The Two Faces of Power

Giorgio Agamben

In his commentary on the cruel repression of Italian anarchist Alfredo Cospito, Giorgio Agamben argued that anarchy “is first and foremost the radical disavowal not so much of the state or simply of administration, but rather of power’s claim to make the state and administration coincide in the government of men.” How does this coincidence of two poles within a single machine function, and why is it so essential to the operation of power? What does it mean to say that the state of exception under which we live today is “not only anomic but also anarchic”? 

The following text, first published in four installments on the Quodlibet website in March 2023, traces the binary machine of Western politics from Aristotle to the present day, passing through St. Thomas Aquinas, Napoleon, and Carl Schmitt, among others. In it, the philosopher highlights the relation between the anarchy of power and its everyday operation in the form of laws, constitutions, and governments. “The time has come,” he argues, “to ask whether the fractured political machine of the West has not reached a threshold in recent years beyond which it can no longer function.”

1. The Two Faces of Power

Any inquiry into politics is marred by a preliminary terminological ambiguity, which condemns those who undertake it to misunderstanding. Consider the passage in the Book III of the Politics in which Aristotle, while “studying the politeiai to determine their number and qualities,” peremptorily states: “since politeia and politeuma signify the same thing and politeuma is the supreme power of cities [to kyrion ton poleon], this supreme power must be one, few, or many” (1279a 25-26).1 Current  translations read: “since constitution and government mean the same thing and government is the sovereign power of cities…”2 Whether this translation is more or less correct, it exposes what we could call the amphiboly of what is perhaps the most fundamental concept of our political tradition, presented sometimes as “constitution” and sometimes as “government.” In a dizzying contraction, the two concepts are at the same time identified and distinguished, and it is precisely this equivocation that defines the kyrion or sovereign, according to Aristotle. That the amphiboly is not occasional is quickly confirmed by a reading of the Athenaion politeia, which we translate as The Athenian Constitution. In describing the "demagogy" of Pericles (27:1), Aristotle writes that in it demotikoteran eti synebe genesthai ten politeian, which the translators render as “the constitution became still more democratic”; immediately afterwards, we read that the many apasan ten politeian mallon agein eis hautous, “centralized in their hands the whole government” (obviously, translating “the whole constitution,” as terminological consistency would require, did not seem possible).3 The ambiguity is confirmed by dictionaries, where politeia is rendered as “constitution of the state” just as frequently as “government, administration.”

Whether it is referred to by the hendiadys “constitution/government” or “state/administration,” the fundamental concept of Western politics is a dual concept, a Janus-face that sometimes displays the austere and solemn face of the institution, at other times the more obscure and informal face of administrative practice, without it being possible to identify or dissociate either side.

In his 1932 essay Legality and Legitimacy, Carl Schmitt distinguishes four types of state. Setting aside the two intermediate figures of the jurisdictional state, in which the judge deciding a particular legal dispute enjoys the final say, and the governmental state, which Schmitt identifies with dictatorship, we are here interested in the two extreme types: the legislative state and the administrative state. In the former, the legislative state based on the rule of law, “the highest and most decisive expression of the common will” consists of norms having the character of law. “The justification for such a state system rests on the general legality of any exercise of power by the state."4 Those who exercise power here act on the basis of a law or “in the name of the law,” and the legislative and executive powers, law and its enforcement, are consequently separated. It is with this type of state that modern parliamentary democracies have identified themselves, with less and less reason.

Last on the list, unsurprisingly, as if the other state forms finally all converged toward it, is the administrative state. Here, "command and decision-making do not appear authoritarian and personal, but neither can they be reduced to the mere application of higher norms"5; rather, they take the form of concrete arrangements, made from one moment to the next on the basis of the current situation, and in view of practical objectives or needs. It can also be said that in the administrative state, “men do not rule, nor do norms count as something superior," but rather, according to a famous formula, “things administer themselves.”

As is evident today, yet as Schmitt could already deduce in those years from the emergence of totalitarian states in Europe, the legislative state tends gradually to transform itself into an administrative state. “Moreover, our state form is undergoing a transformation, and the ‘turn toward the total state’ characteristic of the moment [...] seems typical today of the turn toward the administrative state.” While political scientists appear to have forgotten this today, Schmitt unreservedly asserts, as a “universally recognized fact,” that an “economic state” cannot function as a legislative-parliamentary state and must necessarily transform itself into an administrative state, in which the law gives way to decrees and ordinances.

For those of us who have witnessed the completion of this process, it is the meaning of this transformation — if it is indeed a transformation — that deserves to be questioned. The idea of transformation implies that the two models are formally and temporally distinct. Of course, Schmitt knows perfectly well that “in historical reality, linkages and mixtures continuously appear,” and that “each political community must legislate and judge, govern and administer.” It is possible, however — and this is our hypothesis — that the mixture is even more intimate and that the legislative state and the administrative state, legislation and administration, constitution and government are essential and inseparable parts of a single system, which is the modern state as we know it. While it is therefore tactically possible to play one of the two elements against the other, it would be quite misleading to believe that we can isolate in a stable manner something that forms an integral part of the same bipolar system.

A different politics will only be possible from the realization that State and administration, constitution and government are two sides of the same reality, which must be radically questioned. There is no power that can legitimize its exercise through laws without presupposing an extra-legal order that founds it, nor can there be a pure administrative practice that claims to remain legal on the basis of decrees issued out of necessity. These are, as Schmitt himself suggests, two different ways of making obedience compulsory. As we can see today, the truth of both is in fact the state of exception. Whether one acts in the name of the law or in the name of administration, what is at issue in the final analysis is always the sovereign exercise of a monopoly of violence. And this is the kyros, the hidden ruler who, in the words of Aristotle, holds together in one system the two faces of state power.

2. Politics and Economics 

The lapidary phrase Napoleon uttered when meeting Goethe in Erfurt in October 1808 is well known: Le destin c'est la politique: “destiny is politics.” This statement, which was at the time perfectly intelligible, even if seemingly revolutionary, has lost its meaning entirely for us today. We no longer know what the term “politics” means, much less do we dream of seeing in it our destiny. “Destiny is economics” is instead the refrain that so-called "political men" have been repeating to us for decades now. And yet, not only have they not given up defining themselves as such, but “politicians” continue to refer to themselves based on the party to which they belong, and “politics” is said to refer to the coalitions they form in governments and the decisions they endlessly make. 

What then do we mean today when we pronounce, albeit without much conviction, the word “politics”? Is there in it something like a unitary meaning or, rather, is the meaning that the term conveys constitutively splintered? The terminological uncertainty in the translation of the term politeia, which we have already analyzed, is nothing new. The Latin translation of Leonardo Aretino's Politics, published in Rome in 1942 with Thomas Aquinas' commentary, renders the term with gubernatio and respublica (more rarely with civitatis status). If the passage we have quoted (1279a, 25-26) reads in its Latin translation: Cum vero gubernatio civitatis et regimen idem significant…, in the preceding passage politeia is instead rendered respublica (est autem respublica ordinatio civitatis). In the commentary of Thomas Aquinas, who obviously had another translation in front of him, politeia is sometimes translated as policia, sometimes as respublica. The proximity of the term policia to our "polizia" is not surprising: polizia is indeed, until the beginning of the 19th century, the Italian term for politeia. We can still read "polizia" in Marcello Adriani's translation of Plutarch, published in Florence in 1819: “it means the order with which a city is governed and its common needs are administered; and so it is said that there are three types of polity [tre polizie], the monarchical, the oligarchical, and the democratic.” 

For the German theorists of cameralism and police science, which took shape and spread throughout Europe in the course of the 18th century, the science of the state became a science of government (Regierungwissenschaft), whose essential object was the Polizei, defined — in opposition to Politik, which was only responsible for the fight against external enemies — as the administration of the good order of the community and the care for the well-being and life of its subjects, in all its aspects. And it is certainly no coincidence that Napoleon, who resolutely affirmed politics as destiny, was also the sovereign who gave the administration and the police the modern form with which we are familiar. The administrative state theorized by Sunstein and Vermeule, which is taking hold in advanced industrial societies, is in its own way faithful to this model, in which the state seems to resolve itself into administration and government and “politics” transforms itself entirely into “police.” It is significant that, precisely in a state conceived in this way as a “police state” that the latter phrase ends up designating the least edifying aspect of government, i.e., the bodies required to ultimately ensure by force the realization of the state's governmental vocation. And yet, the formal apparatus of the legislative state does not disappear, just as the laws that governments continue to enact in spite of everything do not disappear, nor are the offices and dignities that, according to the constitution, embody and guard the legitimacy of the system abolished. Beyond its transformations, the essential bipolar nature of the political machine is at least formally maintained.

3. The Kingdom and the Government 

 “Le roi règne, mais il ne gouverne pas” — “The king reigns, but he does not govern.” That this formula, which lies at the heart of the debate between Peterson and Schmitt on political theology and which, in its Latin formulation (rex regnat, sed non gubernat) reaches back to the seventeenth-century polemics against the Polish king Sigismund III, contains something like the paradigm of the dual structure of Western politics, is what we tried to demonstrate in a book published nearly fifteen years ago. Here again, what is underlying it is a genuinely theological problem, that of the divine government of the world, itself ultimately an expression of an ontological problem. In chapter X of book Λ of the Metaphysics, Aristotle asks whether the universe possessed the good as something separate (kechorismenon) or as an internal order (taxin). In other words, it was a question of resolving the radical opposition between transcendence and immanence, articulating them together through the idea of an order of entities in the world. The cosmological problem also has political implications, as Aristotle immediately compares the relation between the transcendent good and the world to that which binds the strategist of an army to the ordering of its soldiers and a house to the mutual connection of the creatures living in it. “Entities,” he adds, “do not want to have a bad political constitution (politeuesthai kakos) and therefore there must be a single sovereign (heis koiranon),” who manifests himself in them in the form of the order that connects them. This means that, ultimately, the unmoved mover of Book Λ and the nature of the cosmos form a single two-sided system and that power — whether divine or human — must hold the two poles together and be both transcendent norm and immanent order, both kingdom and government.

It will be the task of medieval scholasticism, and in particular of Thomas Aquinas, to translate this ontological paradigm into a theological problem of the divine government of the world. Essential to this end is the notion of order, which expresses, on the one hand, the relationship between God and creatures (ordo ad Deum) and, on the other hand, the relationship of creatures to each other (ordo ad invicem). The two orders are closely related, yet their relationship is not as perfectly symmetrical as it seems. That the problem once again has a political aspect is evident in the comparison Aquinas makes between the law and its execution. “Just as in a family,” he writes, “order is imposed through the law and precepts of the head of the family, who is for each of the ordered beings of the house the principle of the execution of the order of the house, so the nature of natural entities is for each creature the principle of the execution of what is due to it in the order of the universe.” Yet how can the law, as the command of one alone, translate into the execution by the many of that which is ordered? If order — as the examples, which are certainly not coincidental, of the strategist and the head of the family seems to indicate — depends on the command of a leader, how can its execution be inscribed in the nature of entities so different from each other?

The aporia that will increasingly mark both the order of the cosmos and the order of the city begins to become visible here. Entities stand in a certain relation to one another, but this relation is only the expression of their relation to the unique divine principle, and, conversely, entities are ordered insofar as they stand in a certain relation to God, but this relation consists only in their relation to each other. The immanent order is only the relation to the transcendent principle, but the latter has no other content than the immanent order. The two orders refer back to each other and are reciprocally founded. The perfect edifice of medieval cosmology rests on this circle and has no consistency outside it. Hence the complex and subtle dialectic between first and second causes, absolute power and ordered power, by which scholasticism will attempt, without ever completely succeeding, to come to terms with this aporia.

If we now return to the problem of political order from which we started, and which explicitly refers to this theological paradigm, it will not be surprising to find in it the same circularities and aporias. State and administration, kingdom and government, norm and decision are mutually connected and exist through each other; and yet — precisely because of this — their symmetry can be neither perfect nor unequivocally guaranteed. The king and his ministers, “politics” and “police,” the law and its execution may enter into conflict, and nothing ensures that this conflict can be resolved once and for all. The bipolar machine of Western politics is always in the process of corrupting and breaking down, perpetually at the mercy of changes and revolutions that challenge its functioning and bipolarity to the same extent that they seem to reaffirm them at every turn.

The primacy of government over kingdom and administration over the constitution that we are witnessing today is actually not without precedents in the history of the West. It reached its first and most radical formulation in the elaboration of the doctrine of rex inutilis by the canonists of the 13th century. It was on the basis of these elaborations that, at the request of the Portuguese clergy and nobility, Pope Innocent IV issued the Grandi non immerito decree in 1245, which deposed King Sancho II from the government of the kingdom (which he had proved incapable of administering), assigning the cura et administratio generalis to his brother Alfonso de Boulogne instead, while leaving Sancho with his royal dignitas. The double structure of the governmental machine contains the possibility that the bipolarity in which it is articulated may be called into question if it ceases to be functional for the system. However, since neither side of power has its foundation in itself, it is significant that even in this extreme case royal dignity is not taken away. The duality of legitimacy and legality is but one aspect of this bipolarity: the kingdom legitimizes government, yet legitimacy has no other meaning than the legality of the government's actions and measures.

4. Anarchy and Politics 

It was a German constitutional scholar of the late 19th century, Max von Seydel, who posed the question that today seems inescapable: “What is left of the kingdom, if we do away with the government?” Indeed, the time has come to ask whether the fractured political machine of the West has not reached a threshold in recent years beyond which it can no longer function. Already in the 20th century, fascism and Nazism had answered this question in their own way by establishing what has rightly been called a “dual state,” in which the legitimate state, based on law and the constitution, is flanked by a discretionary state that is only partially formalized. The unity of the political machine is therefore only apparent. The administrative state into which European parliamentary democracies have more or less consciously slipped is in this sense technically only a descendant of the Nazi-fascist model, in which discretionary organs extraneous to constitutional powers are placed side by side with those of the parliamentary state, which is gradually emptied of its functions. And it is certainly singular that a separation between kingdom [regno] and government has even manifested itself today at the highest reaches of the Roman Church, where a pontiff, finding himself unable to govern, has spontaneously deposed the cura et administratio generalis, while retaining his dignitas.

The most extreme demonstration of the fracture of the political machine, however, is the emergence of the state of exception as the normal paradigm of government, which, now in place for decades, reached its apex during those years we now refer to as the pandemic. From the perspective that interests us here, what defines the state of exception is the rupture between constitution and government, legitimacy and legality — and, at the same time, the creation of a zone in which they become indiscernible. Indeed, sovereignty here manifests itself in the form of a suspension of law and the consequent establishment of a zone of anomie, in which the government nevertheless claims to act legally. While suspending the legal order, the state of exception claims, in fact, to still be in relation to it, to be, so to speak, legally outside the law. From a technical point of view, the state of exception in fact invents a “state of the law” in which, on the one hand, the law theoretically exists but has no force, and on the other hand, measures and provisions that do not have the force of law acquire the force of law. One could say that, at the limit, what is at stake in the state of exception is a fluctuating force-of-law without the law, an illegal legitimacy matched by an illegitimate legality, in which the distinction between norm and decision loses its meaning.

What is essential is to understand the necessary relationship that unites the state of exception and the political machine. If the sovereign is the one who decides on the exception, the state of exception has always constituted the secret center of the bipolar machine. Between kingdom and government, between legitimacy and legality, and between constitution and administration, there can be no substantive articulation. Insofar as it marks their point of coincidence, the hinge that connects them can belong neither to one pole nor to the other and can be in itself neither legitimate nor legal. As such, it can only be the subject of a sovereign decision, which articulates them punctually through their suspension. 

For this very reason, however, the state of exception is necessarily temporary. A sovereign decision made once and for all is no longer such, just as a permanent articulation between the two poles of the machine would eventually undermine its functionality. A normal state of exception becomes undecidable and therefore abolishes the sovereign, who can only define himself through decision. It is certainly no coincidence that both Nazism and the contemporary administrative state have resolutely adopted the state of exception as the normal and not temporary paradigm of their government. In any case, however one defines this situation, the political machine has renounced its operation and the two poles — the kingdom and the government — are mirrored in each other without any articulation.

It is in the threshold between kingdom and government that the problem of anarchy can be properly situated. If the political machine functions by articulating the two poles of kingdom/government, what the sovereign exception reveals clearly is that the space between them is actually empty, it is a zone of anomie without which, however, the machine could not function.

Just as the norm does not contain its own application, but for this requires the decision of a judge, so the kingdom does not contain in itself the reality of government and the sovereign decision is what, by rendering them indiscernible, opens up the space of governmental practice. The state of exception is therefore not only anomic but also anarchic, in a twofold sense: the sovereign decision has no foundation, and the praxis it inaugurates moves within the indistinction between legality and illegality, norm and decision. And since the state of exception constitutes the hinge between the two poles of the political machine, this means that it functions by capturing anarchy at its center.

Thus, we may define a truly anarchic power as one that is capable of liberating the anarchy that has been captured in the machine. Such a power can exist only as the halting and destitution of the machine; that is to say, it is a power that is integrally destituent and never constituent. In Benjamin's words, its space is the “real” state of exception, as opposed to the virtual one on which the machine rests, which claims to maintain the legal order through its own suspension. In it, the kingdom and government exhibit their definitive disconnection, and it can no longer be a question of restoring their legitimate articulation, as its benevolent critics would have it, nor (according to a misunderstood conception of anarchy) of playing administration against the state. We have known for some time now, with lucid awareness and without any nostalgia, that we tread daily on this impervious and risky threshold where any articulation between kingdom and government, state and administration, norm and decision is irrevocably severed; all the while, the deadly specter of the machine continues to circle around us.

First published on the Quodlibet website in March 2023: Part I , II, III, IV.  

Translated by Ill Will.

Images: Mitch Epstein


1. Translations of Aristotle and Schmitt have been modified to suit the Italian renderings. All footnotes have been added by the translators. —IW  

2. Aristotle, Politics, Trans. C.D.C Reeve, Hackett, 77. 

3. Aristotle, The Athenian Constitution, Trans. H. Rackham, Loeb Classical Library, 79, 81. 

4. Carl Schmitt, Legality and Legitimacy, Trans. Jeffrey Seitzer, Duke, 2004, 4. 

5. Schmitt, Legality and Legitimacy, 5. Translation modified. The remaining Schmitt quotes are all taken from this page.