Copyright © 1996 National Endowment for Democracy and the Johns Hopkins University Press. All rights reserved. This work may be used, with this header included, for noncommercial purposes within a subscribed institution. No copies of this work may be distributed electronically outside of the subscribed institution, in whole or in part, without express written permission from the JHU Press. This revolutionary publishing model depends on mutual trust between user and publisher.

Bernard Lewis, "Islam and Liberal Democracy: A Historical Overview," Journal of Democracy 7.2 (1996) 52-63

In a necessarily brief discussion of major issues, it is fatally easy to go astray by misuse or misinterpretation of some of the words that one uses. Therefore, I ought to say first what I mean by the terms "Islam" and "liberal democracy."

Democracy nowadays is a word much used and even more misused. It has many meanings and has turned up in surprising places--the Spain of General Franco, the Greece of the colonels, the Pakistan of the generals, the Eastern Europe of the commissars--usually prefaced by some qualifying adjective such as "guided," "basic," "organic," "popular," or the like, which serves to dilute, deflect, or even to reverse the meaning of the word.

Another definition of democracy is embraced by those who claim that Islam itself is the only authentic democracy. This statement is perfectly true, if one accepts the notion of democracy presupposed by those who advance this view. Since it does not coincide with the definition of democracy that I take as the basis of this discussion, I will leave it aside as irrelevant for present purposes.

The kind of democracy I am talking about is none of these. By liberal democracy, I mean primarily the general method of choosing or removing governments that developed in England and then spread among English-speaking peoples and beyond.

In 1945, the victors of the Second World War imposed parliamentary democracy on the three major Axis powers. It survives in all three, [End Page 52] precariously, perhaps, in one. In none of them has it yet confronted any crisis of truly major proportions. Among the Allies, Britain and France bequeathed their own brands of democracy--with varying success--to their former colonies during the postwar retreat from empire.

Perhaps the best rule of thumb by which one can judge the presence of the kind of democracy I mean is Samuel P. Huntington's dictum that you can call a country a democracy when it has made two consecutive, peaceful changes of government via free elections. By specifying two elections, Huntington rules out regimes that follow the procedure that one acute observer has called "one man, one vote, once." So I take democracy to mean a polity where the government can be changed by elections as opposed to one where elections are changed by the government.

Americans tend to see democracy and monarchy as antithetical terms. In Europe, however, democracy has fared better in constitutional monarchies than in republics. It is instructive to make a list of those countries in Europe where democracy has developed steadily and without interruption over a long period, and where there is every prospect that it will continue to do so in the foreseeable future. The list of such countries is short and all but one of them are monarchies. The one exception, Switzerland, is like the United States in that it is a special case due to special circumstances. In the French Republic, established by revolution more than two centuries ago, the march of democracy has been punctuated by interruptions, reverses, and digressions. In most of the other republics of Europe, and, for that matter, in the rest of the world, the record is incomparably worse.

In all this, there may be some lesson for the Middle East, where the dynastic principle is still remarkably strong. The most purely Arab and Muslim of Middle Eastern states, Saudi Arabia, derives its name and its identity from its founding and ruling dynasty. So, too, did the Ottoman Empire--the most recent and by far the most enduring of all the Islamic empires. Even such radical revolutionary leaders as Hafiz al-Assad in Syria and Saddam Hussein in Iraq endeavor to secure the succession of their sons. In a political culture where the strain of dynastic legitimacy is so strong, democracy might in some places fare better by going with it rather than against it.

What of our other term, "Islam"? It too has multiple meanings. In one sense, it denotes a religion--a system of belief, worship, doctrine, ideals, and ideas--that belongs to the family of monotheistic, scriptural religions that includes Judaism and Christianity. In another sense, it means the whole civilization that has grown up under the aegis of that religion: something like what is meant by the once-common term "Christendom."

When we in the West today talk of Christian art, we mean votive art, religious art. If we talk of Islamic art, however, we mean any art [End Page 53] produced by Muslims or even by non-Muslims within Islamic civilization. Indeed, one can still speak of Islamic astronomy and Islamic chemistry and Islamic mathematics, meaning astronomy, chemistry, and mathematics produced under the aegis of Islamic civilization. There is no corresponding "Christian" astronomy or chemistry or mathematics.

Each of these terms, Islam in the sense of a religion and Islam in the sense of a civilization, is itself subject to many variations. If we talk about Islam as a historical phenomenon, we are speaking of a community that now numbers more than a billion people, most of whom are spread along a vast arc stretching almost 10,000 miles from Morocco to Mindanao; that has a 14-century-long history; and that is the defining characteristic of the 53 sovereign states that currently belong to the Organization of the Islamic Conference (OIC). For obvious reasons, it is extremely difficult (though not impossible) to make any kind of valid generalization about a reality of such age, size, and complexity.

Even if we confine ourselves to speaking of Islam as a religion, significant distinctions must be drawn. First, there is what Muslims themselves would call the original, pristine, pure Islam of the Koran and the hadith (the traditions of the Prophet Mohammed) before it became corrupted by the backsliding of later generations. Second, there is the Islam of the doctors of the holy law, of the magnificent intellectual structure of classical Islamic jurisprudence and theology. Most recently, there is the neo-Islam of the so-called fundamentalists who introduce ideas unknown alike to the Koran, the hadith, or the classical doctrines of the faith.

Clearly this last version of Islam is incompatible with liberal democracy, as the fundamentalists themselves would be the first to say: they regard liberal democracy with contempt as a corrupt and corrupting form of government. They are willing to see it, at best, as an avenue to power, but an avenue that runs one way only.

History and Tradition

What then of the two others--historic Islam and Islam as a system of ideas, practices, and cultural traits?

A first look at the historical record is not encouraging. Predominantly Muslim regions show very few functioning democracies. Indeed, of the 53 OIC states, only Turkey can pass Huntington's test of democracy, and it is in many ways a troubled democracy. Among the others, one can find democratic movements and in some cases even promising democratic developments, but one cannot really say that they are democracies even to the extent that the Turkish Republic is a democracy at the present time.

Throughout history, the overwhelmingly most common type of regime in the Islamic world has been autocracy--which is not to be confused [End Page 54] with despotism. The dominant political tradition has long been that of command and obedience, and far from weakening it, modern times have actually witnessed its intensification. With traditional restraints on autocracy attenuated, and with new means of surveillance, repression, and wealth-extraction made available to rulers by modern technologies and methods, governments have become less dependent than ever on popular goodwill. This is particularly true of those governments that are enriched by revenues from oil. With no need for taxation, there is no pressure for representation.

Another noteworthy historical and cultural fact is the absence of the notion of citizenship. There is no word in Arabic, Persian, or Turkish for "citizen." The cognate term used in each language means only "compatriot" or "countryman." It has none of the connotations of the English word "citizen," which comes from the Latin civis and has the content of the Greek polites, meaning one who participates in the affairs of the polis. The word is absent in Arabic and the other languages because the idea--of the citizen as participant, of citizenship as participation--is not there.

At the same time, however, we can discern elements in Islamic law and tradition that could assist the development of one or another form of democracy. Islam boasts a rich political literature. From the earliest times, doctors of the holy law, philosophers, jurists, and others have reflected carefully on the nature of political power, the ways in which political power ought to be acquired and used and may be forfeited, and the duties and responsibilities as well as the rights and privileges of those who hold it.

Islamic tradition strongly disapproves of arbitrary rule. The central institution of sovereignty in the traditional Islamic world, the caliphate, is defined by the Sunni jurists to have contractual and consensual features that distinguish caliphs from despots. The exercise of political power is conceived and presented as a contract, creating bonds of mutual obligation between the ruler and the ruled. Subjects are duty-bound to obey the ruler and carry out his orders, but the ruler also has duties toward the subject, similar to those set forth in most cultures.

The contract can be dissolved if the ruler fails to fulfill or ceases to be capable of fulfilling his obligations. Although rare, there have been instances when such dissolutions took place. There is, therefore, also an element of consent in the traditional Islamic view of government.

Many hadith prescribe obedience as an obligation of a subject, but some indicate exceptions. One, for example, says, "Do not obey a creature against his creator"--in other words, do not obey a human command to violate divine law. Another says, similarly, "There is no duty of obedience in sin." That is to say, if the sovereign commands something that is sinful, the duty of obedience lapses. It is worth noting that Prophetic utterances like these point not merely to a right of [End Page 55] disobedience (such as would be familiar from Western political thought), but to a divinely ordained duty of disobedience.

When we descend from the level of principle to the realm of what has actually happened, the story is of course checkered. Still, the central point remains: there are elements in Islamic culture that could favor the development of democratic institutions.

One of the sayings traditionally ascribed to the Prophet is the remark, "Difference of opinion within my community is a sign of God's mercy." In other words, diversity is something to be welcomed, not something to be suppressed. This attitude is typified by the acceptance by Sunni Muslims, even today, of four different schools of Islamic jurisprudence. Muslims believe the holy law to be divinely inspired and guided, yet there are four significantly different schools of thought regarding this law. The idea that it is possible to be orthodox even while differing creates a principle of the acceptance of diversity and of mutual tolerance of differences of opinion that surely cannot be bad for parliamentary government.

The final point worth mentioning in this inventory is Islam's emphasis on the twin qualities of dignity and humility. Subjects--even the humblest subjects--have personal dignity in the traditional Islamic view, and rulers must avoid arrogance. By Ottoman custom, when the sultan received the chief dignitaries of the state on holy days, he stood up to receive them as a sign of his respect for the law. When a new sultan was enthroned, he was greeted with cries of "Sultan, be not proud! God is greater than you!"

The Influence of the West

For the first thousand years of its history, Islamic civilization's relationship to Christendom was one of dominance. The loss of Spain and Portugal on the remote western periphery had little impact in the heartlands of Islam, and was more than compensated by the advance toward the heart of continental Europe. As late as 1683, an Ottoman army was encamped before the very gates of Vienna. Earlier in the seventeenth century, North African corsairs were raiding as far north as the British Isles. By the early nineteenth century, however, Islamic power was clearly in retreat as European power grew. Finding themselves the targets of conquest and colonization, Muslims naturally began to wonder what had gone wrong. Islam had always been generally "successful" in worldly terms. Unlike the founder of Christianity, who was crucified and whose followers saw their religion made the official faith of the Roman Empire only after centuries as a persecuted minority, Mohammed founded a state during his lifetime, and as ruler he collected taxes, dispensed justice, promulgated laws, commanded armies, and made war and peace. [End Page 56]

Educated Muslims, chagrined by the newfound potency of their European rivals, asked: What are they doing right and what are we doing wrong, or not doing at all? Representative, constitutional government was high on the list. The nineteenth century saw the rise of elected assemblies in a number of Western countries, and democracy in our current sense was beginning to take hold. Many Muslims suspected that here--in this most exotic and alien of Western practices--lay the secret to the West's wealth and power, and hoped that the adoption of constitutions and the creation of elected legislatures in the Islamic world would redress the civilizational balance.

Getting used to the idea was not easy; the first Muslim visitors to the West disliked much of what they saw. The earliest detailed description of England by a Muslim traveller is a fascinating account by Mirza Abu Talib Khan, a Turko-Persian resident of Lucknow who was in England between 1798 and 1803. He watched the House of Commons in action, and his comments are enlightening. The government and opposition MPs sitting on their benches facing each other across the chamber reminded him of trees full of parrots squawking at each other, a common sight back home in India. When he learned that the purpose of the noisy assemblage was to make laws, he was shocked. The English, he explained to his readers, had not accepted a divine law and so were reduced to the expedient of making their own laws, in accordance with the experience of their judges and the requirements of their time.

Later accounts were more positive. The first Egyptian student mission went to France in 1826. Their chaplain, a sheikh from al-Azhar, learned a great deal (probably more than his student wards) and wrote a remarkable book about Paris. In it he discusses the National Assembly and the freedom of the press, among other things, and makes the very astute observation that the French, when they speak of freedom, mean roughly what Muslims are getting at when they talk of justice. With this insight, he cuts right to the heart of a key difference between European political culture and its Islamic counterpart.

To Muslims, the use of "freedom" as a political term was an imported novelty, dating only from the time of the French Revolution and General Napoleon Bonaparte's arrival in Egypt in 1798. Before that, it had only legal and social connotations, and meant simply the condition of not being a slave. For Muslim thinkers, as the sheikh from al-Azhar implied, justice is the ideal, the touchstone by which one distinguishes good governments from bad.

By the latter part of the nineteenth century, Islamic rulers were coming to think of a constitution as something that no well-dressed nation could afford to be without. Just as gentlemen were abandoning traditional garb in favor of Western-style frock coats, neckties, and trousers, so the state would sport a constitution and an elected legislature as essential accoutrements. [End Page 57]

Yet the idea of freedom--understood as the ability to participate in the formation, the conduct, and even the lawful removal and replacement of government--remained alien. This notion, which belongs to the inner logic of constitutionalism and parliamentarism, is obviously a troublesome one for dynastic autocracies, which can hardly accept it and remain what they are. The real question, then, was whether consti-tutions, elections, and parliaments--the institutional trappings of democracy--would be only that, or would actually become means that the governed could use to gain some say in their government.

The first serious elections in the Islamic world were held to choose the parliament called for by the Ottoman Constitution of 1876. This parliament was no doubt meant to be a tame body that would supply the ceremonial ratification of the sultan's authority. But the Chamber of Deputies soon developed a mind of its own. On 13 February 1878, the deputies went so far as to demand that three ministers, against whom specific charges had been brought, should appear in the Chamber to defend themselves. The next day, in response, the sultan dissolved the parliament and sent its members home. It did not meet again until the year after the Young Turk Revolution of 1908. That phase, too, was of brief duration, and a military coup ended the stormy interval of parliamentary rule.

Since then, parliamentarism has not fared especially well in the Islamic world. All too often, elections are less a way of choosing a government than a ritual designed to ratify and symbolize a choice that has already been made by other means--something like a presidential inauguration in the United States or a coronation in Britain. This is not always so--there are intervals and cases where elections mean something, and they become more common in the record as one goes from the nineteenth into the twentieth century, in spite of (or perhaps because of) a number of dramatic moves in the opposite direction.

A Rough Classification of Regimes

Another complication surrounding the term "freedom" is a legacy of imperialism. When outsiders ruled much, though not all, of the Islamic world, freedom came also, or even primarily, to mean communal or national independence, with no reference to the individual's status within the body politic.

Most of the countries in the Islamic world today are free from external domination, but not free internally: they have sovereignty, but lack democracy. This shared lack, however, does not preclude the existence of very great differences among them. Predominantly Muslim societies (Turkey, as we saw earlier, being the great exception) are ruled by a wide variety of authoritarian, autocratic, despotic, tyrannical, and totalitarian regimes. A rough classification would include five categories. [End Page 58]

1) Traditional autocracies. These are the countries, like Saudi Arabia and the Gulf sheikhdoms, where established dynastic regimes rest on the traditional props of usage, custom, and history. These regimes are firmly authoritarian in character, but the same traditions that sustain them also bind them: their legitimacy relies heavily on acceptance, and too much open repression would shatter it. Their props are not quite what they used to be, however, having been partly undermined by new ideas and forces. The rulers use modern devices to help maintain themselves, but the same devices--especially electronic communications media--are now also available to those who would overthrow the existing order.

The Iranian Revolution, which overthrew the Shah in 1979, was the first electronic revolution in history. It will not be the last. Khomeini could do nothing while he was in Iran, and very little from nearby Iraq. But when he went to Paris and began recording cassette tapes and calling Iran via the direct-dial telephone system that the Shah had installed, he reached a vast audience, with results we know all too well. Satellite television, the fax machine, and electronic mail can all carry the message of subversion in ways difficult to prevent or control. The methods used by the Islamic revolutionaries against the Shah are now being used--in a more sophisticated form--by those who seek to overthrow the Islamic Republic. Other dissident groups--ethnic, religious, ideological--are using the same methods against the regimes that rule in their countries.

2) Modernizing autocracies. These are regimes--one thinks of Jordan, Egypt, and Morocco in particular--that have their roots in traditional autocracy but are taking significant steps toward modernization and democratization. None really fits the description of liberal democracy as given above, but none is anything like a total autocracy, either. All three are moving toward greater freedom. Difficulties, setbacks, and problems may abound, but the basic direction of change is clear.

3) Fascist-style dictatorships. These regimes, especially the one-party Ba'athist governments in Hafiz al-Assad's Syria and Saddam Hussein's Iraq, are modeled on European fascism. In matters of precept, practice, and style, they owe a great deal to the example of Benito Mussolini and, to a lesser extent, Adolf Hitler.

4) Radical Islamic regimes. There are two of these so far, Iran and Sudan. There may be others to follow, perhaps in Afghanistan or Algeria, though the latter possibility now seems to be dwindling rather than growing. Egypt has a potent radical Islamic movement, but the Egyptian political class also has a remarkable knack for maintaining itself in power. Moreover, the threat to the sovereign state posed by pan-Islamic radicalism has been greatly exaggerated. Khomeini used to say that there were no frontiers in Islam, but he also stipulated in the constitution that the president of the Islamic Republic of Iran must be of Iranian birth and origin. In Khomeini's own practice, let alone that [End Page 59] of his successors, the Iranian element remained paramount. Elsewhere, there is a similar disinclination among even the most fanatical Islamic groups to sink their national or territorial integrity into some larger whole.

5) The Central Asian republics. A final group of countries, classifiable more by history and geography than by regime type, are the six former Soviet republics with mostly Muslim populations, sometimes known nowadays as "the five 'stans" plus Azerbaijan. I can venture no characterization of the regimes in these countries, but will only observe that they seem to be having the same problems disentangling themselves from their former imperial masters as the Egyptians, North Africans, Syrians, and Iraqis had with their respective former masters earlier in this century. After the formal recognition of independence comes the postimperial hangover, a period of interference, unequal treaties, privileges, basing agreements, and so on. The big difference this time, of course, is that the former colonial peoples are dealing not with London or Paris, but with Moscow. This may give rise to different results.

Muslims Outside the Middle East

There are also hundreds of millions of Muslims in South and Southeast Asia, but space limitations and my own relative ignorance of these lands lead me to offer only a brief and superficial impression of what is happening in them. Pakistan, Bangladesh, Malaysia, and Indonesia all appear to resemble Egypt or Morocco more than Syria or Iraq, which is encouraging. I say that the South Asian countries resemble the Middle Eastern or North African countries (and not vice versa) for a reason. There are almost as many Muslims in Indonesia, for example, as in the whole of the Arab world, but the lines of influence run from the latter to the former. The historical heartlands of Islam have hitherto enjoyed the kind of influence in the Islamic world that the outlying regions could rarely, if ever, achieve. With the overwhelming numerical preponderance of South and Southeast Asian Islam and the growing importance of the Islamic communities in the West, this may change.

Another relatively small group of Muslims who may matter a great deal are those adherents of Islam who have emigrated to non-Muslim countries in Western Europe and North America. These groups are extremely important, not so much because of what is happening in the countries of their present residence, as because of the impact that they have on their countries of origin. As Muslim minorities go, of course, they are a tiny handful. India's Muslim minority (equal to 11 percent of its total population) is by far the largest concentration of Muslims in a non-Muslim country. Indeed, only two other countries (Indonesia and [End Page 60] Bangladesh) have more Muslims living within their borders. In the Middle East, there is a sizeable Muslim minority in Israel. Ethiopia, a Christian country whose church traces its origins to apostolic times, has a significant Muslim minority, and many other countries in sub-Saharan Africa have Muslim majorities or substantial minorities.

There are also old Muslim minorities in Europe, in the Balkan states, and above all in the Russian Federation itself, which may be as much as 15 percent Muslim.

Like Khomeini amid the Iranian-exile community in the 1970s, some of the political groups that move among the new Muslim communities in Europe and North America are seeking to recruit support for struggles against those in power at home. The separatist movement of Turkish Kurds, for instance, is highly active among the Kurdish population in Germany. The Islamic-fundamentalist movement in North Africa collects money, buys weapons, and organizes in France, and various movements are now using the United States in the same way.

The vast majority of Muslim immigrants in Western Europe and North America, it should be noted, has no interest in extremist or revolutionary movements. On the contrary, these immigrants are increasingly taking part (sometimes as citizens) in the democratic processes of their adopted societies while remaining in touch with their countries of origin. The views that they form as a result of their experience of democracy may well be among the most significant factors shaping the political future of the Islamic world.

Religion and the State

In Islam, as was mentioned above, there is from the beginning an interpenetration, almost an identification, of cult and power, or religion and the state: Mohammed was not only a prophet, but a ruler. In this respect, Islam resembles Old Testament Judaism and looks quite different from Christianity. Christianity, to repeat, began and endured for centuries under official persecution. Even after it became the state religion of Rome under the Emperor Constantine in the fourth century, a distinction was maintained between spiritual and temporal powers. Ever since then, all Christian states without exception have distinguished between throne and altar, church and state. The two powers might be closely associated, as under the caesaropapism of the Byzantine Empire, or they might be separated; they might work in harmony or they might come into conflict; one might dominate for a time and the other might displace it; but the duality remains, corresponding to the distinction in Christian Rome between imperium (imperial power) and sacerdotium (priestly power).

Islam in its classical form has no organizational equivalent. It has no clergy or clerical hierarchy in anything like the Christian sense of the [End Page 61] word, and no ecclesiastical organization. The mosque is a building, not an institution in the sense that the church is. At least this was so until comparatively recently, for Khomeini during his rule seems to have effected a kind of "Christianization" of Iran's Islamic institutions, with himself as an infallible pope, and with the functional equivalent of a hierarchy of archbishops, bishops, and priests. All of this was totally alien to Islamic tradition, and thus constituted an "Islamic revolution" in a sense quite different from the one usually conveyed by references to Khomeini's legacy.

Islamic civilization has produced a wealth of theological, philosophical, and juridical literature on virtually every aspect of the state, its powers, and its functions. What is not discussed to any great extent is the difference between religious and temporal powers. The words for "secular" and "secularism" in modern Islamic languages are either loanwords or neologisms. There are still no equivalents for the words "layman" and "laity." Jurists and other Muslim writers on politics have long recognized a distinction between state and religion, between the affairs of this world and those of the next. But this in no way corresponds to the dichotomy expressed in such Western pairs of terms as "spiritual" and "temporal," or "lay" and "ecclesiastical." Conceptually, this dichotomy simply did not arise. It has arisen now, and it may be that Muslims, having contracted a Christian illness, will consider a Christian remedy, that is to say, the separation of religion and the state.

Of course, I am well aware that the Reformation was a stage in the evolution of Christendom and the Enlightenment a stage in the history of Europe, and I am not suggesting that the past of the West can somehow be grafted onto the future of Islam. There is no reason whatever why the Muslims can or should be expected to follow precisely the same pattern, by the same route. If they take up the challenge at all, they will have to tackle it in their own way. So far, alas, there is little sign that they are willing to take it up, but one may hope.

Turkey alone has formally enacted the separation of religion and the state. Its constitution and laws declare it a secular republic. In many practical respects, however, Islam remains an important and indeed a growing factor in the Turkish polity and in the Turks' sense of their own identity.

As a rule, gradual and unforced change is better than sudden and compulsory change. Democracy cannot be born like Aphrodite from the sea foam. It comes in slow stages; for that reason, places like Egypt and [End Page 62] Jordan, where there is evolution in a broadly democratic direction, seem to offer the best prospects. In Iraq and Syria, an overthrow of the present dictators is unlikely to lead to the immediate establishment of a workable democracy. The next change of regime in those countries will probably just produce less-brutal dictatorships, which might then evolve into reforming autocracies in the Egyptian or Jordanian style. That would not be democracy, but it would be a huge step forward nonetheless.

The places that offer the best prospects for democracy are those where there is a process of gradual change in the direction of freer institutions. Democracy usually evolves out of a movement toward freedom. The liberal democracies of the West certainly did not come about all at once. One need only think of the history of slavery in the United States or the disenfranchisement of women in most of the Western world to see that, even under favorable conditions, democratic progress takes time and effort and may be hard-won indeed.

Imperialist powers deprived most of the Islamic world of sovereignty; the prime demand, therefore, was for independence. Foreign rule was equated with tyranny, to be ended by whatever means possible. But tyranny means different things to different people. In the traditional Islamic system, the converse of tyranny is justice; in Western political thought, the converse of tyranny is freedom. At the present day, most Islamic countries are discovering that while they have gained independence, they enjoy neither justice nor freedom. There are some--and soon, perhaps, there will be many more--who see in democracy the surest way to attain both.

Bernard Lewis is Cleveland E. Dodge Professor (emeritus) of Near Eastern Studies at Princeton University. Formerly professor of history at the University of London, he is the author of many books, including The Political Language of Islam (1988) and The Arabs in History (1960). The present essay is based on remarks he delivered on 13 October 1995 at the International Forum for Democratic Studies in Washington, D.C. Copyright ©Bernard Lewis, 1996.

http://calliope.jhu.edu/journals/journal_of_democracy/v007/7.2lewis.html