1. What Does It Mean to Govern?
We have grown so accustomed to talking about government, to feeling reassured because societies are governed, and even to putting up with government occasionally, that we have lost sight of a more radical reality that the normal condition is no government. Most things have transpired throughout the history of mankind and are occurring still today with no visible or invisible hand to govern them. Usually it is destiny and chance, the disorderly course of events, the fact that more things happen than what are anticipated in the rules, are codified by the law or have derived from explicit political decisions. Any examination of what it means to govern should begin with this finding, which challenges our usual perception of things.
The fact that we human beings govern ourselves is not something that can be taken for granted. The question “why not anarchy?” (Nozick, 1974: 4) is at the source of every justification of the act of governing. It is asked by a liberal like Nozick, but it could be also be shared by an anarchist, and in between there is a whole variety of complaints claiming that there is too much government or too little, there are interventions that must be justified, decisions with questionable legitimacy, people who think that there is too much regulation and people who lament the ineffectiveness of regulations, there are democratic deficits that are deficits of demos or cratos… But long before we have resolved the question of whether we are being governed well, we must answer another more worrisome question as to whether we are being governed at all, whether governing is possible and what that means.
Some years ago, Niklas Luhmann proposed to turn our view of reality upside down and spoke of government as something “improbable”, triggering a discussion that years later Baecker (2004: 55) summarized as follows: “from the need to govern politically, we assumed that it would also be somehow possible”. People then took another look at the finger without seeing what it was pointing at, and theories abounded as to how to improve something that was being taken for granted, proposals for good governance that did not want to consider the question of government. Since then, there has been a steady increase in the number of proposals about how to radicalize democracy (improving participation, transparency, accountability, etc.), but of course they all assume that contemporary democracies are governable, as if the current conditions of a global knowledge society would only pose problems of democratization but not of governability. And what if we were proposing to improve something that cannot be done, or that at least can no longer be done as we have been accustomed? If we do not want to spare ourselves any inconvenient possibility, we should begin any investigation into how to govern in advanced societies by asking ourselves what it means to govern per se, why it is something so improbable and fraught with failures.
So what does it mean to govern? To govern is to be unsatisfied with the natural course of things. Straight off, because it presupposes politicizing an area of reality, understanding that what happens to us can be configured differently from how it is presented to us (by destiny, tradition, uncoordinated individual decisions or the mere aggregation of events). That is why Bobbio (1976) said that democracy is subversive: because it alters the traditional stratification of power and opens to discuss something that has been understood as commonplace or natural for years and centuries. But there is another type of dissatisfaction that is more forward-looking than backward-looking. In this sense, the action of governing is explained because we want a future situation that is different from the current one. Governing is a type of intervention that aims to correct, to direct, to intervene, to improve, to balance and change processes or states of things that do not satisfy us in their current form. For Luhmann (1989: 5) , every governmental action seeks to forestall the situation we would be in if there were no governmental intervention. We intervene to achieve a status quo (equality or regulated markets, urban planning, for example) that would not occur if there were no government.
2. Failures of Government
To understand what it means to govern, we must realize that it is an action accompanied by so many failures. Those interventions, from which we expect so much in terms of breaking with the past and creating a different future, are often times unsuccessful, costly, incorrect, discouraging and even catastrophic. We could say that it is almost normal for an intervention not to have the expected success and especially that strategies that have trivialized the problems will fail, and the problems will ultimately get worse. Politics always has a tragic aspect to it, such that we do not always achieve everything, and we do not always achieve something without causing undesired effects. To phrase it in less dramatic terms, we could blame excessive expectations about the transformation of societies or incorrect decisions that have had less effect than what was desired and even some effects opposite to what was desired.
When I speak about failures, I am referring to the extraordinaries and the ordinaries ones. Among the most outstanding failures, I can cite those that have been grouped under the expression “tragedy of the commons” (Hardin, 1968) to refer to situations in which an interested individual behavior can be contrary to the common good when certain resources are shared. Here we could mention global warming, environmental exploitation or financial bubbles. It is true that public policies have unquestionable successes, which we often take for granted, but there are also very disheartening global failures involving the eradication of poverty, the limited effectiveness of aid to development, the imbalance of freedom and equality or demographic explosion. Government failures also include the failures experienced by anyone who has tried to change a status quo, involving all the factors at play in configuring the social fabric… and which put up a remarkable resistance when someone seeks to transform them or simply to reshape them. Human decisions and plans frequently end up badly or are unsatisfactory compared to the expectations because the long-term side effects were not taken into account or because the measures were too ambitious or too weak, etc. Whether we are referring to dramatic global issues or to the petty impotence of local politics, the action of governing is always accompanied by failure.
What are these failures due to? Before discussing the conditions needed for successful political intervention in a complex society, I would like to emphasize what we should not do, the methods of altering reality that cannot achieve the desired effect because they have not sufficiently considered societal resistance. When understanding the act of governing, the model of “trivial machines” is obsolete, which presupposes naively that a given intervention has a completely foreseeable social result, such that societies can change in a direct and linear way. When we are talking about managing non-trivial systems, it is not enough to just push a button, issue an order, make a law or prescribe a medication.
Here, on the one hand, we run up against the limits of direct intervention: legislation usually comes late and it is an illusion to think that a legislative initiative can substitute for the ethical infrastructure of a society; at other times, governments try to fix problems with vertical interventions and punitive institutions, which are expensive and not very effective; there are also ineffective interventions, such as the issue of whether the injection of capital by central banks is useful or not, and how much it should be, because that is a short-term view of the problem. Its limited effectiveness has to do with the fact that usually they are one-time interventions on systems which have not been acted upon sufficiently to provide them the proper stability.
At other times, the failure has to do not so much with inefficacy as with causing undesired effects. There are disasters that are not the mere result of bad luck but rather the result of inappropriate institutional interactions and configurations. On many occasions, they are the result of a failure to understand the counter-intuitive nature of the behavior of systems. For example: penalties that can foment social conflict, security measures that encourage terrorism, promotions of innovation that contribute to economic decline…
Finally there are what we could call impossible areas for imperative politics because the desired results are unattainable simply because they do not lend themselves to being ordered: from the public sector, many things can be done to foster innovation or the creation of jobs, but if we were to try to do it directly and immediately, we would be creating the opposite of what we want: a subsidized society or a society that obeys the instructions of the authority, that is to say, a society in which there cannot be economic vitality or cultural creativity. The very logics of those systems are what impede vertical intervention and control.
We would spare ourselves many disappointments if we started by acknowledging the possibilities of governmental action and its limits, if we understood the logic of government of complex systems. I am not saying this to lower our expectations of intervention in society (on the contrary), but rather so that we will understand its logic.
3. Government as Self-Government
What if, ultimately, government―above all government of others, hetero-government―were impossible, in a real and normative sense, that we cannot or should not govern others or, stated positively, we can and should only govern ourselves? The entire history of political thought is marked by a paradox that has been formulated in various ways but could be summarized as follows: the only legitimate government is self-government. This paradox has been formulated in many ways: as an anarchical exigency (challenging authority), as a melancholy lament against ungovernability (societies do not allow external management), as a neoliberal program (dismantling the State to return the leadership to civil society)… From this, there is a right-wing version (liberalization, evocation of civil society, minimal state) and another left-wing version (citizen participation, evocation of social movements), but both versions, in different ways, exhibit the same conviction that government of complex societies is limited and requires the leadership or collaboration of the governed, who cannot be understood as passive recipients of governmental intervention, either because they do not tolerate excessive intervention or because they aspire to take active part in the decision-making processes. With this thesis, I am not maintaining any specific ideological option, but rather I am pointing to the existence of an axis on which the various options can be placed. The ideological spectrum goes from the right wing (which would be characterized by the desire not to be bothered by government) to the left wing (whose fundamental concern is not to be excluded from governmental decisions); what they all share, from one end to the other, is the conviction that a mature society is a society that governs itself, whether we construe it in liberal or socialist terms, from libertarian neoliberals to the democratic radicalism of the new left.
Our political theory comes from the Hegelian juxtaposition of state and civil society, whose relationship is construed vertically and in one direction. A large part of our current discord comes from the difficulty in understanding the new societal logics and proposing procedures for its legitimate government. The dominant theories of government have not developed a concept of government in keeping with the current transformation of our societies, with their increasing complexity and the reality of distributed knowledge.
One of the conceptions that we must break from is the conception of understanding power as the property of one individual actor and we must instead consider it a social relationship that structures the possibilities of action of the various actors. Understood as a relationship, power lies not only with the holder of sovereignty or the recipients of government action, since both of them mutually influence, limit and enable each other, in relationships that are not symmetrical, of course, but which also cannot be described as a relationship between authority and subordinates, or even a strict separation between subject and object in governmental action. There is the Official Gazette of the State, certainly, but also the sovereign governor who anxiously scrutinizes surveys or toadies to the people (who are also sovereign). We have to make a space in political theory for bilateral relationships, horizontality, self-limitation and network structures.
We need to think of the State less as a vertical and unitary actor and understand it more pluralistically as an institution that articulates a sort of game and works the various social problems with help from the various social networks that come into play (from their various administrative authorities to NGOs and social movements, including social subsystems like universities, mass media, the health system or economic agents, involving political parties and labor unions) (Mayntz, 2001: 18) . In addition to the typical forms of vertical control, we would need to explore other possibilities that are nonhierarchical, indirect, informal, negotiation-based, and decentralized, like those pertaining to incentives or self-regulation. We would be talking about finding functional equivalents of power in complex, horizontal societies.
When we do not understand this complexity of exercising sovereignty, we end up with theoretical perplexity and practical inefficiency. Any theory of government must realize the structure of mutual influence between the governors and the governed, between the political system and society. The desire for society to play a leading role is especially intense in knowledge societies, but it is part of the same social logic as always, made acute in contemporary political revolutions. I would like to illustrate this confusion in a text by the historian and politician Guizot, written in 1821, in which he denounces the error of power, which is unable to recognize what its true resources are: “there are ministers, prefects, mayors, teachers, soldiers, everything that it calls means of government. When it possesses them, when it has arranged them in a network over the landscape of the country, it says that it is governing and is surprised to still find obstacles and that it does not possess its people as agents (...). This is absolutely not what I understand by means of government. If this were sufficient, what is the government complaining about? It has such machines; it has never seen so many of them and they have never been so good. However, the government says over and over that France is ungovernable, that everything is in revolt and anarchy; it is dying of hunger in the midst of its forces, like Midas dying of hunger in the midst of his gold. The fact is that the true means of government lie not in these direct and visible instruments of action of power; they reside within society itself and they cannot be separated from it. Human society is not a field that comes to be exploited by a master. It is futile to seek to govern it by forces exterior to its own forces, by machines set up on its surface, that do not have roots within it and do not establish the principle of their movement within it” (Guizot, 1821: 128-139) . The fundamental problem of the post-Napoleonic era, Guizot (1821: 121) said, was “to form the government by the action of society and the society by the action of the government”: the government cannot be effective if it does not interact with the things that constitute the life of the country. The government’s resources are found in the interests, passions and opinions of the society.
Any claim to govern must be understood within the praxis of the self-organization of society and civic empowerment. In order to identify and elaborate political problems, governments need distributed knowledge from all sectors of society, from science to mass media, and the consumers; let’s think about how financial markets are regulated (the Basel accords), which are an example of that “reflective governance” (Schutter & Lenoble, 2010) in which the regulators and the regulated, public and private actors, design possible scenarios in an environment of collaboration in which knowledge and legitimacy are exchanged; states that request help from hackers to solve problems of communication security, i.e. those who know best how to make a system secure are those who have transgressed it; the economy cannot work without certain behavior on the part of consumers (a certain level of confidence, innovation, moderation of risk) which are attitudes that cannot be forced upon people, but can only be encouraged or fostered; health management depends more on the creation of healthful habits than on health planning, i.e., the health system is not about trying to cure so much as it is about avoiding being in that situation in the first place; even law and order would be impossible if there were not a subtle monitoring of neighbors and cooperation of citizens, as if the police were ultimately recognizing their ineffectiveness. These are some examples of government action in which the official office-holders are the ones governing, but in fact the governed seem rather to be the ones who actually exercise the authority, as if it were the political transposition of that lack of distinction between producers and users that we owe to the internet.
This bilateralism or mutual influence of government action has a democratizing power that the theories of government of a complex society should explore. Authoritarian regimes do not fall because an opposing force topples them but rather they usually explode because they understand government action so unilaterally that they are lack the bottom-to-top information that they need in order to govern. If democratic systems are more adaptive, it is because they favor open spaces in which people participate, criticize and protest, which are things that definitely inconvenience the people in power, but which ultimately provide information without which the people in power lose a sense of reality and end up governing badly.
All this can also be formulated in the language of government of social subsystems. Adapting to the semantics of the system that one seeks to govern is an essential condition if the act of governing is to be a success. Any government operation must reconcile itself to the autonomous operating method of the system that it aims to govern. Luhmann (1984: 654) has posited it as a cognitive strategy, but it could be formulated pragmatically: “the object can be investigated only by putting its self-reference into motion, that is to say, by taking advantage of its own motion”. It would be about taking advantage of, rather than combating, the inherent tendency of complex systems to self-organize. The most promising alternative for government of complex systems is “guided self-organization” (Helbing, 2015: 72) , an action oriented toward permitting their self-organization and impeding only those dynamics that endanger the self-organization of other subsystems.
All the difficulties of governing derive from the fact that it is always a contradictory operation because it aims to modify something that in turn it is required to respect. If society were not so suspiciously inclined to organizing itself apart from government, if society were simpler and the democratic exigencies less severe… But this paradox of the act of governing also characterizes other types of intervention in complex systems, in which a form of authority presents itself to an environment where things tend to order themselves without that authority. Doctors are confronted with the logic of an autonomous body, therapists are confronted with the logic of the psyche, consultants are confronted with the logic of organizations, professors are confronted with that strange logic that nests in the heads of students, and politicians are confronted with the contradictory logic of societies. Governing means accepting the inherent limitations of the intervention and recognizing that, basically, as Rousseau well knew, government is always self-government.
4. Indirect Government
In general, the task of governing has been understood as control, management, power, leadership, hierarchy, sovereignty and order. All of that can no doubt be found in any government action. However, that perspective has prevented us from exploring other possibilities that have to do with government of context, indirect government, soft power, cooperation, regulation or shared sovereignty. I propose that we investigate the possibilities of what we could call indirect government, which moreover is not something alien to many of our current practices, but which needs to be theorized and primarily legitimized with greater precision. Here we could find many solutions to some of the paradoxes that I have mentioned, primarily the inevitability and impossibility of governing. That way, it would be possible to achieve a certain reconciliation between effectiveness and legitimacy, technocracy and populism, the reasons of the technique and the reasons of the people. And we would clear the way to a harmonization of freedom and authority or, at least, a framework in which to realize both aspirations at the same time, as tense and contradictory as it may need to be.
We have a clearer knowledge of what is impossible than what is politically possible. Systems theory taught us a long time ago that social systems have their own semantics and do not allow themselves to be governed from outside. It does not make sense (and it will not be heeded) to tell the economic system to consider different values other than money, just as an artist will not accept moral censorship, nor can we be constantly demanding financial profitability from scientists, and it is just about ridiculous to accuse a politician of trying to get votes. What else can they do? The only limitation of all those self-referential logics is that they all can discover by themselves their possible self-threat when it is taken to the extreme and forgets that it must coexist with other systems that have other priorities: financial markets accept political regulation because they are aware of how harmful absolute financialization of the economy can be; the artist exhaust the expressive possibilities of transgression and see that beauty has other possibilities; the scientist accept some social obligations even if it is only because he thereby obtains the financing for his experiments, or the politician discover that the public trust on which he ultimately depends can force him to certain self-limitations.
I am referring to forms of governing that could be encompassed under the metaphor of “circonvenir” introduced by Merlau-Ponty (1949; 1960) , which suggests, in a strong sense, the act of misleading or deceiving with trickery, but also circumventing, circumlocution or delimitation. “Neither pure fact, nor absolute right, power does not compel, does not persuade: it circonvient and it circonvient itself by appealing to freedom rather that by terrorizing”. In this sentence, we have the full semantic scope of indirect government, its incentivizing nature and its infrastructural functions.
Many of the components of democratic governance, of what we could call “soft government” (Göhler, Höppner, & De la Rosa, 2009) , are not per se exercises of power in the classic sense but rather incentives that are implemented through rational argument, the expectation of mutual benefit or the fear of damaging one’s own reputation. Hard power (without knowledge, without persuasion, unilateral, like an order) is not a suitable way to govern highly complex systemic processes. It is not a suitable way to govern the financial system, or to protect consumers, reduce pollution or provide security. The more politics depends on developing processes to form an intelligent political will, the more archaic the idea of sovereignty or strong power seems.
From this perspective, political reason appears as a motivating reason, which aims to encourage some options and dissuade people from others, in accordance with the desirable end goals, knowing that most of the time it concerns behavioral changes that people cannot be forced to make, that they can only be encouraged, incentivized or convinced to do. This concerns infrastructural measures that serve “to channel the actors in a certain direction by manipulating their structures of preference and opportunity” (Rosewitz & Schimank, 1988: 322) . Examples of what Lindblom (1977) called “preceptorial systems”, would be behavioral changes in relation to health or traffic, attitudes toward immigration, cooperation on ecological programs or energy saving, improving the level of education, fostering innovation or environmentally sustainable forms of consumption.
In recent years, forms of governing have been proposed that resemble indirect or soft government and that require a proper conceptual justification. These suggestions could be grouped in three areas: 1) soft and argumentative power in relations among subjects; 2) government of context referring to social subsystems and 3) infrastructural conditioning of choices.
1) The theory of “soft power” comes largely from a new conception of international relations that aspires to replace unilateralism with diplomacy and multilateralism (Nye, 1990; 2004) . It is a theory of power because it tries to make others do certain things, but it is soft power because it seeks to do so without coercion. In the European Union, the Open Method of Coordination (OMC) was introduced in the 1990s as a set of procedures for governing a Union made up of sovereign states which had to govern themselves in a post-sovereign manner. The effectiveness of its results is questionable but, in any case, it sought to put into play non-binding goals and directives, commonly agreed indicators, benchmarking and persuasion in areas in which EU institutions did not have jurisdiction or had very limited jurisdiction. Ultimately, forms of “soft power” are all the persuasive procedures, which replace unilateral domination with procedures of reciprocal influence in deliberative spaces in which the subject and the object of government swap their functions.
2) Other strategies of indirect government can be grouped in the concept of “government of the context” (Willke, 1989) , especially when we refer to the governments of differentiated social subsystems. This would consist of combining the self-organization capacities of each one of the systems and the possibilities that politics has to establish the conditions and the frameworks in which such autonomous systems are to be deployed. Thus the logics of distributed intelligence would be combined, in a new balance, with those of the last word, which would belong to the political system, crowd wisdom and Kompetenzkompetenz, horizontality of the masses and verticality of politics; we could liberate politics from the weight of many decisions for which it has hierarchical authority but lacks cognitive competency, while we would spare autonomous systems from the errors that come from their inability to see the compatibility of the whole, its side effects and miscalculated risks.
3) A third group of strategies of indirect government is contained in the idea of “nudging” with which behavioral sciences, political theory and economics defend the possibility of achieving unforced acceptance in the behavior of individuals or groups more effectively than legislation by coercion (Thaler & Sunstein, 2008) . “Nudge” is an elbowing, a prod or push, which seeks to promote conduct that benefits the common interest, moderate and subtle procedures to get people to place the public interest above their individual desires. It’s less about ordering that it is about encouraging, promoting, involving, shaping, prodding, and stimulating. For that reason, it has been possible to characterize this strategy of government as “laissez-faire activism” (Colander & Kupers, 2014) or “soft paternalism” (Sunstein, 2014) . There can be a public conception of the public interest and it can be promoted (versus the libertarianism that underlies the merely aggregative logics of neoliberalism or the supposed rationality of homo economicus), but this promotion does not trust in the old forms of authoritarian paternalism. Everything is staked on the importance placed on the political design of the “choice architecture”, which does not force anyone to do anything and maintains freedom of choice, but makes it easier to get people to do what is considered most politically appropriate. We are referring to a multitude of procedures that do not impede choice while they steer it: information, warnings, awareness, procedures, protocols, stimuli (pro or anti-cyclic), emulation and comparison (ranking), “default rules,” the whole area of regulation and incentives, in relation to health, traffic, consumption, debt, the environment… In these forms of soft or indirect conditioning, governments (whose capacity to compel is limited) and individuals (who would be overburdened with no orientation to guide their decisions) are released.
5. Critique of Indirect Political Reason
If the two preceding hypotheses are true―the difficulty of governing by means of direct procedures and the inevitability of governing in any way if we want to avoid certain catastrophic chain reactions risked by societies that entrust everything to mere aggregation―then the territory of what we have called indirect government presents itself as a space full of possibilities, yet not exempt from certain ambiguities, which I can only indicate here as a brief conclusion and suggestion of possible future investigations.
The task of governing would consist of “getting meta-conditions right” (Colander & Kupers, 2014) . This is a normative ideal but also a matter of facts. Government of societies is increasingly done by means of indirect procedures and we trust less and less that real change in societies can be the inexorable result of an order. Even legislation seems like an insufficient means of promoting true social transformations.
The space of indirect conditioning is a place for promoting the common good for the same reasons as for exercising domination. Good politics rely little on direct imposition, and bad politics also explore the possibilities of this type of procedure. The logic that promotes responsible consumption and the logic that creates those contagion effects that are the source of financial bubbles are formally identical; there is nothing to assure us that someone who can condition our behaviors for good cannot also do so for evil; the procedures to get us to do the right thing so as not to damage our reputation also are the source of the worse forms of competitiveness; we are warned of the risk but also our fears are manipulated; there is a call to innovation and we all end up innovating in the same way; the difference between argued persuasion and propagandistic manipulation is very tenuous. The promises of “nudging” coincide with the subtle threats of the method of discipline that Foucault (1994) denounced with the term “gouvernementalité” or the configuration of hegemonies elucidated by Laclau and Mouffe (2001) through which a certain normalcy is established with no need for anyone to order it directly.
Therefore, this exploration of the possibilities of indirect government must be accompanied by a critique of indirect political reason. Although the conception of explicit and vertical power is unable to discover indirect possibilities of governing, neither is it especially sensitive to forms of domination that are not based on naked imposition. With these forms of conditioning, there occurs something similar to what occurs with spell checkers, which enable us to avoid certain errors but oftentimes induce us to say something different from what we intended to say. Future debates about the democracy of our societies are going to be less about explicit emancipation and repression than about the implicit conditioning that is contained in the architecture of our decisions.