of violence.

For Arendt, “the implements of violence, like all other tools, are designated and used for the purpose of multiplying natural strength until, in the last stage of their development, they can substitute for it” (Arendt 1970, 46). That is, the purpose of tools is to multiply the strength—a natural property—of whatever or whoever is wielding them. Violence, being the “most elemental experience of human strength” (Arendt 1998, 140), is the use of strength as a tool (i.e., to achieve an end). Note Arendt’s insistence that the tools of violence are like all other tools. However, the very definition of tools—those things which have the purpose of “multiplying natural strength…until they can substitute for it”—means also that all tools consist of violence. In other words, while violent means seem at first to be types of tools, ultimately it is all tools which appear to constitute the category of violent means. Everything made with the use of tools is made violently: “We kill a tree in order to obtain wood; we destroy the wood qua wood to make a table” (Klein 1971, 101; Arendt 1998, 139).

In other words, violence and instrumentality are synonymous. For this reason, in Arendt’s work “instrumentality is always formulated negatively” (Dietz 2002, 240). Its modern manifestations, which extend from totalitarian domination to technology, are neither identical nor equally threatening; yet all represent a certain danger. If totalitarian domination aims to “fabricate something that does not exist, namely, a kind of human species resembling other animal species whose only ‘freedom’ would consist in ‘preserving the species’” (Arendt 1976, 438; emphasis mine), the question of technology is “whether machines still serve the world and things, or if, on the contrary, they and the automatic motion of their processes have begun to rule and even destroy world and things” (Arendt 1998, 151).

One could obviously retort here that eating my food with a fork and knife, or writing this essay on my computer, both involve using instruments, but neither is “violent.” To say that the act of using cutlery is violent would be preposterous, and, worse, would invalidate the very notion of violence. If everything is violent, nothing is. To understand Arendt’s excessive identification of violence with instrumentality, one must realize that she is responding not only to Marx, as she emphasizes when rejecting the idea that human beings create themselves through labor (Arendt 1970, 13; 1998, 79), but also and more importantly to Heidegger, whose essay “The Question Concerning Technology” was published a few years before The Human Condition.⁷ Put differently, her aim is not to trivialize violence but, on the contrary, to underline the danger of violence in endorsing and, at the same time, criticizing Heidegger’s views on instrumentality.

In “The Question Concerning Technology,” Heidegger questions the essence of the “instrumental” (Heidegger 1977, 6). He defines the term as “enframing” [Ge-stell], by which he means a process of calling-forth or challenging, together with a process of gathering or ordering, through which the real reveals itself (Heidegger 1977, 19–20).⁸ “Technology, so understood, is in no sense an instrument of man’s making or in his control. It’s rather that phenomenon, ruled from out of Being itself, that is centrally determining all of Western history” (Heidegger 1977, xxix). However, the “ordering” part of the instrumental contains a “supreme danger,” that of reification, of stopping the challenging claim in which the process of revealing moves forward. As such, enframing consists of concealing no less than revealing (Heidegger 1977, 27). Heidegger concludes that the essence of technology is “ambiguous”: On the one hand it involves a paradoxical relationship between revealing and concealing in which concealing threatens revealing; on the other, it includes a “more primally granted revealing,” namely, art, which is the only “saving power” that can rescue Being from concealing (Heidegger 1977, 33–34).

It has seldom been noticed that the chapter on “Work” in The Human Condition echoes Heidegger’s text in its terms and structure: It focuses on instrumentality and reification, and ends with the work of art. It reiterates Heidegger’s warning against reification: “Fabrication, the work of homo faber, consists in reification… This element of violation and violence is present in all fabrication” (Arendt 1998, 139). Here, Arendt concurs with Heidegger’s views on instrumentality as highly dangerous. However, she strongly opposes him on one major point.⁹ For Arendt, that which is instrumental can never be redeemed, because it lacks the slightest ability to reveal. Even art, which she understands as “transfiguration,” is conceived by Arendt in terms of reification (Arendt 1998, 168; 187). Therefore, Arendt keeps Heidegger’s warning against “the essence of the instrumental” without accepting the possible salvation that could come from it.

There is a process of revealing in Arendt’s philosophy, but it takes place radically beyond instrumentality—namely, in the next chapter of The Human Condition, “Action,” whose first section deals with the “disclosure” of the agent through words and deeds (Arendt 1998, 175). Revealing, in Arendt, has nothing to do with instrumentality. The agent’s “revealing” becomes political in power, namely, in the revealing-together of many actors. Power engenders a “space of appearance,” which “comes into being wherever men are together in the manner of speech and action” (Arendt 1998, 199). Power and the space of appearance are two dimensions of, or two perspectives on, a single phenomenon; neither preexists the other. The acting-in-concert of a group of people materializes a spatial “in-between,” a visual “web of relationships” (Arendt 1998, 181–182) in which people appear to each other, namely, act in concert. Explaining that the Greek equivalent of “power” is dynamis, Arendt resists Heidegger when she writes: “Men can enact and save their greatness at the same time and, as it were, by one and the same gesture, and… the performance as such will be enough to generate dynamis and not need the transforming reification of homo faber to keep it in reality” (Arendt 1998, 205). It is not instrumentality that is needed to reveal the real. The real appears not through technology (and therefore violence) but through power—namely, through people living together (Arendt 1998, 201).

For Arendt, therefore, violence is non-political not because it is never part of the political space of appearance, but because, as instrumentality, it is not a necessary part of that space. While “nothing […] is more common than the combination of violence and power” (Arendt 1970, 47), this situation is contingent. In other words, Arendt never says that violence should be avoided at all costs, only that violence and politics are two different things. The circumstances which cause violence to be combined with power can be positive or negative, and the effects of this combination can be positive or negative as well, in the sense that violence can strengthen or weaken power (violence can destroy power too, though it can never create it). On the positive side, violence is sometimes the only way to “dramatize grievances and bring them to public attention,” and as such it can be a “weapon of reform” (Arendt 1970, 79. Note that a “weapon” is clearly instrumental). In this respect, Arendt recalls Machiavelli’s “famous statements on the necessity of violence for the founding of new political bodies and for the reforming of corrupt ones” (Arendt 1977a, 139). It should be stressed here that the use of violence for the reform or even the founding of “political bodies” does not contradict Arendt’s premise that violence is non-political, because her case is that the violence of foundations is pre-political: Only when the violence ends can real politics begin (Arendt 1990, 142; Finlay 2009, 34–35; McGowan 1997, 275). Likewise, she underlines that violence sometimes enhances group cohesion. Quoting Fanon, she writes: “It is perfectly true that in military as well as revolutionary action ‘individualism is the first [value] to disappear’; in its stead we find a kind of group coherence which is more intensely felt and proves to be a much stronger, though less lasting, bond than all the varieties of friendship, civil or private” (Arendt 1970, 67).

This being said, for Arendt, violence in the public sphere is potentially dangerous and destructive. Being directed toward future goals, it tends to spread beyond these original goals, threatening laws and organized structures. While all actions produce results outside the actor’s control, “violence harbors within itself an additional element of arbitrariness” (Arendt 1970, 4). Indeed, violence is an “intoxicating spell” (Arendt 1970, 67) that cannot be restrained. As such, violence tends to lead to more violence: “The practice of violence, like all action, changes the world, but the most probable change is to a more violent world” (Arendt 1970, 80). Even more: The spell of violence is an “immense feeling of vitality in the face of death, [an] intensification of being alive,” an “ecstasy,” which “has no consequences,” by which Arendt means that it has no political consequences, because it does not lead people to act together (Klein 1971, 102). To the contrary, it tends to divide people from each other, and thereby to diminish or dilute their power.

2 Violence and the private sphere

In using violence, politics becomes something that it is not—a kind of fabrication. When politics becomes something that it is not, power is corrupted and often destroyed. Violence can be “justified” only when it is employed in the context of “work,” or even “labor” when tools are used for the sake of survival.¹⁰ It is necessary only in order to achieve ends through the process of making (Arendt 1998, 139–159; 1970, 52). However, the political sphere should have nothing to do with the process of making, because revelation and appearance constitute their own end. As Arendt said, “In the sphere of fabrication, violence is indeed justified and justifiable by the end, the end-product… In the sphere of human affairs violence as an absolute loses its creativity together with its instrumental character because nothing is achieved in the sphere of action that could be likened to an end-product” (Klein 1971, 101).

Violence as instrumentality should therefore be kept outside the public sphere. But what is “outside” this sphere? To what realm does violence belong? To the surprise and aggravation of all readers, Arendt’s answer seems straightforward: to the private realm, as defined in ancient Athens. As she explains at length in The Human Condition, the polis was divided into public and private spheres, with the latter being defined by a system of means and ends. Work and labor involved means and ends, and so were part of that private realm. The master (male, free and wealthy) used his slaves, wife, children, and beasts for his convenience, and was allowed to exercise violence on all of them. In modernity, with the regrettable “rise of the social” and the concomitant “decline of the political” (Walsh 2014, 135)—that is, with the blurring of the distinction between the public and private domains and the admission of “household and housekeeping activities,” i.e., economic issues, into the public sphere—violence became prominent in politics.¹¹ These modern developments, however, are hardly new: “Plato and, to a lesser degree, Aristotle, who thought craftsmen not even worthy of full-fledged citizenship, were the first to propose handling political matters and ruling political bodies in the mode of fabrication” (Arendt 1998, 230).

If Arendt indeed meant to justify the use of violence in the private sphere and, ultimately, in the social realm as well, her position is highly problematic. As Habermas famously argued, she would seem to call for “a state which is relieved of the administrative processing of social problems; a politics which is cleaned of socio-political issues; an institutionalization of public liberty which is independent of the organization of public wealth; a radical democracy which inhibits its liberating efficacy just at the boundaries where political oppression ceases and social repression begins” (Habermas 1977, 15). She would stand guilty of “endorsing a nostalgic vision of the ‘political’ that is utterly impractical under modern conditions” (Breen 2007, 354). And as Breen rightly adds, her conception of power would not offer a real alternative “to the time-honored emphasis on dominion since her alternative presumes an indirect affirmation of that emphasis, albeit in the realm of the ‘social’” (Breen 2007, 256).

I would like, however, to suggest another interpretation of Arendt’s position. As we have seen, Arendt contends that in the realm of fabrication, the use of violence is justified by its ends. Therefore, when people are used in a system of means and ends, violence is justified. However, she does not say that the use of people in such a system is legitimate. Further, she says that it was the Greek way to separate the realm of labor and fabrication from that of power, and she regards it as unfortunate that politics has become identified with fabrication. However, she does not say that the only way to separate politics from violence is to use people as tools in the private sphere. And indeed, in her chapter on “Work” in The Human Condition, she writes:

Although we find earlier (for instance, in Locke’s insistence that no man can be permitted to possess another man’s body or use his bodily strength) an awareness of the fateful consequences which an unhampered and unguided thinking in terms of means and ends must invariably entail in the political realm, it is only in Kant that the philosophy of the earlier stage of the modern age frees itself entirely of the common sense platitudes which we always find where homo faber rules the standards of society. The reason is, of course, that Kant… wanted first of all to relegate the means-end category to its proper place and prevent its use in the field of political action (Arendt 1998, 155–156).

With this in mind, Arendt’s claims about violence take a different tone. That is, in invoking—and radicalizing—Locke and Kant (since Kant did not condemn thinking of others as means, but only merely as means), Arendt rejects the means-end category when applied to people in general, including in the private realm. Indeed, Locke’s and Kant’s positions are not connected to the “field of political action” by contradistinction with any kind of private realm. For Locke and Kant, people—as people, not as citizens belonging to the public realm—should not be used (merely) as tools. Hence, in speaking of the separation between the political realm and the proper place of the means-end category in the context of Locke and Kant (in the context, therefore, of natural rights!), Arendt implies both that the realm of violence is a non-political realm, and that people as people should not be the recipients of violence.

To say that Arendt radicalizes Kant and utterly rejects the means-end category when applied to people does not mean that violence cannot appear in the private realm. It means that violence is not inherent to that realm. In the chapter on “Labor” of The Human Condition, Arendt emphasizes the necessities of life itself (anagkai). The needs of the body fall within the realm of biological life, and necessarily lead to labor. Arendt never calls this essential human activity “violent.” What does constitute violence, however, is using someone or something else—people or tools—in the performance of labor. To put it more bluntly, labor is aimed at meeting the necessities of life, which constitute a “natural force” (Arendt 1998, 129) that is inherently painful; violence seeks to deny this pain by using tools, or by using others, and is therefore purely “man-made”: “Since [the darkness of pain and necessity] is natural, inherent in the human condition—only the act of violence, when one group of men tries to rid itself of the shackles binding all of us to pain and necessity, is man-made—the price for absolute freedom from necessity is, in a sense, life itself, or rather the substitution of vicarious life for real life” (Arendt 1998, 119–120). Violence is always “the violence of others” (Arendt 1998, 118), while nature is a force distinct from violence (Arendt 1970, 44–45). Violence is neither essentially political nor inherent to the necessities of life.

In short, far from legitimizing violence in the private realm, Arendt criticizes the “generalization of the fabrication experience in which usefulness and utility are established as the ultimate standards for life and the world of men” (Arendt 1998, 157; Villa 1995, 199). Violence is no more the essence of the private realm than of the political. As she writes, “what needs justification by something else cannot be the essence of anything” (Arendt 1970, 51). Rather, the essence of the private realm is those activities connected to meeting the necessities of life and, hence, to sustaining life based on natality. This private realm is found between the “four walls” of the family, which “enclose a secure place, without which no living thing can thrive” (Arendt 1977b, 186) and which should never be violated (Arendt 1998, 29).

For Arendt, the major phenomenal manifestation of the private domain is education (rather than the Greek male mastery of the household, used only to illustrate the dichotomy between public and private). Its purpose is to help children grow—the metaphor of vegetative growth, the emergence from the dark earth into the light of the outside world, permeates Arendt’s reflections on education. Children, if they are to grow, must be protected and cherished (Arendt 1977b, 192) by adults whose responsibility “takes the form of authority” (Arendt 1977b, 189). Arendt makes clear everywhere in her work that authority has nothing to do with violence (Arendt 1970, 45; 1977a, 93; b 190).¹² This is so because authority shares no features with instrumentality. Like power, but in a different way, authority is based on a relationship with the past. If power derives from the people acting in concert, authority is grounded in tradition, which is constituted by sedimentary layers of past power (Arendt 1977b, 193). In other words, the public and private spaces are different but connected. Neither should be ruled by violence.

3 Violence as third genus

Arendt did not legitimize violence in any area of human existence (note that even in her highly controversial “Reflections on Little Rock,” she does not for a moment confer legitimacy upon violence in either the social or private sphere. See Arendt 1959). However, the realm of instrumentality cannot be confined or delimited; it is not separated or detached from everything else. As Villa puts it, “the world created by homo faber… is not a different world from that inhabited by human beings qua political actors. The world created by work and the world ‘illuminated’ by action are the same world” (Villa 1995, 137). For Arendt, labor, work and action are different moments of the same thing: the human condition on earth (Arendt 1998, 7). My position here is very close to Benhabib’s when she contends that “the only tenable and productive way of distinguishing the social from the political [in Arendt’s work] is in the light of attitudinal orientations.” Arendt’s distinction between the different forms of the human condition is based not on content (whereby certain activities or concerns might essentially be deemed social and others, political) or on institutions, but on attitudes (Benhabib 1996, 139, 141). Indeed, Arendt says explicitly that the content of “acting together” will vary at different points in time: “Life changes constantly, and things are constantly there that want to be talked about. At all times people living together will have affairs that belong in the realm of the public… What these matters are at any historical moment is probably utterly different… So what becomes public at every given period seems to me utterly different” (Arendt 1979, 316).¹³ Therefore, it must be that the distinction between labor, work and action can be maintained even though their content is constantly transformed.

I argue that this distinction can be maintained because, for Arendt, instrumentality is always there, in parallel with natality and power, but it can be conflated neither with the former nor with the latter. Violence characterizes things that are part of neither the public nor the private sphere. It is therefore only by contrast with violence that we can think about those things which are either political or private. With respect to the political, the content of “words and deeds” can be anything that comes from group formation, so long as this “thing” is not reified. With respect to the private, the content of the private sphere can be anything that needs protection and darkness to grow, so long as this thing is not turned into fabrication. In other words, there are no specific activities, issues or concerns that constitute the political, but nothing that has to do with instrumentality is political. At the same time, the fact that violence is not political does not mean that in essence it belongs to the private sphere. Violence seems to be a third category or place by contradistinction with which both the private and the public spheres can be conceived.

As mentioned above, Arendt maintains that “power and violence are opposites” (Arendt 1970, 56). What this means is that “being its own end” and instrumentality are logical opposites. For this reason, the Greek public and private spheres were opposites, or were thought to be so. In modern times, these opposites became intermingled in the social, in the philosophies of Hegel and Marx. However, Arendt has no trust in the dialectic, in which opposites “develop into each other” (Arendt 1970, 56).¹⁴ She does not believe a thing can be derived from its opposite. The empirical existence of the social cannot be denied; but what can be done, against Hegel and Marx, is to re-conceptualize clear-cut domains that do not develop into each other. Yet in re-conceptualizing the public and private spheres in opposition to dialectic, Arendt had to distance them from that which, for Marx (and largely also for Hegel), allowed dialectic, namely, violence (Arendt 1977c, 22). What, then, to do with violence? It cannot be simply ignored. Dislodged from the movement of history, it could be conceived as a pre-political stage, as in Hobbes. However, Arendt unequivocally contends that violence is not natural (Arendt 1970, 82), and she does not understand politics as reflecting a state monopoly over the legitimate use of force which puts an end to generalized violence. Therefore, violence exists but is not a natural stage, and is not inherent to the private and public spheres.

Arendt emptied the concept of violence of its Marxist function of “midwife of history” (Arendt 1977c, 22) without relegating it to some pre-political hypothetical state. In that context, violence, defined as instrumentality inherent neither to the public nor the private sphere, and with no dynamic power, becomes a static category against which dynamic life (political and biological) must be defined. Its definition is only negative. As such, it allows us to draw a line between the political and the non-political, but this delineation is never restricted to specific issues or to specific segments of the population, and it has no proper spatial location. Violence does not belong anywhere or to anyone in particular, but its presence and absence delineate groups and realms.

To return to where this essay began, violence, like khôra, constitutes a third genus, without which the distinction between the private and the public domains (the only existing domains) cannot be understood. As I said at the start, it is the discourse on violence, like the discourse on khôra, that gives life to both domains, not in the sense that one is violent and the other is not—since, as we have said, violence is the essence of neither—but because the discourse on violence generates the need to define and redefine them. Violence, says Arendt, is a rupture in acting-in-concert (Arendt 1970, 83). As such, it calls for a regeneration of power. Again, this does not mean that power grows out of violence (Arendt 1970, 53), because it could not possibly grow out of its negation (recall that Arendt had no trust in the Hegelian dialectic). But instrumentality as a void of power outlines the contours of what and where power should be. By the same token, the discourse on violence calls for a redefinition of a secure private space, which, like power, is put in jeopardy by reification. Indeed, Arendt’s “recovery of the public world” (Hill 1979) would have no meaning if it did not come together with the recovery of a privacy destroyed in modernity by that same violence which destroyed the public sphere. As Arendt wrote in her description of “loneliness” at the end of The Origins of Totalitarianism, “totalitarian government, like all tyrannies, certainly could not exist without destroying the public realm of life, that is, without destroying, by isolating men, their political capacities. But totalitarian domination as a form of government is new in that it is not content with this isolation and destroys private life as well. It bases itself on loneliness, on the experience of not belonging to the world at all” (Arendt 1976, 475).

Take the example to which Arendt comes back in several texts, the student revolt of the late 60s. The violence of these events leads Arendt to conceptualize the two spheres that should not be ruled by violence: the “private” sphere of education, which should maintain high academic standards under all circumstances, and the “public” sphere, where the right of participation belongs to all independently of academic criteria (Arendt 1970, 18–20; 1972, 201; 1977b). It is only against the background of violence that Arendt defines these two spheres as different from one another. But the background of violence reveals that the inhabitants of the two spheres are the same people, the students, whose public protests relate to the private realm of their education. As paradoxical as khôra, which generates all discourse while being part of none, the discourse on violence is a revolving door between two realms that are in fact the selfsame and only world. In that context, the two realms can never be definitively characterized, and are always in a process of being defined.

In Arendt, both the private and the public sphere acquire their specific “place” because of and from violence. Arendt’s entire enterprise is to redefine both power and the security of privacy “against the background of the twentieth century,” namely, omnipresent violence. Put differently, her categories appear in the process of distancing themselves from violence. The public and private spheres are defined by their rejection of violence; if there were no violence they would not be defined. To put it simply: Without violence Arendt’s philosophy would not be. On the one hand, she refuses to give any place to violence; on the other hand, her entire project is meaningful only against its backdrop. She is not a realist, because she refuses to assign a political role to violence. She is not a non-realist, because the political (and the non-political) exist only in a constant dialogue with violence. Violence in Arendt is neither a Hobbesian state of nature that can only be overcome by consent and will endure only as a frightening ghost, nor a Machiavellian mythological foundation. It is and remains there as instrumentality against which the political and non-political spheres must constantly be conceived. In a striking irony, violence is the instrument thanks to which Arendt conceptualizes the categories of her theory. It is a major tool of a thinking that refuses to give any legitimacy to instrumentality, and attempts to save the human condition from reification.