"What It Means to Be Liberal", Michael Walzer
Like all adjectives, “liberal” modifies and complicates the noun it precedes. It determines not who we are but how we are who we are—how we enact our ideological commitments.
Is liberalism an “ism” like all the other “isms”? I think it once was. In the nineteenth century and for some years in the twentieth, liberalism was an encompassing ideology: free markets, free trade, free speech, open borders, a minimal state, radical individualism, civil liberty, religious toleration, minority rights. But this ideology is now called libertarianism, and most of the people who identify themselves as liberals don’t accept it—at least, not all of it. Liberalism in Europe today is represented by political parties like the German Free Democratic Party that are libertarian and right-wing, but also by parties like the Liberal Democrats in the UK that stand uneasily between conservatives and socialists, taking policies from each side without a strong creed of their own. Liberalism in the United States is our very modest version of social democracy, as in “New Deal liberalism.” This isn’t a strong creed either, as we saw when many liberals of this kind became neoliberals.
“Liberals” are still an identifiable group, and I assume that readers of Dissent are members of the group. We are best described in moral rather than political terms: we are open-minded, generous, tolerant, able to live with ambiguity, ready for arguments that we don’t feel we have to win. Whatever our ideology, whatever our religion, we are not dogmatic; we are not fanatics. Democratic socialists like me can and should be liberals of this kind. I believe that it comes with the territory, though, of course, we all know socialists who are neither open-minded, generous, nor tolerant.
But our actual connection, our political connection, with liberalism has another form. Think of it as an adjectival form: we are, or we should be, liberal democrats and liberal socialists. I am also a liberal nationalist, a liberal communitarian, and a liberal Jew. The adjective works in the same way in all these cases, and my aim here is to describe its force in each of them. Like all adjectives, “liberal” modifies and complicates the noun it precedes; it has an effect that is sometimes constraining, sometimes enlivening, sometimes transforming. It determines not who we are but how we are who we are—how we enact our ideological commitments.
The conservative writer Bret Stephens recently defined populism as the triumph of democracy over liberalism. I think he meant the triumph of majoritarian democracy over its liberal constraints. Liberal democracy sets limits on majority rule—usually with a constitution that guarantees individual rights and civil liberties, establishes an independent judicial system that can enforce the guarantee, and opens the way for a free press that can defend it. Majorities can only act, or act rightly, within constitutional limits. Like everything else in democratic politics, the limits are disputed both legally and politically. But these disputes are not settled by majority rule but only by much more complex processes that work slowly over time, making it difficult to overturn any existing set of rights and liberties.
I don’t mean to deny the importance of popular agency. The great achievement of democracy is to bring ordinary men and women, you and me, into the decision-making process. Indeed, the adjective “liberal” guarantees that everyone is in fact brought in—as they mostly haven’t been in the history of actually existing democracies from Athens to the United States. Civil rights and civil liberty are the rightful possession of every member of the political community—Jews, black people, women, debtors, felons, the poorest of the poor. All of us join the democratic arguments, organize social movements and political parties, and participate in electoral campaigns. But even when we are victorious, there are limits on the reach of our decisions. So populist demagogues are wrong to claim that once they have won an election, they represent or embody “the will of the people,” and they can do anything they want. There is, in fact, a lot they can’t do.
What they most want is to pass laws that ensure their victory in the next election, which may turn out to be the last meaningful election. They attack the courts and the press; they erode constitutional guarantees; they seize control of the media; they reshape the electorate, excluding minority groups; they harass or actively repress the leaders of the opposition—all in the name of majority rule. They are, as Prime Minister Viktor Orbán of Hungary has said, “illiberal democrats.” Populist victories are disasters for everyone on the losing side, perhaps especially for liberal journalists, the everyday voice of opposition, who are often falsely charged with corruption or sedition and locked up. And if the populists, despite all their efforts to ensure victory, were ever to lose an election, that would be a disaster for them—since we (liberal democrats) believe that their attacks on the constitution and their violations of civil rights are criminal acts. The stakes are high in this kind of politics. Lose an election, lose power, and go to jail.
The liberal constraints on democracy are a kind of disaster avoidance for everyone involved. They lower the stakes of political conflict. Losing an election leaves you still possessed of all your civil rights—including the right of opposition, which carries with it the hope of victory next time. Rotation in and out of office is a regular feature of liberal democracy. Obviously, no officeholder wants to rotate out of office, but all officeholders accept and live with the risks of rotation. Those risks don’t, however, include repression and imprisonment. Lose an election, lose power, and go home.
The constraints imposed by the adjective “liberal” are understood in exactly this way by the Italian socialist Carlo Rosselli, one of the leaders of the anti-fascist resistance in the 1920s and 1930s and the author of Liberal Socialism, which is one of my texts for this article. “Liberal” describes, he writes, “a complex of rules of the game that all the parties in contention commit themselves to respect, rules intended to ensure the peaceful co-existence of citizens …; to restrain competition … within tolerable limits; [and] to permit the various parties to succeed to power in turn . . .” So Rosselli’s liberal socialism incorporates liberal democracy. For him, as for the democrats he follows, the adjective “liberal” is not only a constraining but also a pluralizing force: it guarantees the existence of “various parties” (which means more than one) and sustains for each the possibility of success. Liberal socialism, writes Nadia Urbinati in her introduction to the American edition of Rosselli’s book, requires “loyalty to a framework that presupposes an antagonistic and pluralistic Society . . .”
Marx argued long ago that the final victory of the proletariat in the class struggle would bring the end of all forms of social antagonism. There would be only one class of equal citizens: one class, one set of interests; nothing important to argue about. Pluralism might still exist, but this would be a pluralism of architectural styles, of literary theories, of sports organizations—definitely not of “various parties” competing for power. The end of antagonism is not something I would look forward to; it would be an illiberal outcome. The triumph of “liberal” communism, if there is such a thing, would look very different.
“Liberal” is a strong adjective, and its constraints are binding, obviously, not only on populist demagogues who win elections but also on our favorite leftists if and when they win. Looking back, we liberal democrats would have had to reject FDR’s court-packing scheme of 1937. That was an example of left populism, but it is comparable to President Trump’s attack on the courts—populism from the right. Yes, judicial decision-making is in part, probably in large part, a political process. So liberal democrats should favor judicial deference to the legislative branch—except in cases involving human rights and civil liberty, where we want activist judges. More generally, judicial professionalism can be an important part of liberal constraint, as we learned when many of Trump’s executive orders were declared unconstitutional by the courts.
It is an old doctrine of socialist militants that the overthrow of capitalism will require a period of dictatorship or, at least, a temporary suspension of civil liberties—a democratic dictatorship of the proletariat or, more likely, an undemocratic dictatorship of the vanguard of the proletariat. Courts defending civil liberty would certainly be repressed—or the judges replaced by vanguard loyalists who do as they are told. Liberal socialists don’t necessarily deny that the final overthrow of capitalism might require measures of that sort. If you believe in finality, then you can’t allow “various parties to succeed to power in turn.” But the coercive measures necessary to prevent that succession won’t produce the socialism we (liberal socialists) want.
The adjective “liberal” means that socialism can only be achieved with the consent of the people; it must be fought for democratically. The struggle has already been long, and there have been and will be compromises along the way—with opponents whose rights we have to respect. Two steps forward, one step back is much better than three steps forward over the bodies of our opponents.
“Liberal” also means that there will be room for socialists to disagree among themselves about the strategy and tactics of the struggle and its short- and long-term goals. So there will be many socialisms, and we should expect to find parties, unions, and ideological groupings of different sorts competing for members and influence within a liberal-democratic framework. As Rosselli argued, the competition will be continuous because, finally, “liberal” means that “socialism is not a static and abstract ideal that can one day be completely realized.” The world changes; new inequalities emerge in place of old ones; we never stop arguing among ourselves; socialist politics is steady work. As Eduard Bernstein suggested long ago, the movement is more important than the end. Or, as Rosselli wrote, “The end lives in our present actions.” The adjective “liberal” is hostile to actual endings.
Socialists are defined, according to Rosselli, by their “active adhesion to the cause of the poor and the oppressed.” But this attachment can’t itself be defined by a comprehensive doctrine. It isn’t expressed in a single correct ideological position that an elite of knowers, a political vanguard, can impose on the rest of us. “Grief will result,” Rosselli says, “from trying to fetter a movement with a development spanning centuries, a movement irrepressibly polyphonic, to a given philosophical creed.” Certainly, we have had a lot of grief over the years from trying to do that. Liberal socialists will be skeptical even about the creeds to which they are committed; a touch of irony is inherent in all liberal commitments.
The revived democratic socialism in the United States today seems irrepressibly polyphonic, though some of the voices are devoid of skepticism—a little too eager to deny the political correctness of all the others. Avoiding grief will require a sustained commitment to the adjective “liberal.”
Nationalists are people who put the interests of their own nation first. Liberal nationalists are people who do that and who, at the same time, recognize the right of other people to do that. And then they insist that all the “firsts” must accommodate each other. They acknowledge the legitimacy and the legitimate interests of the different nations. As liberal democrats set limits on the power of triumphalist majorities, and liberal socialists set limits on the authority of theory-obsessed vanguards, so liberal nationalists set limits on the collective narcissism of nations.
We (defenders of the adjective “liberal”) don’t deny that majorities have rights, that theories about society and economy are politically useful, and that national belonging is a genuine value. But we defend minorities against majority tyranny and ordinary activists against vanguard arrogance. And we defend nations that need states against any opposing nation-states: Kurds, Palestinians, and Tibetans, for example, against Turkey, Israel, and China—but we do this without denying the national rights of Turks, Israelis, and Chinese.
By contrast, people calling themselves “cosmopolitans” condemn all nationalisms and deny the moral value of national membership. Can there be a liberal cosmopolitanism? Since cosmopolitan philosophers recognize a world of rights-bearing individuals, surely they should be called liberal. But most of these individuals set a high value on their particular memberships and identify themselves as French, Japanese, Arab, Norwegian—and not as citizens of the world. The refusal to acknowledge these identifications and to value the pluralism that results from them seems to me illiberal. A global, cosmopolitan state would have to be brutally repressive of (almost) everyone’s national identity or ethnic loyalty. To avoid the brutality, liberal cosmopolitans should make their peace with liberal nationalists. The name of the peace is “internationalism.”
The adjective “liberal” turns nationalism into a universalist doctrine. Yael Tamir, an Israeli academic and sometime politician—and the author of Liberal Nationalism, my second text for this essay—makes the point very clearly when she writes that “the acknowledgment of the importance of cultural membership, and … the assertion of a general right to cultural and national self-determination, must therefore be at the center of any [liberal] theory of nationalism.” One meaning of this “general right” is that all nations must recognize the claims of the others and make room for the nation that comes next.
The English political theorist Thomas Hobbes, thinking of the plight of refugees fleeing famine or persecution, wrote that people living in neighboring states may have to “inhabit more closely together” in order to make room for the refugees. You could call that the moral requirement of a (very) liberal nationalism, but it is a hard demand to make—and taking in refugees rarely requires such a clustering of the natives. There is another demand of liberal nationalism that is easier to make: imperial nation-states that have expanded at the expense of other nations must withdraw from those others and contract their size. I doubt that there is such a thing as “liberal imperialism,” but if there were, it would be an imperialism genuinely committed to its future contraction—making room for the subject nations.
The radical defenders of “Little England” in the late nineteenth and early twentieth centuries were anti-imperialists and, at the same time, good liberal nationalists. “Greater Israel” today is an example of illiberal nationalism, while the defenders of “little Israel” are liberal Zionists—as is Tamir, who invokes the memory of the Girondins in the French Revolution: they wanted to create free nation-states, she writes, “in the territories that France had conquered … ” So also Tamir, in the territory that Israel has conquered.
The adjective “liberal” accommodates the interests of existing and aspiring nations; it also recognizes the rights of minorities within the states that nations create. Most nation-states include ethnic and religious minorities, and their liberalism is tested by their treatment of these groups. Do minority members have the same rights and obligations as all other citizens? Do they have the same economic opportunities? If they are regionally concentrated, do they have a degree of political or cultural autonomy that fits their history and current condition? Are federal arrangements worked out democratically? Canada’s “asymmetric federalism,” which grants greater rights to French-speaking Quebec, is the democratic and collaborative work of an assertive minority and a liberal nation.
The liberal qualification of nationalism makes for the plurality of nations; it is paralleled by the liberal qualification of each particular nationalism. Liberal nations are not created and defined by “blood and soil” or by divine appointment or by a history that starts at the beginning of time and is never interrupted. The blood is always mixed; the geography changes over the years; God isn’t involved; and the history is entangled with other histories. The national story is part true, part imagined, and revisionist historians periodically challenge the going version.
Liberal nations are also not ideologically cohesive; their members are monarchists and republicans, libertarians and socialists, conservatives and radicals. A multinational, multiracial, multi-religious country like the United States is to a much greater degree defined by its politics. It is held together by the commitment of its citizens to a certain political regime and by their recognition of the authority of founding documents like the Declaration of Independence and the Constitution. People who reject that politics or question that authority are called “un-American”—as members of the Communist Party were in the 1950s. “But in a society where social cohesion is based on national, cultural, and historical criteria,” Tamir writes, “holding nonconformist views does not necessarily lead to excommunication.” Right-wing French politicians do not accuse French communists of engaging in “un-French activities.” Or, a better example: “de Gaulle never doubted that Sartre was a respected member of the French nation.”
Communitarianism describes the close connection of a group of people who share a commitment to a religion, a culture, or a politics. Like nationalists, they aim to advance the interests of their community, but the emphasis of their commitment is internal; they are focused on the quality or the intensity of their communal life. Civic republicanism is probably the best known version of communitarianism. Jean-Jacques Rousseau is one of its prophets, and he is definitely not a liberal. Rousseau describes the ideal citizen—a man (women were not yet included) who rushes off to one meeting after another and who derives a greater proportion of his happiness from his political life than from his private life. Citizenship involves a commitment that excludes all others; secondary associations are a threat to the integrity of the republic.
Rousseau’s civic republic is also an illiberal nation-state, as he makes clear in his Government of Poland, where he describes the education of future citizens: they are to study Polish history, Polish geography, Polish culture, Polish literature—and nothing else. “It is education that must give souls a national formation, and direct their opinions and tastes in such a way that they will be patriots by inclination, by passion, by necessity.” Here communitarianism and nationalism are brought together in a radically illiberal union.
I used to teach Rousseau’s politics, and I always felt that his republic was an overheated community. A liberal communitarianism would turn down the heat. It would allow citizens to avoid (some) meetings for the sake of their private happiness—to watch a baseball game, go to the movies, play with the kids, work in the garden, make love, or just sit with friends and talk. It would combine the zeal of participatory democracy with the coolness of representative democracy, so that men and women who don’t love politics would still have a say in political decisions. Its schools would aim at creating patriots by inclination but not by necessity. Students would read novels translated from other languages and study the history and geography of other countries.
Alternatively, liberal communitarians might avoid the civic republic altogether, arguing that the state should be a liberal democracy or a liberal social democracy that provides a framework for a plurality of communities, some heated and some not. This is my own preferred version of communitarianism. Let there be many communities! Of course, some people will choose one, relish the intensity of its common life, and set themselves radically apart from (and perhaps against) their fellow citizens. Identity politics usually follows from a narrow focus on group interest, but it is aided and abetted by an illiberal communitarianism.
Many of us will choose instead to be members of different communities, and the intensity of our commitment will vary across the plurality of our memberships. I can be, all at once, a Jew, a socialist, an academic political theorist, a New Yorker, a husband and father (and grandfather), and an active, but part-time, citizen of the American republic.
I assume that the adjective “liberal” works in much the same way with regard to Catholics, Protestants, Muslims, Hindus, and Buddhists—and I will try to say something about liberal religions generally in a moment. But liberal Jews are different since the Jews are both a nation and a religious community. So we are, or should be, nationally as well as religiously liberal—which means that no theological or ideological, religious or secular commitments can ever be described as un-Jewish. Atheist Jews are not “lapsed” Jews; they are as Jewish as orthodox Jews, since all of us are members of the Jewish people.
Liberal Jews who are religiously identified are no different from liberal Catholics, Protestants, Muslims, and so on. Presumably all these people believe in the legitimate existence of other religions; “liberal” is still a pluralizing adjective. It should work with regard to religion in much the same way that it works with regard to ideology. Liberal believers acknowledge the right to differ—hence the rights of heretics and infidels. Hence also the multiplication of denominations and sects that inhabit the open spaces of civil society and make room for the groups that come next. The members of all these groups will hold their own beliefs fervently, perhaps, but not fanatically. Just as liberal socialists reject the idea of vanguard dictatorship, so liberal believers reject any use of coercion in religious matters. Faith is free.
Liberal believers would recognize not the equal rightness of other beliefs but the equal sincerity of the men and women who hold those beliefs. “These others,” liberals might say, “believe what they believe in the same way that we believe what we believe—and so we can acknowledge the value of their beliefs to them (because we know the value of our beliefs to us). And then we must accommodate the activities, and sometimes the refusals to act, that those beliefs produce.” Accommodation for radical non-believers is probably more difficult, though still required by the adjective “liberal.”
Illiberal religion is easy to describe; it is at least as common as ideological zealotry. Every religion that subordinates women—which means pretty much every religion in its orthodox or fundamentalist versions—is obviously illiberal. Similarly, men and women who believe that the religion or irreligion of the others consigns them to eternal subordination (or damnation) and that they, the true believers, are morally obligated to rescue them from that fate—they are illiberal, and actively so. But the description also fits those who think that rescue isn’t necessary or possible. Jews who believe that most non-Jews will never see the world-to-come are illiberal Jews, just as evangelical Protestants who believe that Jews are doomed to hellfire are illiberal Christians. Far more dangerous, however, are the zealots who aim to “force the end” and establish the messianic kingdom—or the Islamic Caliphate or Jesus Christ’s Holy Commonwealth or any other religious version of the end of secular history. Most liberal believers probably think skeptically or ironically about the end time.
It follows from this account of liberal and illiberal religion that state power cannot be used to indoctrinate future citizens in the orthodox version of Judaism or Catholicism (or any other religion) or to persecute heretics or infidels. A liberal nation-state might emphasize the majority religion in its educational system, since that religion has probably played an important part in the nation’s history. But it wouldn’t turn that religion into a school catechism—any more than liberal socialists in power would turn socialist ideology into a school catechism (as illiberal socialists did in the Soviet Union). And it would also teach the history of local minority religions and of other countries and their religions: the ancient Greeks, the ancient Israelites, the origins of Islam, Chinese Confucianism, and much else. It wouldn’t endorse or promote any particular version of any religion (or of any ideology). There are many ways of being religious—all of them recognized, all of them protected, none of them prioritized, by the adjective “liberal.”
Most people probably think that a liberal Jew or Catholic (or a liberal of any other faith) is a Jew or Catholic who votes Democratic. That is partly right, since the adjective “liberal” is transferable, and so liberal believers are likely to be liberal democrats and (in the United States) liberal New Dealers or social democrats. The Democratic Party has for many years been the favored home for men and women like that. But we have seen (in the past at least) liberal Republicans who defend constitutional democracy, believe in an independent judiciary, feel comfortable in a pluralist society, and expect to rotate out of as well as into political office.
It is an interesting question whether there are groups, parties, ideologies, identities that can’t be modified by the adjective “liberal.” Can you be, for example, a liberal ultra-orthodox Jewish man or a liberal fundamentalist Christian man? The adjectives sit uneasily together. Maybe talented and flexible individuals could accommodate both (they would have to be ready to imagine women as their equals), but I suspect that their fellow believers would say that they had left the fold. Religious dogmatists, whatever the dogma, can’t be liberal. Liberal Republicans are possible, as I’ve just said, even if not currently visible; liberal conservatives, too. I have already raised doubts about a liberal communist; the Stalinist version of communism certainly can’t abide the adjective, though I am sure that there are liberal communists—in the nineteenth century certainly, and perhaps today—who believe in a plurality of communes of different sorts. Fascists and Nazis obviously can’t be liberal. Totalitarianism is the ideal type of an illiberal politics.
A liberal monarchy is possible, which is why we use the adjective “absolute” to describe the illiberal version. A liberal monarch rules alone and doesn’t rotate in and out of office, but he or she recognizes a pluralist politics with constitutional restraints and a plurality of religions. I think that despotism can be enlightened, as some eighteenth-century despots claimed to be, but not liberal. Nor can tyranny live with a liberal modifier. I am dubious about the possibility of a liberal oligarchy, but a liberal aristocracy (along Jeffersonian lines) is conceivable, so long as membership is not hereditary. Competition in excellence and virtue and the social mobility it produces might have some of the features of rotation in office.
Most of these possible uses of the adjective “liberal” are not relevant today. But the ones that I started with seem to me not only relevant but critically important for contemporary politics. We need liberal democrats to fight against the new populism; liberal socialists to fight against the frequent authoritarianism of left-wing regimes; liberal nationalists to fight against contemporary xenophobic, anti-Muslim, and anti-Semitic nationalisms; liberal communitarians to fight against the exclusivist passions and fierce partisanship of some “identity” groups; and liberal Jews, Christians, Muslims, Hindus, and Buddhists to fight against the unexpected return of religious zealotry. These are among the most important political battles of our time, and the adjective “liberal” is our most important weapon.
Michael Walzer is editor emeritus of Dissent.