To switch for a moment to conceptual language: Power is indeed of the essence of all government, but violence is not. Violence is by nature instrumental; like all means, it always stands in need of guidance and justification through the end it pursues. And what needs justification by something else cannot be the essence of anything. The end of war – end taken in its twofold meaning – is peace or victory; but to the question “And what is the end of peace?” there is no answer. Peace is an absolute, even though in recorded history periods of warfare have nearly always outlasted periods of peace. Power is in the same category; it is, as they say, “an end in itself.” (This, of course, is not to deny that governments pursue policies and employ their power to achieve prescribed goals. But the power structure itself precedes and outlasts all aims, so that power, far from being the means to an end, is actually the very condition enabling a group of people to think and act in terms of the means-end category.)

And since government is essentially organized and institutionalized power, the current question “What is the end of government?” does not make much sense either. The answer will be either question-begging – to enable men to live together – or dangerously utopian – to promote happiness or to realize a classless society or some other nonpolitical ideal, which if tried out in earnest cannot but end in some kind of tyranny.

Power needs no justification, being inherent in the very existence of political communities; what it does need is legitimacy. The common treatment of these two words as synonyms is no less misleading and confusing than the current equation of obedience and support. Power springs up whenever people get together and act in concert, but it derives its legitimacy from the initial getting together rather than from any action that then may follow. Legitimacy, when challenged, bases itself on an appeal to the past, while justification relates to an end that lies in the future.
Violence can be justifiable, but it never will be legitimate. Its justification loses in plausibility the farther its intended end recedes into the future. No one questions the use of violence in self-defense, because the danger is not only clear but also present, and the end justifying the means is immediate.

Power and violence, though they are distinct phenomena, usually appear together. Wherever they are combined, power, we have found, is the primary and predominant factor. The situation, however, is entirely different when we deal with them in their pure states – as, for instance, with foreign invasion and occupation. We saw that the current equation of violence with power rests on government’s being understood as domination of man over man by means of violence. If a foreign conqueror is confronted by an impotent government and by a nation unused to the exercise of political power, it is easy for him to achieve such domination. In all other cases the difficulties are great indeed, and the occupying invader will try immediately to establish Quisling governments, that is, to find a native power base to support his dominion. The head-on clash between Russian tanks and the entirely nonviolent resistance of the Czechoslovak people is a textbook case of a confrontation between violence and power in their pure states. But while domination in such an instance is difficult to achieve, it is not impossible.

Violence, we must remember, does not depend on numbers or opinions, but on implements, and the implements of violence, as I mentioned before, like all other tools, increase and multiply human strength. Those who oppose violence with mere power will soon find that they are confronted not by men but by men’s artifacts, whose inhumanity and destructive effectiveness increase in proportion to the distance separating the opponents. Violence can always destroy power; out of the barrel of a gun grows the most effective command, resulting in the most instant and perfect obedience. What never can grow out of it is power.

In a head-on clash between violence and power, the outcome is hardly in doubt. If Gandhi’s enormously powerful and successful strategy of nonviolent resistance had met with a different enemy – Stalin’s Russia, Hitler’s Germany, even prewar Japan, instead of England – the outcome would not have been decolonization, but massacre and submission. However, England in India and France in Algeria had good reasons for their restraint.

Rule by sheer violence comes into play where power is being lost; it is precisely the shrinking power of the Russian government, internally and externally, that became manifest in its “solution” of the Czechoslovak problem, just as it was the shrinking power of European imperialism that became manifest in the alternative between decolonization and massacre.

To substitute violence for power can bring victory, but the price is very high; for it is not only paid by the vanquished, it is also paid by the victor in terms of his own power. This is especially true when the victor happens to enjoy domestically the blessings of constitutional government. Henry Steele Commager is entirely right: “If we subvert world order and destroy world peace we must inevitably subvert and destroy our own political institutions first.”

The much-feared boomerang effect of the “government of subject races” (Lord Cromer) on the home government during the imperialist era meant that rule by violence in faraway lands would end by affecting the government of England, that the last “subject race” would be the English themselves. The recent gas attack on the campus at Berkeley, where not just tear gas but also another gas, “outlawed by the Geneva Convention and used by the Army to flush out guerrillas in Vietnam,” was laid down while gas-masked Guardsmen stopped anybody and everybody “from fleeing the gassed area,” is an excellent example of this “backlash” phenomenon.

It has often been said that impotence breeds violence, and psychologically this is quite true, at least of persons possessing natural strength, moral or physical. Politically speaking, the point is that loss of power becomes a temptation to substitute violence for power – in 1968 during the Democratic convention in Chicago we could watch this process on television – and that violence itself results in impotence. Where violence is no longer backed and restrained by power, the well-known reversal in reckoning with means and ends has taken place. The means, the means of destruction, now determine the end with the consequence that the end will be the destruction of all power.

Now here is the self-defeating factor in the victory of violence over power more evident than in the use of terror to maintain domination, about whose weird successes and eventual failures we know perhaps more than any generation before us. Terror is not the same as violence; it is, rather, the form of government that comes into being when violence, having destroyed all power, does not abdicate but, on the contrary, remains in full control. It has often been noticed that the effectiveness of terror depends almost entirely on the degree of social atomization.

Every kind of organized opposition must disappear before the full force of terror can be let loose. This atomization – an outrageously pale, academic word for the horror it implies – is maintained and intensified through the ubiquity of the informer, who can be literally omnipresent because he no longer is merely a professional agent in the pay of the police but potentially every person one comes into contact with. How such a fully developed police state is established and how it works – or, rather, how nothing works where it holds sway – can now be learned in Aleksandr I. Solzhenitsyn’s The First Circle, which will probably remain one of the masterpieces of twentieth century literature and certainly contains the best documentation on Stalin’s regime in existence. The decisive difference between totalitarian domination, based on terror, and tyrannies and dictatorships, established by violence, is that the former turns not only against its enemies but against its friends and supporters as well, being afraid of all power, even the power of its friends. The climax of terror is reached when the police state begins to devour its own children, when yesterday’s executioner becomes today’s victim. And this is also the moment when power disappears entirely. There exist now a great many plausible explanations for the de-Stalinization of Russia – none, I believe, so compelling as the realization by the Stalinist functionaries themselves that a continuation of the regime would lead, not to an insurrection, against which terror is indeed the best safeguard, but to paralysis of the whole country.

To sum up: politically speaking, it is insufficient to say that power and violence are not the same. Power and violence are opposites; where the one rules absolutely, the other is absent. Violence appears where power is in jeopardy, but left to its own course it ends in power’s disappearance. This implies that it is not correct to think of the opposite of violence as nonviolence; to speak of nonviolent power is actually redundant. Violence can destroy power; it is utterly incapable of creating it.

Hegel’s and Marx’s great trust in the dialectial “power of negation,” by virtue of which opposites do not destroy but smoothly develop into each other because contradictions promote and do not paralyze development, rests on a much older philosophical prejudice: that evil is no more than a privative modus of the good, that good can come out of evil; that, in short, evil is but a temporary manifestation of a still-hidden good. Such time-honored opinions have become dangerous. They are shared by many who have never heard of Hegel or Marx, for the simple reason that they inspire hope and dispel fear – a treacherous hope used to dispel legitimate fear. By this, I do not mean to equate violence with evil; I only want to stress that violence cannot be derived from its opposite, which is power, and that in order to understand it for what it is, we shall have to examine its roots and nature.
___
Excerpt from Hannah Arendt’s On Violence (New York: Harcourt, 1969).