Stanley I. Benn, "Democracy," Encyclopedia of Philosophy, ed. Paul Edwards, 1967.


Stanley I. Benn

"Democracy" is difficult to define, not only because it is vague, like so many political terms, but more importantly, because what one person would regard as a paradigm case another would deny was a democracy at all. The word has acquired a high emotive charge in the last hundred years; it has become good tactics to apply it to one's own favored type of regime and to deny it to rivals. The most diverse systems have been claimed as democracies of one sort or another, and the word has been competitively redefined, to match changes in extension by appropriate changes in intention. However, there is still this much agreement: democracy consists in "government by the people" or "popular self-government." As such, it would still be universally distinguished from, say, a despotism that made no pretense of popular participation -- the despotism of Genghis Khan or of Louis XIV, for instance -- or from a theocracy, like the Vatican. There remains plenty of room for disagreement, however, about the conditions under which the people can properly be said to rule itself.

In the first place, what is "the people"? In ancient Greece, the demos was the poorer people; democracy meant rule of the poor over the rich. This is still the usage of those who identify the people with the proletariat and democracy with the rule of the working class. However, the word "people" is often used to differentiate the subject mass from the ruling elite, as, for instance, when Locke speaks of a tyrannical government putting itself into a state of war with the people. In this sense, "the people" necessarily means the ruled. Can the people, however, be said to rule itself in the same sense as it is said to be ruled by monarchs, oligarchs, and priests? To rule is, generally, to prescribe conduct for someone else. There is a sense, it is true, in which moralists speak of ruling oneself, when by a kind of metaphor they speak of reason governing the passions. Again, a former colony becomes self-governing when its people is no longer ruled by outsiders; but this is not inconsistent with its still being ruled by native masters.

The usual paradigm of a people governing itself is the direct democracy of ancient Athens. Admittedly, citizenship was a hereditary privilege, excluding slaves and metics, and it is very doubtful whether, without this limitation, the citizen body would have been small enough for it to have operated as it did. Aside from this, however, the Athenian people governed itself in, the sense that every individual could participate personally in policy decisions by discussion and voting, in a face-to-face situation. Athenian procedures are held to have been democratic in the sense that everyone was supposed to have an equal opportunity to state a case and influence decisions, even if, in some cases, individuals had ultimately to accept decisions that they had previously resisted. So today, in a similar sense, if a school or a department is said to be democratically run, we should expect its head to consult his staff on important issues and to concur in decisions to which he himself is opposed when the weight of opinion is against him. Self-government for a small group consists in general participation in the deliberative process, in which each person's voice carries a weight appropriate not to his status but to the merits, in the judgment of others, of what he has to say. If, despite continuing disagreement, a decision is essential, then it must be arrived at by majority vote. For it is not consistent with equal participation in decision making for any one individual to be privileged to say in advance that regardless of the distribution of opinions, his own or that of his group must prevail. That privilege excluded, decisions may be reached by lot or by vote; and if by vote, the opinion of either the lesser or the greater number may prevail. Deciding by lot was in fact used in Athens to fill certain public offices; it is a way of giving everyone an equal chance where advantages or privileges cannot be equally and simultaneously enjoyed; but to decide policy by lot would make nonsense of the procedure of public discussion, which is as integral to the democratic process as the idea of equality. The same would apply to a rule whereby whatever opinion received the fewest votes would prevail; for what point would there be in persuasion if it had no effect on the outcome or, still worse, if it actually reduced the chance of one's view being implemented? If a democratic decision is thought of, then, as the result of a fair confrontation of opinions, it must, at best, be generally agreed upon, and at worst, agreed upon by the majority.

Conditions of political democracy. Obviously, the conditions of face-to-face democracy, with direct participation, cannot be fulfilled within the political structure of modern states, both because of the size of their populations and because of the specialized knowledge needed to govern them. So although everyone may agree on what makes a small group democratic, when it comes to applying the concept to mass organizations, there is plenty of room for different interpretations of the principles to be applied and of the way to realize them under these very different conditions. Democracy now becomes representative government, that is, government by persons whom the people elect and thereby authorize to govern them.

"Election" and "representation" are themselves complex notions, however. In one sense, to be representative of a group may mean no more than to possess salient characteristics common to and distinctive of most of its members. In another, quasi-legal sense, one person may be said to represent another if, according to some code of rules, the consequences attached to an act of the representative are precisely those that would be attached to the act had it been performed by the principal himself; the representative can, in this case, commit the represented. In yet a third sense, one may represent another by looking after his interests, with or without his authorization (for example, the representation of infants in law). Now, democratic representation need not imply representation in the first sense, that of resemblance. Since an elected member of a legislature is taken to represent those who voted against as much as those who voted for him, he need not resemble those he represents, even in his opinions. Nor does he commit them as if they themselves had acted; the fact of their having legal duties does not depend on the fiction that, if their representative votes for a law, they have personally agreed to it. Their legal duties remain even if their representative voted against it. Nor must we necessarily accept moral responsibility for what is done by those who politically represent us, for in voting against them, we may have done the only thing open to us to disavow them.

Political representation is closer to the third sense of the term -- the representation of interests; a democratic representative is usually thought to have the duty to watch over either the interests of his constituents or, as a member of an assembly representing the whole people, the interests of the people at large. Nevertheless, he could still represent the interests of a group of people without their having had any part in choosing him. Some members of colonial legislatures in Africa used to be nominated by the governor to represent the interests of the unenfranchised native population. Precisely analogous, from the standpoint of the liberal democrat, is the case of a single-party system, where the ruling party invites the electors to endorse the candidate it has chosen to represent them. No matter how zealously the representative watched their interests, this would not count as democratic representation, precisely because the electors had had no part in selecting him. This view of democracy, therefore, is not compatible with tutelage; it implies the possibility not only of rejecting but also of freely proposing candidates, if none put forward by others is acceptable. Choosing and rejecting representatives is, indeed, the central act of participation by the citizens of a mass democracy, from which any effectiveness that they might have in other respects derives.

Closely related to election is the notion of the responsibility of the democratic representative. This means, in practice, that representatives must submit themselves periodically for re-election and, as a corollary, that they must be prepared to justify their actions and to attend to the experience and needs of their constituents, whose good will they must retain so long as they wish to remain in office.

Democracy and popular sovereignty. It is often said that in a democracy the people's will is sovereign. But can the people be said to have a will? Opinions are divided on most things; there may be ignorance and apathy; on many questions only sectionally interested groups may have any clear opinions at all. Small groups, like committees, may reach agreed policies to which everyone feels committed; or in time of grave national danger, whole nations may discover a collective devotion to a single objective, overriding all conflicts of interests. However, although it might be intelligible to speak of a collective will in such cases, they are too limited or too rare to provide a framework for a general theory of democratic government. Such cases apart, one may speak of action, will, or decision in relation to collectivities only if their collective acts can be identified by some more or less formal procedure or if there are rules authorizing some identifiable individual to act in the name of the whole group. Thus, "Parliament has decided . . ." presupposes rules determining who are members of Parliament, defining their roles, and giving their several actions a collective significance and validity as "legislation." Are there analogous procedures, by virtue of which the people can be said to act or to express a will? Only by voting and by applying the majority principle in elections and referenda. And of course, applied to any particular collection of individual votes, different systems of voting or different arrangements of constituency boundaries can yield quite different results, each in its own rule context expressing "the people's will." Nevertheless, some people consider a system democratic to the extent that it approximates to government by referendum, though they would agree that this could not work as a day-to-day procedure. The doctrine that a government ought not to initiate policy changes without putting them to a vote in a general election (or, in a stronger form, that having done so, it is entitled -- or obliged -- to implement them forthwith) is a practical application of the popular-sovereignty view of democracy. A possible corollary sometimes derived from this last view is that it is undemocratic to oppose or impede any government acting with the people's mandate. Moreover, since the people is sovereign, the traditionally important safeguards against the abuse of power become otiose; for, in Rousseau's words, "the sovereign, being formed wholly of the individuals who compose it, neither has nor can have any interest contrary to theirs." Popular-sovereignty theory is always, therefore, on the brink of totalitarianism, since -- as the French Jacobin party showed -- it is only a short step from proclaiming the sovereignty of the people to claiming the unlimited authority of its elected representatives, to proscribing opposition, and to denying individuals any rights other than those which the government with majority support deems fit.

There is, of course, another view, closer to the tradition of liberal individualism, which sees democracy as a way of safeguarding and reconciling individual and group interests. For James Madison, the virtue of the new constitution of the United States was that it permitted no faction, not even a majority, to deprive minorities of their natural rights, since it demanded the concurrence in action of independent authorities. The constitution was designed to balance diverse interests against one another, so that none might ever become a dominant and entrenched majority. Recent pluralistic accounts of democracy (or of what R. A. Dahl calls "polyarchy"), while more sophisticated, follow a similar approach. To be democratic, policy-making agencies must be sensitive to a wide range of pressures, so that no interest significantly affected by a decision will be left out of account. Popular participation consists not merely in voting, but also in wide consultation with interest groups and in the whole process of public criticism and governmental self-justification. Democracy, according to this view, requires the dispersal, not the concentration, of power; every voter has his quantum, making him worth the attention of those who want to govern. The people is not homogeneous, but a highly diversified complex of interest groups with crisscrossing memberships. It rarely makes sense to talk of the majority, except with reference to the result of a particular election or referendum, to describe how the votes were cast. A sectional majority, if there were one, would have no intrinsic claim to rule. To govern, a party would have to piece together an electoral majority; but every elector would have his own reasons for voting as he did, and no party could say in advance that, since it had no potential supporters among the members of some particular group, that group could, therefore, be safely neglected. Admittedly, wherever group divisions coincide over a wide range of interests (as, for instance, in many polyethnic societies), these conditions might not be fulfilled, and there might be a built-in majority and minority. In such a case, no party aiming at majority support could afford to uphold a minority interest, and democracy would tend to give way to majority tyranny. Thus, where popular-sovereignty theorists see the majority as the expression of the supreme will of the people, writers like Madison, Alexis de Tocqueville, J. S. Mill and, more recently, Walter Lippmann and the pluralists have seen it as either a myth or a potential tyrant.

The possibility of democracy. According to elitist sociologists like Vilfredo Pareto, Gaetano Mosca, and Robert Michels, there is always, behind the democratic facade, an oligarchy, even though its members take turns at playing the key governing roles. Now obviously, in every organization leaders initiate action and followers concur, but the power relations between leader and led are not on that account always the same. Precisely because democracy is a form of political organization, it must also be a pattern of leadership; nevertheless, the way leaders gain and retain their authority; the extent to which their initiatives respond to the interests of those they lead; their need to listen to and answer criticism -- these things distinguish a democracy in important ways from what we usually mean by an oligarchy.

For the Marxist, bourgeois democracy is a sham because equal political rights cannot equalize political power where economic power is unequal. This does not amount to saying that democracy is necessarily impossible, only that economic equality and a classless society are necessary conditions for it.

According to other critics, popular self-government is delusory because government calls for expertise which few voters possess. Most accept the directions of some party, to whose image they are irrationally committed, and are incapable of a rational choice of policy. However, except in the popular-sovereignty variant, democracy does not require the electors to choose policies. Their role is merely to choose governors whom they trust to deal fairly and efficiently with problems as they emerge, and to look for new governors when they are disillusioned. A party's public image need not be an irrational construct; it may accurately epitomize deep-rooted tendencies and traditional preferences and be a reliable guide to the spirit in which the party would govern.

Justification of democracy. Democracy, it is sometimes said, asks too much of ordinary men, who would never be prepared to maintain the lively and informed interest in politics that ideally it demands. This, however, presupposes a particular view of the purpose and justification of democratic government. For some writers, as J. S. Mill, men and women cannot be fully responsible, adult, moral persons unless they are "self-determining," that is, concerned about the ways in which their lives are to be controlled. This view is a development from an older natural-rights theory of democracy, according to which (in the words of Colonel Rainborough, the Leveller), "Every man that is to live under a government ought first by his own consent put himself under that government," this being a condition for preserving his natural autonomy as a rational being. Or again, for democrats in the tradition of Rousseau, men achieve moral fulfillment only as participants in the collective self-governing process, helping to give expression to the "General Will" for the "Common Good"; failure in this constitutes failure in one's moral duty as a citizen.

There is, however, a more strictly utilitarian theory, sketched by Bentham and James Mill and implicit in a good deal of the work of democratic political scientists today. According to this view, the test of the adequacy of a political system is whether it tends to provide for the interests of the governed and protect them against the abuse of power. Democracy, they maintain, is likely to do this better than other systems. Active participation has no intrinsic virtue. James Mill would have limited the franchise to men over 40, on the grounds that the interests of women and younger men would be adequately safeguarded by their husbands and fathers, and therefore universal suffrage would be an unnecessary expense. For many modern writers, politics is a second-order activity: if things are going well, there is really no reason for people who prefer to spend their time on other things to devote it to politics. Political activity, indeed, is often most vigorous, as in Germany before 1933, when passions are high and democracy is in imminent danger of collapse. Apathy may be a sign of political health, indicating that there are no irreconcilable conflicts nor serious complaints. If there is ground for disquiet, it is only that apathy may become so habitual that democracy's defenses may be found unmanned in the face of some future attack.

This is a prudential model of democracy, in which satisfaction is maximized and conflicts reconciled by pressures bringing countervailing pressures into operation. It leaves out of account, perhaps, the sense in which democracy moralizes politics. Because decisions have to be publicly justified, political debate is conducted in moral terms, reviewing the impact of decisions on all interests affected, not just on this or that pressure group. Moreover, the quantum of power one has as a citizen can be represented not simply as a lever for personal or sectional protection or advantage, but also as a public responsibility; for even when one's own interests are not affected, one is still a member of a court of appeal. The bystanders in a democracy are, in a sense, the guarantors that a political decision shall not simply register the strongest pressure but shall be a reasoned response to diverse claims, each of which has to be shown to be reasonable, in the light of whatever standards are widely accepted in the community.



Benn, S. I., and Peters, R. S., Social Principles and the Democratic State. London, 1959. Reissued as Principles of Political Thought. New York, 1964.

Dahl, R. A., Preface to Democratic Theory. Chicago, 1956. A formal analysis of types of democratic theory.

De Grazia, A., Public and Republic. New York, 1951. On theories of representation, with useful annotated bibliography.

Mayo, Henry B., An Introduction to Democratic Theory. New York, 1960.

Pennock, J. R., Liberal Democracy: Its Merits and Prospects. New York, 1950. Includes an extensive bibliography.

Sartori, G., Democrazia e definizione, 2d ed. Bologna, 1958. Translated by the author as Democratic Theory. Detroit, 1962. Notes include extensive bibliographical references.

Wollheim, R., "Democracy." Journal of the History of Ideas, Vol. 19 (1958), 225-242. Includes brief history and bibliographical references.


McKeon, R., ed., Democracy in a World of Tensions. Chicago, 1951. Edited for UNESCO.

Naess, A., Christopherson, J. A., and Kval0, K., Democracy, Ideology, and Objectivity. Oslo, 1956. A semantic analysis of the McKeon volume.


Crosa, E., La sovranitd popolare dal medioevo alia rivoluzione francese. Bocca, 1915.

Glover, T. R., Democracy in the Ancient World. Cambridge, 1927.

Gooch, G. P., English Democratic Ideas in the Seventeenth Century, H. J. Laski, ed. Cambridge, 1927.

Mill, James, Essay on Government. London, 1821; Cambridge, 1937.

Mill, John Stuart, Considerations on Representative Government. London, 1861. Recent edition by R. B. McCallum. Oxford, 1946.

Rousseau, J. J., Political Writings of Jean-Jacques Rousseau, C. E. Vaughan, ed. Oxford, 1962. In French.

Rousseau, J. J., Rousseau: Political Writings, translated and edited by F. Watkins. London, 1953.

Spitz, D., Patterns of Antidemocratic Thought. New York, 1949.

Talmon, J. L., Origins of Totalitarian Democracy. London, 1952.

Tocqueville, Alexis de, De la Democratic en Amerique, 2 vols. Brussels, 1835-1840. Translated by H. Reeve as Democracy in America, P. Bradley, ed., 2 vols. New York, 1945.

Woodhouse, A. S. P., Puritanism and Liberty. London, 1938.