While Luhmann did not believe that democracy exists in the sense of “rule of the people,” he did not deny that there is a mode of government in contemporary society that is named “democratic” and that this term usually refers to a “specific structural arrangement” of the political system.1 That is to say, for Luhmann, democracy is not a term that should be discarded. It indeed says something, just not what it actually means. While the idea of democracy seems quite unfit to describe the functioning of politics in today’s world, the term nevertheless signifies a specific way of doing politics. But what exactly does it signify, if not the rule of the people? What is, sociologically speaking, “actually existing democracy”? I suggest that Luhmann’s answer to this question can be divided into three parts: First, being a type of politics, democracy denotes a form of exercising power, or more precisely, of establishing “collectively binding decisions” in society. Second, and more specifically, it denotes a political structure that allows for a continual alternation between government and opposition and thus manages to provide the political system with a balance of stability and variety that has proven successful in the sense of maintaining systemic durability. Third, it denotes a symbolism that provides the political system with what it needs most in order to be able to fulfill its social function of establishing collectively binding decisions in a durable way, namely, it denotes legitimacy.
Political power, according to Luhmann, is the power to make collectively binding decisions. This power typically rests with the government. These decisions may concern anything in society, ranging from regulations on what to wear (on your head, for instance), to what is taught in schools, right up to a proposal to fly to the moon. Such decisions express a specific political will and are made by governments that are authorized to express and enforce this will. Such a definition of politics, including democratic politics, is not radical but rather commonsensical. However, in the context of social systems theory, and in connection with a theory of functional differentiation, it leads to some rather paradoxical conclusions, namely that political power is not very powerful and that political decisions do not decide much.
I discuss these two consequences by referring to two examples of political decision-making in areas that have been fairly relevant (not only) in current democracies, namely governmental regulations and warfare. The power to regulate public institutions—even if they function within other systems—rests with the political system. It can politically regulate universities and hospitals, the legal system, or even the economy by, for instance, supporting private banks or other companies. Nevertheless, at least under democratic conditions, the power to regulate simultaneously leads to the creation of other nonpolitical powers—by politically regulating the education system, the military, or the economy, these very systems are then empowered to function on their own terms. The exercise of political power is not a zero-sum game; it does not absorb all social power, but enforces a dynamic social power structure. Hospitals, for instance, and not the government, have the power to declare people insane or sane; schools have the power to let people graduate or not; courts have the power to declare people guilty or innocent; and the economy has the power to make people rich or poor. These powers are not immediate political powers. Under the conditions of functional differentiation the power of establishing collectively binding decisions within the political system is necessarily limited by the complex social powers exerted by its intrasocial and extrasocial environment. It is limited, for instance, not only by the powers of law and science, but also by the powers of nature (such as, for instance, by the fact that the oil reserves are not inexhaustible). Under the conditions of functional differentiation, that is, the coexistence of a number of autopoietic social subsystems, there is no social center and thus also no absolute power center in society. The diffusion of power does not simply follow a center-periphery structure (which is not to say that there are no center-periphery power structures at all) but a highly complex assemblage of power feedback mechanisms. Political power simultaneously regulates and is regulated by, limits and is limited by, legal power, economic power, media power, and so on.
Similarly, the power of the political system to make important decisions, for instance the decision to go to war, should not be confused with the power to determine what happens after the decision has been made. The power to declare war or not obviously does not translate into the power to decide who wins the war or even to end the war when one wishes to. Under the conditions of functional differentiation, the number of decisive social factors cannot be controlled. A political decision to go to war only necessitates further political decisions that will in one way or another react to decisions made by other systems, such as, once again, the economy, the media, the legal system, and so on. Just as the political system has no power monopoly in society, it has no monopoly on decisions. It has a monopoly on making political decisions and for exercising political power, but these monopolies exist in the context of ongoing and simultaneous decision-making and the exercise of power by other systems. In this sense, political decisions ultimately decide or determine nothing (and neither do those of any other system), they only enable the political system to make further decisions. Once the troops are sent to another country, other decisions must (and can and will be) made, such as what they will do there? how long will they stay? will troop numbers be increased or reduced, and so on.2 Obviously, all these decisions will, in turn, make it necessary for the political system to generate further decisions.
An example of the powerless power of the political system and its inability to come to decisive decisions is the present issue of global warming. There is significant demand for political decisions that are supposed to help or even stop climate change. But how can the political system “control” the climate? The outcome of a war on global warming is no more predictable than the outcome of a war on terror or a war on drugs. The media system, the economy, the legal system, and, obviously, the extrasocial environment (“heaven and earth,” as the Chinese used to say), do not and cannot stop exercising their own powers and making their own decisions once the political system has made its. There is no doubt that the political system can and does exercise its powers and make collectively binding decisions with respect to climate change, but it is quite inconceivable that the political system might determine the outcome of these decisions. If the political system were able to make final decisions, it would, paradoxically, undermine its very function in society. If a political decision were final, there would be no need for further political decisions and thus politics would become obsolete. Final decision-making would be the end of politics—particularly in an actually existing democracy. This is not to say that the political system is entirely powerless or entirely nondecisive, but its powers and decisions are not and cannot be privileged within society and the world.
Collectively binding decisions can be made and political power can be exercised by different political systems, both nondemocratic and democratic. However, it seems that, on the whole, democratic countries have achieved grander success than others in two important respects: they seem to be superior, at least at present, in establishing stable and durable political structures, and they seem superior insofar as they are commonly (which, nowadays, typically means in the mass media system) perceived to be legitimate.
Expressed in the terminology that Luhmann uses in his analysis of the mass media system, it may well be said that the democratic process produces a “stabilization of a relationship of redundancy and variety.”3 Democracy produces stability by allowing a certain degree of instability: governments change, but the system thereby remains intact. This kind of stability could also be called “stability by flexibility”—a kind of stability like that of an airplane wing, which does not break because it is not too rigid. Similarly, one could conceive of economic stability as being based on flexibility. In the capitalist economy, for instance, prices are not stable. Economies that are able to tolerate a certain flexibility in prices and operate with moderate inflation have proven to be more durable than economies that have tried to eliminate inflation altogether, as was unsuccessfully attempted in some Communist countries.
The code that allows for the “stabilization of a relationship of redundancy and variety” in the political system is the distinction between government and opposition. In general elections, our current democracies have established a mechanism that allows for regular alternations of these roles so that those in government and those in opposition frequently (but not too frequently) trade places. On the one hand, there are normally at least two organizations, that is, parties, or candidates, who are seriously competing for power. On the other hand, the individual parties, and even the political personages, tend to remain stable over an extended period of time. We get, so to speak, the best of two worlds: things change, so that the impression that something is happening and that development takes place can be upheld, while there is also a nonchaotic multiplicity of political organizations and drastic differences of ideology, and so there is no need for reorienting ourselves on a daily basis.
In the past, there have been democracies that collapsed due to an overdose of variety. One can think of the period between the two world wars in Germany. The political system of the Weimar Republic suffered from open hostilities between ideological camps (liberals, conservatives, socialists, communists, and fascists) that tended to fight for ultimate supremacy rather than agreeing to coexist. The strong ideological identity of each of the various parties did not favor moderate approaches to politics. In a way, the seriousness of the political parties, their commitment to their respective ideologies, created a spectrum of political variety that turned the political arena into a battlefield and made it impossible for the political system to fulfill its primary function, namely, to make collectively binding decisions. Today, the difficulties that are encountered in trying to establish a democratic political system in countries like Iraq or Afghanistan may also be related to the present struggle to reduce variety. If multiple political forces that represent mutually hostile ideologies, ethnicities, or religious groups compete for power, then political variety may increase to a degree that eliminates all political redundancy. If who governs and who does not is literally of vital importance for all political forces—if it is a matter of life and death—then the very strife for being in government may easily lead to a breakdown of the political system. Paradoxically, a functioning democratic system seems to be based on a good dose of redundancy. It seems that as long as politics are relatively redundant, democracy thrives—as long as no one really cares who governs today, because things may be different (but not too different) as soon as tomorrow. Even a chaotic political landscape, with a great number of political parties and frequent changes between government and opposition—as in Italy—can remain remarkably stable so long as this chaos remains redundant and is not taken too seriously.
On the other hand, a lack of variety may also threaten a democracy and thus political stability. Governments of countries that called themselves democratic but operated factually on the basis of a single-party system typically ended up with tremendous difficulties in legitimizing themselves. A total lack of variety, the seemingly endless continuation of the government, and, in some cases, the suppression of any meaningful opposition lead to a situation in which the political system must pay for its ability to enforce collectively binding decisions by not only losing its perceived legitimacy, but also by the breakdown of functioning couplings with other social systems. The ruling party effectively functioned as a self-serving organization that hijacked the political system for its own self-interest. And other social systems, such as the economy, the law, the media, were unable to develop independently. The economy was supposed to fulfill a political dream, namely equality; the law was supposed to deliver political justice; science was supposed to prove the truth of the political ideology, and so on. Functional differentiation did not function well in these societies. None of the goals were realized. Instead, one organization, the ruling party, had become a parasite on all social systems—and had taken over not only politics, but also the economy, law, science—and, unsuccessfully, tried to steer these. What actually happened was not a steering and controlling of society by politics, but a quixotic fight against the windmills of functional differentiation on the scale of a world society that led to a more or less complete social breakdown.
The basic code that underlies democratic politics is the distinction between government and opposition. It seems that the stability of democratic politics correlates with a “stabilization of a relationship of redundancy and variety”—in particular with respect to the government/opposition distinction. There is no general principle or ratio that would define, once and for all, the specific degrees of redundancy and variety that lead to stability. However, the elimination of one of these two elements seems to present a serious danger to any specific political system. Democracies have developed a highly effective social mechanism for maintaining this stability: periodical elections.
General elections allow for both the determining and the concrete distinction between government and opposition in a democracy for a certain time—and thus redundancy—as well as for the possibility of variety, namely the exchange of roles between the government and the opposition. Perhaps even more important, with respect to the stabilization of the political system, is that they also produce what Luhmann calls a “reflexivity of power relations” (Reflexivwerden der Machtverhältnisse) by, as he puts it, the “inclusion of the audience into the differing-out of the political system.”4 This rather awkward formulation essentially means that through general and free elections the audience (or the public) becomes a factor in both distributing and legitimizing political power; it takes part in establishing both redundancy and variety. The audience is, of course, the electorate. In an election, the voters, who are otherwise restricted to the passive role of observing, are activated, and do indeed, according to strictly regulated processes and rules, contribute to the delegation of political functions. The specific rules of elections vary from state to state and country to country (how the votes are counted, who is eligible to vote, how many votes a voter has, how frequent elections are to take place, which political functions are distributed through elections, how candidates are nominated, etc.) and thus are, to a significant degree, contingent, arbitrary, and subject to change. There is no specific formula that would rationally or logically translate any given election result into a “correct” distribution of political positions and power. There are democratically elected presidents who have received significantly less votes than their opponents. There are democratically elected presidents whose number of votes was less than 20 percent of the overall electorate. There are parties participating in governments who only received about 5 percent of the popular vote while others who received more than 30 percent are in the opposition, and so on. None of these cases are exceptions, but are in fact rather common. This is perhaps why Luhmann calls the election process a periodic invitation to the electorate to “cast dice.”5
The electorate produces certain relations of numbers that, according to the respective rules of the election, may result, for instance, in a certain number of seats in Parliament for certain parties. Voters typically have little knowledge about the individual representatives actually sitting in Parliament (only a few voters will be able to name more than a handful) and have no power to control the actual political decisions of these representatives in between elections. There is also no correlation between whatever intentions individual voters have when they cast ballots and the concrete political effects of the numerical results of the election. It is likely that not a single voter intended the particular distribution of parliamentary seats and governmental positions or the particular set of political decisions that follow an election. This is why an election is, in Luhmann’s view, not much less random than throwing dice. If an election were to be repeated a week later, let’s say for some technical reason, the results would certainly be different even though the world would hardly have changed. The democratic election process includes the audience in the process of distributing political power, but it does this in a highly contingent way. In other words, what the political system does with the numbers generated by an election is up to the political system itself. It is up to the system, according to its own respective rules and procedures, to interpret the numbers politically. Once the numbers have been generated, only the political system, not the electorate, can transform the numbers into political decisions.
Empirically speaking, the democratic election process is not only a random procedure, it is also not really a (and much less the) central factor in the making of democratic politics. In the words of Edwin Czerwick, Luhmann distinguishes between the official (but mostly imaginary) narrative about the “circle of power” and its unofficial but actual circle: “According to the official circle of power, the audience determines through elections the persons that are to become the parliamentary representatives and then elect the government. The government makes the decisions—along with the Parliament and the administration—that the administration then implements and that concern the audience. Within the unofficial circle of power—which, according to Luhmann, is the rule—the public administration prepares the decisions on which the governments and the parliaments then bindingly decide and then justify their decisions to the audience. The audience then reacts with either approval or disapproval.”6
The decisive difference between Luhmann’s analysis of democracy and the common understanding of its “official circle of power” is that, for him, the voters do not represent the people as the source of all political power. They merely constitute the periodically activated audience of the political system which, by this very activity, is integrated into a circle of political decision-making. An election is not constitutive of political power; it is just one element within an ongoing political process. Actually existing democracies are not run by the people and their elections, but are “specific structural arrangements” of a number of political organizations, procedures, and institutions that enable themselves to remain relatively stable by, for instance, allowing for a smooth and often relatively inconsequential change between government and opposition. As soon as the election is lost, the opposition can begin preparing for the next one.
Under these circumstances, the role of the people is primarily (but not “merely”) symbolic, or, in other words, the democratic narrative is a myth that serves the function of closing the political circle described above. Luhmann explains that in a democracy based on the government/opposition distinction, the people are only needed as the unitary formula for alternating differences: “As in the 18th century, the people is only a construct by which political theory accomplishes closure. Or, put differently: Who would notice it if there would be no people at all?”7 The people, as a unit, is a fiction produced by the quasi-magical event of the election process. It is, so to speak, much more a mathematical effect than anything else. The election generates a single result, a single “will of the people”—to which the politicians unequivocally refer and from which they derive their legitimacy. Empirically speaking, however, it would be difficult to identify this will of the people with the will of any given individual. It is not really the people that are needed for a functioning democracy, but the symbolical construct of a numerical relation that is generated by an election.
Luhmann sarcastically alludes to religious beliefs when he says: “In the text of its self-description, democracy still presupposes the ‘people’ as a kind of superior instance within which the miracle of melting the individual will into the general will takes place.” Public opinion is not representative of individual mindsets or interests; it is, in the form of a numerical election result, a “miracle” that produces unity: “What the individuals actually have in mind (meinen) (if anything at all) when they mark ballots, remains unknown. This alone is occasion enough not to… conceive of public opinion (Meinung) as the general expression of the opinions of individuals.”8 What opinions people have, that is, what they “have in mind” or what they “mean” (meinen), has not much to do with “public opinion” (öffentliche Meinung). People’s opinions (in the sense of their thoughts, beliefs, feelings, etc.) are far too diverse and idiosyncratic to be adequately expressed in the form of a ballot. An election result does not indicate the incomprehensibly complex mental or volitional states of millions of people. Rather, it serves as a means by which to establish what is called public opinion (as the successor of the “general will,” which already for Rousseau, should not have been mistaken for a mere sum of individual wills). Public opinion is a communicative construct that enables the political system to distribute positions and roles. It is not a human or mental phenomenon, but a communicative one—it is a social construct (one that, given the current electoral procedures, is produced by the “structural coupling” of the political and the mass media system). By the symbolic act of free elections society legitimizes a government to fulfill this function, and, miraculously, in the form of a “mystic transformation,” generate a general will of the people.9
There is a passage in one of Luhmann’s last published articles in which he stresses the symbolical core of democratic politics:
If one should wish, in accordance with an old convention, to designate this as “rule” of the people, then it seems appropriate to adopt the old distinction between symbolic and instrumental politics. Given this, what we are dealing with can only be symbolic rule. Symbolic are operations that actualize the unity of the separate and thus have an effect on it—which in this case is the unity of the structural differentiation of the organization of the state, political parties, and the audience. The problem with this distinction is that it is connected with a criticism and a demand to cross the boundaries between the symbolic and the instrumental, that is, the demand for “more democracy” and that the latent function of the political election, namely to secure that the future remains unknown, is overlooked.10
This very dense passage is in need of some dissection. First, Luhmann somewhat sarcastically refers to the notion of democracy as a rather obsolete “old convention” that cannot be taken too seriously if it is meant to designate the “rule of the people.” If, however, one is inclined to rescue the term, then one would have to admit, according to Luhmann, that it does not designate an actual or instrumental ruling of society, but a symbolic ruling. By alluding to its ancient Greek meaning, a “symbol” can be understood as something that connects or unites what otherwise would be apart. Luhmann mentions the three elements of the political system that he identified in his later works: the state (government, Parliament, administration), the political organizations (the parties), and the audience, that is, the electorate. These are functionally differentiated, and yet coupled; they are distinct, but cooperate within the political system. Because democracy is a form of “rule” (this term has to be used with caution given that, in a strict sense, there is no “ruling” system in our society) that integrates the audience in its constitution of power, its essential feature is that through general elections these three realms of politics are united into a circle: in elections the audience establishes a numerical result that determines the political strength of the various parties which in turn enables the system to form a parliament, a government, and so on. This unity is, according to Luhmann, of a “mythical” nature, since it is based on such narratives as that of a mysterious “will of the people” that somehow expresses itself in politics. However, operating on the basis of a symbolic myth does not prevent democracy from functioning well.
Problems may arise if the mythical aspect of democracy is taken too literally. The demand for an “instrumental democracy” that constitutes more than a “merely” mythical rule of the people may, in fact, pose a danger to democracy because it is an impossible demand. In a sense, the greatest danger for a democracy is to mistake its symbolic nature as real and let the people try to take over, which may result in anarchy, oligarchy, the despotism of a self-styled “people’s party,” or a meaningless democratic media theatre. Paradoxically, a significant danger for the functioning of democracy is the demand for more democracy.
Such a demand, as Luhmann concludes, implies the urge for a more decisive social impact of democratic procedures, such as elections. But, for Luhmann, the limited social impact of democratic procedures is exactly what makes them so functional. Every election in a functioning democratic system is merely one election before another—just as a season in professional sports always leaves hope for the next one—and does not decide forever the future of a society or a state. If a democratic election is supposed to be an actual instrument for “realizing” the will of the people, this may well be the end of democracy. One of the most important functions of democratic elections is that they do not decide a society’s fate once and for all.
I conclude this section by addressing two issues that go beyond a descriptive analysis of the radical aspects of Luhmann’s theory of democracy. The first issue concerns an implicit conclusion that follows from Luhmann’s analysis, namely that politicizing society too intensely may pose a danger for “actually existing democracies.” The second is the question of where Luhmann actually stood politically.
In existing democracies there is considerable concern regarding political education. From a liberal perspective, political education is not understood as propaganda or indoctrination, but rather as teaching the young about various political options, about the functioning and meaning of democracy, about how to become an active participant in politics, and how to constructively contribute to the further development of democracy. The dearth of political activity among the citizenry is often lamented (in the mass media). When the “alarmingly” low percentage of voters is published, the media and the politicians seem to worry. There are public campaigns funded by tax money to encourage people to cast their votes. It is often stated that which (democratic) party you vote for does not really matter; what really counts in elections, as in the Olympics, is participation. As the slogan goes in German: Dabeisein ist alles (taking part is everything). Historically, concern about political education has had a lot to do with the ideal of democratic citizenship that was a cornerstone of both the American and the French revolutions. Only a politically active member of society is a good member of society, only he (and, somewhat later, she) deserves the honor of living in a truly free and self-managing country. Just as a citizen needs a democracy in order to be a true citizen, a democracy needs politically active citizens in order to be a true democracy. The dependency is mutual.
The reasons why political education (in schools, through the mass media, and through political organizations) is seen as crucial for a functioning democracy are, from a Luhmannian perspective, quite obvious. First, without politically engaged citizens—and just casting your vote on election day already makes one count as politically engaged—the myth of determining the “people’s will” loses its magical power. If only a third of the people vote, then something seems to be wrong with the people’s willpower. If the most important ceremony for generating the people’s will is not well attended, then the people’s will is in jeopardy. This is why taxpayers buy television commercials to encourage themselves to vote. There is no mathematical or procedural problem with a low voter turnout (the number of parliamentary seats can be calculated equally well on the basis of ten thousand or ten million votes), but there may be a mythological one. Only a decent turnout at the election guarantees the survival and strength of the democratic myth. The legitimacy of the government, as well as that of the democratic state as a whole, depends on this myth.
A similar ideal was of prime importance for the political self-description of the totalitarian “democracies” in communist and fascist states. The “people’s parties” that ruled these states conceived of themselves as both the institutionalized form of public political activity and the source of a thorough politicization of society. The insistence on active political participation followed the very same democratic revolutionary demands stemming from the Enlightenment that still inform liberal democracies—the demand for active citizenship. The despotic democracies of the left and the right saw themselves as more democratic than liberal ones, not less so. And they held an even greater belief in the primacy of political activity. In these states every activity was (potentially) political. Going to school, taking part in sports, or going to one’s job, everything was seen as somehow contributing to the strengthening of the people’s political progress. Political liberation was a never-ending project. Students, athletes, workers, and teachers, all were constantly taking part in the politicization and political emancipation of the people. The masses were supposed to permanently engage in political self-education. Politically emancipated workers were supposed to conceive of work as political engagement; politically emancipated athletes were supposed to conceive of their efforts in the context of the socialist reconstruction of society (be it leftist or rightist socialism). These societies also measured their political success by the political participation of their citizens. Political mass ceremonies such as big parades and (rigged) elections were considered central for constructing the democratic myth in these despotic states, just as (free) elections are in liberal democracies.
The forced political activism in despotic democracies produced less stability than the merely encouraged political activism in actually existing democracies. Just as the elections were rigged, and known to be rigged, so were the mass parades. This proved far more detrimental to the political mythology than mere political indifference has in the West. The belief of the ruling party that political engagement could be steered proved to be wrong. These states, paradoxically, ultimately became victims of an overdose of political education and agitation. The political myth that secures a functioning democracy is rather sensitive to communicative climate change: it can be threatened by too much coldness when it is in danger of being forgotten by the people, but also by too much heat when it is prescribed in the manner of an overdose. A myth remains a myth only when it is neither neglected nor turned into something that is supposed to be more than a myth, namely the truth. True democracy is more than the myth of democracy can stand—it can be the death of democracy. Demands for such a democracy and an intensified politicization of society should therefore be met with caution.
There are two common classifications of Luhmann’s political positions, and I find both problematic. The first, and better known, is the accusation that he was a conservative. It is easy to see how this classification came about. In the debate with Habermas in the 1970s, which made him rather famous in Germany, Luhmann was obviously not on the left (as represented by Habermas), so the immediate conclusion by the left was that he was on the right. This is so, because, from the perspective of the left, everything that was not left was an obstacle to social progress and thus necessarily right, or conservative. The various leftist criticisms of Luhmann have been concisely and profoundly listed, analyzed, and refuted by Michael King and Chris Thornhill, and there is nothing I have to add to what they say.11
A second, more appropriate, though still somewhat limited, assessment of Luhmann’s political position is that he is an advocate of functional differentiation, and that this advocacy also makes him an advocate of “actually existing democracy” given that it enables the perpetuation of functional differentiation and thus contributes to social stability and evolution. Simply put, this amounts to saying that Luhmann not only tries to describe contemporary society, but also that he is happy with the result of his description, that he praises our present society (in the West)—though not necessarily in a Leibnizian fashion as “the best of all possible worlds”—and that he affirms current social structures in the hope that they will be sustained. The function of democratic politics within functional differentiation would accordingly be to remain autopoietic and not to interfere too actively in the functioning of other systems. Such a position would be compatible with, though not exactly equal to, a laissez faire approach to politics—and its relation, in particular, to the economy.
I admit that Luhmann often expressed such an affirmative attitude toward certain aspects of functional differentiation and also suggested that it would probably be better for society to simply continue along these lines rather than trying to introduce “revolutionary” changes. He was particularly concerned with the possibly negative effects of a reduction of functional differentiation. I would argue, however, that these normative slips of an otherwise explicitly antinormative and strictly descriptive theorist should be understood more as occasional pragmatic remarks than as ideological commitments. I think that the conclusion reached by Edwin Czerwick overstates Luhmann’s affirmation of democracy: “His [Luhmann’s] systems-theoretically inspired conception of democracy has to be seen as an important attempt at (re)constructing the democracies of the Western world from the perspective of the logic or rationality of political systems that today can only appropriately fulfill their social functions if they practice democracy. Democracy thus becomes in effect the most important condition for the continued existence of political systems that have reached a certain evolutionary stage.”12
Luhmann, on the one hand, was clearly not seeking the abandonment of democracy on the grounds that it is based on a myth or because it would not work well. On the other hand, he did not believe that democracy is the only political structure that would allow for the survival of politics in present society. His defense of democracy is much less specific and, I think, not meant as a plea for democracy. It is meant instead as a warning for those who expect too much from democracy and who want to make society more democratic. He was, I believe, quite disturbed by ideological attempts at taking democracy too seriously and thought that such attempts may paradoxically pose a danger for its existence. I think that Luhmann was no prescriptive thinker, and perhaps even more importantly, a highly non- or even anti-ideological thinker. He looked at the totalitarian political systems that were established on the foundations of leftist or rightist ideologies with great dismay. He was also very sensitive about the problems of “ideologizing” liberal democracy. To criticize current liberal democracy and to demand that it should be “really” democratic smacks of democratic “fundamentalism” and thus of ideological fervor. I think that Luhmann’s affirmation of democracy and of functional differentiation only says that functional differentiation is a bearable outcome of social evolution and that attempts to interfere in evolutionary processes to help it progress are usually not very promising. His defense of democracy is thus not an expression of a substantial prodemocratic ideology. It is based, instead, on the paradoxical insight that politically less charged political systems seem to function in a way that is socially less harmful than those of a highly ideological nature. The benefit of actually existing democracy is not that it somehow realizes any supposed historical goal, but that it allows for social stability. Luhmann believed that “symbolic democracy” had empirically proven to be more successful than any attempt to bring about “true democracy.”